Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: Petrus on August 27, 2015, 02:19:36 pm
-
I made some comments on another photography site about the new Leica S 007 being a medium format camera (ok, it is). My point was that the old definition of Medium Format being anything larger than 135 sized sensor is not really relevant in the digital age.
During the film era things were simple: with the same or similar emulsions available in different sizes, you were able to increase the IQ simply by using bigger piece of film. With that came the effect of diminishing DOF with the same aperture lenses, etc. The MF look, if you wish. And LF look with large plates.
If we put this "scientifically" the formula which defined the "look" had only two variables: film size (which governed the IQ) and the lens aperture, larger formats having less DOF.
Now we have another variable in the equation: Sensor quality. Sensors can not be judged by surface area only, like film was (in a simplified sense). This makes the "format equation" much more complicated, because we can get the same MF quality with small but high resolution sensor with large aperture lenses. So a 135 size sensor (like Nikon D8xx, Sony & new Canon) with fast lens can give you optically the same picture as slightly larger "mid format" (?) sensor with slightly slower lens, as larger format lenses tend to be. As we know, new 135 sensors beat many MF backs and sensors both in resolution and all in DR. With a slightly faster lens all shallow DOF effects of MF cameras can be duplicated with smaller sensors also.
So are we stuck with the old film era definition of MF? How about defining digital MF as something with more than certain resolution (30 MPix?) and the availability of lenses for the system which give the same minimum DOF as f/1.4 on a 135 sensor, or f/2 on a "MF" sensor (whatever it is)? It should be the end result that matters, not the way it is achieved.
-
With a slightly faster lens all shallow DOF effects of MF cameras can be duplicated with smaller sensors also.
This has not been my experience.
-
There's a bit too much prestige in MF gear still I think. Too many want it to be a separate exclusive genre.
In the longer term however, I think forums like this will be called "high resolution photography" rather than "medium format" and include both high res smaller sensors and larger. The shooting techniques, challenges and questions are the same when shooting high res, regardless of sensor size.
That said there are today quite some MFD gear specific questions, and tech cams are quite a different animal, and many are using older second hand gear etc, this forum is good for that to share tips for this relatively rare gear.
-
This has not been my experience.
It is really just descriptive geometry at work. Of course lenses have their own characteristics, but that is not directly tied to the size of their image circle.
-
It is really just descriptive geometry at work. Of course lenses have their own characteristics, but that is not directly tied to the size of their image circle.
It's not just the image circle and it's not just the lenses.
There is a large variety of factors, that COMBINED makes medium format shooting a whole different animal than shooting smaller formats.
Starting from the more deliberate way of planning a shot; (Can be a plus or minus, depending on your POV) more so if you're working with tech cams , the out of camera color renditions including better color seperation and tonal gradation, the shallower DoF for a given aperture and FoV as you mentioned, the post production workflow, per pixel "bite", the thickness of the files... all these combined makes it a very different way of working and output.
Of course, those who like to break everything down to numbers would argue that none of these are real, but these are my observations from actually using these formats to make pictures.
And megapixels have nothing to do with it either. Owning a 40MP MF back and a 36MP 135 format camera, I can consistently see that the difference between them is a lot wider than the 4MP would suggest. For one, the MF fileholds itself together very well, even when blown up to 200%. The 135 file completely fals apart in comparison.
(https://scontent.xx.fbcdn.net/hphotos-xap1/v/t1.0-9/1521273_10151942916528196_800461168_n.jpg?oh=3f246735a8ed28b802790f4f677cdd84&oe=5672F6C9)
(An example of the "Global color" look that is prevalent in most 135 cameras: The one that BC always talks about, vs better gradations in skintones on the MF file)
Even with low megapixel older backs, this can be seen. There is a "Fat pixel magic" thread in GetDPI with plenty of image samples that show this. At base ISO, a 16MP file from an older MF back is still superior to a 20+mp 135 camera in my books (Say, like the 6D).
So no, I don't believe in re-defining MF based on some arbitrary megapixel number or whatever. The old definitions still work great for me.
-
+1000
-
Of course, those who like to break everything down to numbers would argue that none of these are real, but these are my observations from actually using these formats to make pictures.
thank you for being modest and not invoking ghosts of "6 stops of DR advantage" and "16-bit raw files" staying with just "the thickness of the files" ;)
-
It's not just the image circle and it's not just the lenses.
There is a large variety of factors, that COMBINED makes medium format shooting a whole different animal than shooting smaller formats.
Starting from the more deliberate way of planning a shot; (Can be a plus or minus, depending on your POV) more so if you're working with tech cams , the out of camera color renditions including better color seperation and tonal gradation, the shallower DoF for a given aperture and FoV as you mentioned, the post production workflow, per pixel "bite", the thickness of the files... all these combined makes it a very different way of working and output.
Of course, those who like to break everything down to numbers would argue that none of these are real, but these are my observations from actually using these formats to make pictures.
And megapixels have nothing to do with it either. Owning a 40MP MF back and a 36MP 135 format camera, I can consistently see that the difference between them is a lot wider than the 4MP would suggest. For one, the MF fileholds itself together very well, even when blown up to 200%. The 135 file completely fals apart in comparison.
(https://scontent.xx.fbcdn.net/hphotos-xap1/v/t1.0-9/1521273_10151942916528196_800461168_n.jpg?oh=3f246735a8ed28b802790f4f677cdd84&oe=5672F6C9)
(An example of the "Global color" look that is prevalent in most 135 cameras: The one that BC always talks about, vs better gradations in skintones on the MF file)
Even with low megapixel older backs, this can be seen. There is a "Fat pixel magic" thread in GetDPI with plenty of image samples that show this. At base ISO, a 16MP file from an older MF back is still superior to a 20+mp 135 camera in my books (Say, like the 6D).
So no, I don't believe in re-defining MF based on some arbitrary megapixel number or whatever. The old definitions still work great for me.
Between the two examples, I prefer the look of the small-format image in this particular case. Yes, the medium format image has a slightly greater sense of depth to it, but I actually do not find that effect to be preferable for this particular image. I will say, however, that the small format image has a warmer overall color balance, which I find more appealing here, and I think it makes the other version look rather cold and unappealing by comparison. So, perhaps the difference in color balance is having some effect on my overall perception of the two images.
-
thank you for being modest and not invoking ghosts of "6 stops of DR advantage" and "16-bit raw files" staying with just "the thickness of the files" ;)
I am a "reformed" velvia 50 shooter who started on the D70s in digital. So for me, anything over 6-7 stops of DR is a bonus, not a prerequisite. :)
Regarding DR, my D800 definitely can milk a lot more out of the shadows at base DR, but in a real world scenario, I have never really required to do this. In the studio, you're always working with optimal lighting and for landscaping, I work with filters and make some optimal exposure choices. Of course, when I want to do extra long exposures, I use the D800. This is why we have horses for courses, innit?
Below are two throwaway shots that show what the Credo's files can do in post. I am more than happy with that.
(https://scontent-lga1-1.xx.fbcdn.net/hphotos-xfp1/t31.0-8/s2048x2048/1534943_10151947650268196_1035251969_o.jpg)
(https://scontent-lga1-1.xx.fbcdn.net/hphotos-xfp1/t31.0-8/s2048x2048/1090908_10151947704168196_550812448_o.jpg)
with respect to "Thickness", I am not talking about DR or whatever. As mentioned in the post, an MF file holds its own even when scaled up to 200% while the 135 file visibly degrades.
Here are some 100% crops of two files of the same location, shot with similar focal lengths using the same setup (MLU, cable release, very sturdy tripod). Both opened in ACR (It is not even the best converter for the credo files) with no settings changed. Credo on top, D800 at the bottom. Both scaled to 200% in Photoshop.
You can see just how much better the MF file holds up to the enlargement. this is very vital when working on larger prints. Yes, the megapixel counts are similar, but not all pixels are created the same. This is what I mean by "Thickness" :)
The numbers crowd can fight their battles about number of stops and whatever. I work with what I see.
p.s. I know that the Credo file shows visible chroma noise in this screenshot. Before someone brings the knives out about that, no noise reduction was applied to the files (Imported as they are from ACR). In my real workflow, opening the files in C1P and using its noise control tools (Including single pixel noise) cleans the files up beautifully for me.
-
Between the two examples, I prefer the look of the small-format image in this particular case. Yes, the medium format image has a slightly greater sense of depth to it, but I actually do not find that effect to be preferable for this particular image. I will say, however, that the small format image has a warmer overall color balance, which I find more appealing here, and I think it makes the other version look rather cold and unappealing by comparison. So, perhaps the difference in color balance is having some effect on my overall perception of the two images.
Hi David,
No worries, a warmer WB is mostly preferable for portraits. These images are again pretty much SOOC and I stuck with the default renderings. The Credo file can easily be warmed up in a second.
The point I was making though was that the skin in the D800 file has an overall "redness" to it, while the Credo file shows better tonal gradations. This is not something that can easily be corrected in post.
I have also seen this in landscapes where the smaller format muddles the greens up in foliage while the MF camera discriminates them better.
(https://scontent-iad3-1.xx.fbcdn.net/hphotos-xpa1/t31.0-8/1534728_10151954839039209_338595565_o.jpg)
Both cameras in A mode, same setup, similar FoV, MLU, cable release and a sturdy tripod. WB picked manually in post from the same spot.
See how the greens in the D800 image gets polluted from the lighting all around them. The blues in comparison, hold up quite well though.
p.s. The last time I posted this sample, a well know Nikon supporter in the forum had a near-meltdown over it. One of his complaints were that the exposure times were different. Sadly, it seems like he is unaware of the fact that claimed ISO numbers for manufacturers is not necessarily the actual ones and when shooting in aperture priority, exposure times will vary from camera to camera. I invited him to do his own tests, haven't seen one yet. ;)
-
The differences claimed here sits 99% in the camera profile and 1% in the hardware, it has absolutely nothing to do with the size of the sensor.
However as photographers can't make their own profiles it doesn't really matter, but still the difference you see is a software one, not a hardware.
I would also strongly recommend against evaluating camera color in artifical narrow band light (the night landscape case), the results will be pretty random. Differences in color filter responses are exaggerated, and I can assure you no-one designs color filters to render pleasing colors in artificial peaky spectrum lights, as the task is virtually impossible and will damage the possiblities to perform well in good light conditions. It just happen to become what it becomes and in some conditions you're lucky.
One problem is that for most users there are only two raw converters, Capture One and Adobe Lightroom. Adobe doesn't make that great profiles, so any camera tested in Lightroom/ACR will be considered behind. And in Capture One there's a lot of effort put into the MFD profiles, and less effort for the 135 profiles and probably care to make them look a bit different (a really bad business decision if Phase One would strive to make the 135 look the same as their MF cameras).
That is, MFD has in practice a big software advantage, and it's not going to change anytime soon as I see it.
-
The differences claimed here sits 99% in the camera profile and 1% in the hardware, it has absolutely nothing to do with the size of the sensor.
However as photographers can't make their own profiles it doesn't really matter, but still the difference you see is a software one, not a hardware.
I would also strongly recommend against evaluating camera color in artifical narrow band light (the night landscape case), the results will be pretty random. Differences in color filter responses are exaggerated, and I can assure you no-one designs color filters to render pleasing colors in artificial peaky spectrum lights, as the task is virtually impossible and will damage the possiblities to perform well in good light conditions. It just happen to become what it becomes and in some conditions you're lucky.
One problem is that for most users there are only two raw converters, Capture One and Adobe Lightroom. Adobe doesn't make that great profiles, so any camera tested in Lightroom/ACR will be considered behind. And in Capture One there's a lot of effort put into the MFD profiles, and less effort for the 135 profiles and probably care to make them look a bit different (a really bad business decision if Phase One would strive to make the 135 look the same as their MF cameras).
That is, MFD has in practice a big software advantage, and it's not going to change anytime soon as I see it.
Torger,
I have mentioned several times that I have custom profiles for all my cameras. My observations regarding color are with custom profiles, in ALL light conditions. If I did deep into my backups, I can probably find similar examples in daylight, tungsten, sunset etc.
That said, I don't understand how color profiles can affect how a file holds up when scaled to 200% (And that was after an ACR conversion, where the MF file is at a disadvantage). I am asking this because you painted a broad stroke saying ALL the differences shown above can be attributed to profiles, while I am showing many distinct cases.
But I do agree, MF has a big software advantage and if 135 manufacturers don't want that gap to be narrowed, their loss. What I do not agree with is that C1P puts a subpar effort in for 135 files. A default C1P render of a D800 files is miles better than what Adobe can manage. And it only gets better when the IQ250 color profile is used. It's kinda hilarious that a ginormous corporation cannot match up to what a comparatively tiny company in Denmark can do with probably 1/10th the number of programmers.
Back to the topic though, it ties in quite well with what I was saying earlier. It's the whole package that makes MF a different beast to smaller formats. And that includes the tools available for post production too.
-
Out of curiosity, which profiling software do you use to make your own profiles? I also make my own, recently using my own software. You might be interested in that, DCamProf (it's free).
That Adobe's colors are worse is not an objective truth of course, but what folks with MFD tastes generally think. Adobe themselves, which in this case I assume is Thomas Knoll in person, probably think that their color is great. When it comes to color it's probably one or two people that does it, both at Adobe and Phase One.
If you make your own profiles you probably know that cameras can be made to show almost the exact same color for a fixed light condition. Still C1 chooses to make color more different than it needs to be. There's a reason for that, and I think it's about business.
I missed the 200% thing, which as you say is of course not profile related. The D800 has aa filter which makes its blowup quality better due to less aliasing, but many photographers has developed a pixel peep taste that prefers aliasing artifacts over correct micro detail. And indeed remove the aa filter and have sharp lenses you get those "desirable " artifacts. Today 135 has that too though, so it's not format related. MFD has a lead though due to overall sharper lenses, especially in tech cams like I use myself.
I totally agree that one should consider the whole package including post processing, and I think software has more importance than many may think.
-
My tools for color profiling are rather rudimentary by the standards of those who are hardcore into these kinda things.
For LR, I use the Spyder Color Checkr and their associated software to create profiles for different lighting conditions (For each camera, of course), tweak a bit to taste and save them. I use these as starting points.
For C1P, I shoot the Spyder color checkr in different lighting conditions, use a default color profile as starting point, use auto adjust, then do the fine tuning using the color editor and save as ICC profiles.
You"re definitely right that Photographers can benefit from doing custom profiling instead of sticking with the defaults. My cameras, across formats have benefited from this, but I maintain that the difference in color fidelity between MF and smaller formats still remain after this.
Regarding Adobe, it's not just the color rendering that I dislike about their RAW converter. Their default output requires quite some work to match the sharpening of C1P, lacks some features such as corner sharpening for WA lenses, lacks fine tuning tools such as the skin tone editor and so on.
In general, I only work with LR when I have a large batch of images to finish (Such as vacation pictures, which I am doing right now). Adobe certainly has a better library management module, but that is where my praise ends.
I missed the 200% thing, which as you say is of course not profile related. The D800 has aa filter which makes its blowup quality better due to less aliasing, but many photographers has developed a pixel peep taste that prefers aliasing artifacts over correct micro detail. And indeed remove the aa filter and have sharp lenses you get those "desirable " artifacts. Today 135 has that too though, so it's not format related. MFD has a lead though due to overall sharper lenses, especially in tech cams like I use myself.
Yes, the AA filter in the D800 does play a part in that, but I have observed the same with an AA filterless camera such as the D7100 that I owned for a while too (Granted, this is only APSC). I do not have access to a current AA less camera such as the A7RII to do side by tests, but you might be right, the difference might be lesser now.
Good point on the lenses as well. I do not have tech lenses, but the Schneider Kreuznach lenses for Phamiya are quite excellent IMO. 135 format has excellent lenses too, but the dud-to-excellent ratio is a lot more lopsided there.
-
>With a slightly faster lens all shallow DOF effects of MF cameras can be duplicated with smaller sensors also.
>>This has not been my experience.
I agree with Doug.
Interestingly people that like to make these kinds of comparisons look at only one facet - for example sharpness at 100% or in your case a comparison at shallow DOF, but you can actually see differences in format size at all apertures not just shallow. And if you want to look at shallow DOF, here's a shot at 80mm f/2.0 which is focal length that represents normal on the Hy6. Could you do this with a 50mm 2.0 with your DSLR? f/1.4? f/1.2? and what about your MFT camera? My experience is there's 3 stops difference in aperture between MF and 135 but even after you adjust the aperture you still don't get the same look.
-
Hi David,
No worries, a warmer WB is mostly preferable for portraits. These images are again pretty much SOOC and I stuck with the default renderings. The Credo file can easily be warmed up in a second.
The point I was making though was that the skin in the D800 file has an overall "redness" to it, while the Credo file shows better tonal gradations. This is not something that can easily be corrected in post.
I have also seen this in landscapes where the smaller format muddles the greens up in foliage while the MF camera discriminates them better.
(https://scontent-iad3-1.xx.fbcdn.net/hphotos-xpa1/t31.0-8/1534728_10151954839039209_338595565_o.jpg)
Both cameras in A mode, same setup, similar FoV, MLU, cable release and a sturdy tripod. WB picked manually in post from the same spot.
See how the greens in the D800 image gets polluted from the lighting all around them. The blues in comparison, hold up quite well though.
p.s. The last time I posted this sample, a well know Nikon supporter in the forum had a near-meltdown over it. One of his complaints were that the exposure times were different. Sadly, it seems like he is unaware of the fact that claimed ISO numbers for manufacturers is not necessarily the actual ones and when shooting in aperture priority, exposure times will vary from camera to camera. I invited him to do his own tests, haven't seen one yet. ;)
Again, what I seem to be seeing here is an overall difference in color balance, which I think needs to be addressed before close comparisons of color rendition can be made. It looks to me as though the color balance of the small format version is considerably warmer than the medium format version, and a warm overall color balance will obviously subdue cooler colors and may even make them look muddy in some cases. If you are just talking about what comes straight out of the camera, well, that is not set in stone either, since you can make custom color profiles for each camera. Also, it is not just about the cameras. Different manufacturers' lenses can have their own modest differences in overall color rendition.
-
David,
I think you should re read the parts where I have mentioned that manual white balance for both images were done on the same spot and also that I have custom profiles for my cameras.
It is not fair for someone to write a detailed explanation when the other party ignores half of it.
-
David,
I think you should re read the parts where I have mentioned that manual white balance for both images were done on the same spot and also that I have custom profiles for my cameras.
It is not fair for someone to write a detailed explanation when the other party ignores half of it.
Synn. Dunno why you waste time justifying your gear choices. Specially to some of the trolls that reside in this forum. I guess there are some folks here that are honestly asking for information. But the trolls pop out in no time to dissuade any PhaseOne interest or god forbid and actual purchase. It is their duty in life to take down PhaseOne. True haters. (Maybe I got it all wrong and they secretly work for other manufacturers? of 35mm gear? and are sabotaging PhaseOne on purpose, or maybe they just despise capitalism? Maybe they are just angry at everything. Only they know. Some of the haters have actually used PhaseOne gear, others, not even a lens cap. Oh well.
-
David,
I think you should re read the parts where I have mentioned that manual white balance for both images were done on the same spot and also that I have custom profiles for my cameras.
It is not fair for someone to write a detailed explanation when the other party ignores half of it.
Sorry, I missed the comment about manual white balance. However, I am not sure that picking a point within a scene like this for white balance is that precise and would prefer a sampling from a neutral grey card within the scene for maximum precision. In any case, I still think that the overall color balance of the small format shot is again warm, and that it is at least partly for this reason that the cooler tones are not looking as nice as one might prefer.
As for the redness of the skin tone in the portrait shot, I see a modest overall red bias, so I find it hard to evaluate any differences in color rendition between the two systems from this example.
I don't see any reference to custom camera profiles in your comments about either of the two scenes.
-
Synn. Dunno why you waste time justifying your gear choices. Specially to some of the trolls that reside in this forum. I guess there are some folks here that are honestly asking for information. But the trolls pop out in no time to dissuade any PhaseOne interest or god forbid and actual purchase. It is their duty in life to take down PhaseOne. True haters. (Maybe I got it all wrong and they secretly work for other manufacturers? of 35mm gear? and are sabotaging PhaseOne on purpose, or maybe they just despise capitalism? Maybe they are just angry at everything. Only they know. Some of the haters have actually used PhaseOne gear, others, not even a lens cap. Oh well.
I do not hate Phase One or any other medium format gear. I think that medium format has its applications, and there may very well be some differences in color rendition that make medium format preferable for some people and some applications. All I am trying to say is that Synn's examples are not sufficiently convincing to me for the purposes of illustrating differences in color rendition between the two systems in question. I think that it would have been preferable to show a color chart and grey scale within the scenes (or perhaps several, in different parts of the scene), both for overall color reference and to use as a target for white balance.
-
I do not hate Phase One or any other medium format gear. I think that medium format has its applications, and there may very well be some differences in color rendition that make medium format preferable for some people and some applications. All I am trying to say is that Synn's examples are not sufficiently convincing to me for the purposes of illustrating differences in color rendition between the two systems in question. I think that it would have been preferable to show a color chart and grey scale within the scenes (or perhaps several, in different parts of the scene), both for overall color reference and to use as a target for white balance.
If you haven't seen the difference after looking at samples, then that difference doesn't exist for you.
Edmund
-
Synn. Dunno why you waste time justifying your gear choices. Specially to some of the trolls that reside in this forum. I guess there are some folks here that are honestly asking for information. But the trolls pop out in no time to dissuade any PhaseOne interest or god forbid and actual purchase. It is their duty in life to take down PhaseOne. True haters. (Maybe I got it all wrong and they secretly work for other manufacturers? of 35mm gear? and are sabotaging PhaseOne on purpose, or maybe they just despise capitalism? Maybe they are just angry at everything. Only they know. Some of the haters have actually used PhaseOne gear, others, not even a lens cap. Oh well.
Good point, ken. I made the effort in case somebody googles this thread up looking for info. Not gonna make any more effort. There are enough career specialists in the forum to provide the mind numbing charts and graphs proof that some cannot do without. Some of us unfortunately have to be content with shooting only real world images.
Cheers.
-
(https://scontent-iad3-1.xx.fbcdn.net/hphotos-xpa1/t31.0-8/1534728_10151954839039209_338595565_o.jpg)
p.s. The last time I posted this sample, a well know Nikon supporter in the forum had a near-meltdown over it. One of his complaints were that the exposure times were different. Sadly, it seems like he is unaware of the fact that claimed ISO numbers for manufacturers is not necessarily the actual ones and when shooting in aperture priority, exposure times will vary from camera to camera. I invited him to do his own tests, haven't seen one yet. ;)
No, I provided you with feedback that the D800 image appears to be over-exposed. I read your PM in which you blamed the camera meter... which was an acknowledgement of the exposure problem (which is the moment you put me in your black list...). Yet you keep posting the same images as a demonstration of a color issue.
I am more than willing to admit the superiority of whatever camera over the D800 (and it should be superior considering the important price premium), but confusing amount of exposure with color quality is a bit weak in terms of factual data.
Yet I am the "Nikon guy" and you are the "I own both so I don't care guy". ;)
Cheers,
Bernard
-
I can understand some things... but I can't understand what the word "redifined" in the OP's title means... or where it refers to... may he explain please?
-
What a messy thread lol! Tbh I think the op just needs to use a dmf camera to get what the real big deal is. I don't think you can show anyone what you can see over the Internet. also I do like 35mm over the mf file of the first example of the Asian lady. Not because of colors but the composition was better lol. composition over camera.
-
For LR, I use the Spyder Color Checkr and their associated software to create profiles for different lighting conditions (For each camera, of course), tweak a bit to taste and save them. I use these as starting points.
OEM software from DataColor does not create profiles - it creates presets... so you are still using Adobe profiles... so I think we shall stop right here about the quality of demonstrated conversions for the purposes of illustration, no ?
-
What I do not agree with is that C1P puts a subpar effort in for 135 files.
but it does _versus_ the profiles for their MFDBs ...
-
Could you do this with a 50mm 2.0 with your DSLR? f/1.4? f/1.2? and what about your MFT camera? My experience is there's ___3 stops difference___ in aperture between MF and 135 but even after you adjust the aperture you still don't get the same look.
snake oil is alive and well... not 6 stops of DR advantage, but 3 stops difference in aperture for a MF which is at best 2.5 times in sensor size vs FF camera... for all other systems (P&S vs m43 vs APS-C vs FF) 2 times difference = 1 stop... but MFs has some magical dragon urine sprayed over it so that does not work anymore ;D
-
Synn. Dunno why you waste time justifying your gear choices. Specially to some of the trolls that reside in this forum. I guess there are some folks here that are honestly asking for information. But the trolls pop out in no time to dissuade any PhaseOne interest or god forbid and actual purchase. It is their duty in life to take down PhaseOne. True haters. (Maybe I got it all wrong and they secretly work for other manufacturers? of 35mm gear? and are sabotaging PhaseOne on purpose, or maybe they just despise capitalism? Maybe they are just angry at everything. Only they know. Some of the haters have actually used PhaseOne gear, others, not even a lens cap. Oh well.
who is troll here though ? the one who exposes raws differently ? the was one Ph.D here who once illustrated something using raw conversion from MF and JPG from a cell phone camera...
-
but it does _versus_ the profiles for their MFDBs ...
That is not necessarily because C1 *wants to* make SLRs look bad, it may be because they actually have more information about their own sensors via their calibration process. In particular there is substantial variation among sensors; a boutique manufacturer like Phase can afford to characterize each individual unit.
Using Canon's DPP vs. Photoshop ACR used to be a day vs night difference, and Adobe do not go out of their way to give users bad images, they just don't care enough about any *one* model to refine the quality. Their business is decent conversion for all cameras, and good workflow.
So let's put aside accusations against C1, they are competing with Adobe and it is not in their interests to be bad even if they can allow themselves to be mediocre.
Irident's Raw developer, Canon's DPP etc. are the software one should use to compare a prosumer dSLR or a Sigma compact against a Phase back. And unless extreme enlargements are necessary, very good files can be obtained from a Sony sensor dSLR with a Zeiss lens, or a Sigma Quattro. There is little reason for matching a Canon Rebel with a kit lens against a $30K back.
Most of the Phase advantage in quality against other brands is negated if the user uses ACR, as it is tied to the integration of the P1 back with C1. Therefore the ACR user might as well go for a different brand -remember the Pentax has *exactly the same* sensor as the IQ250, in a body that is at least as good as modern as the XF, but at 1/3 the price. And there is a very nice Hasselblad Cmos back out there which will make the users of tech cameras perfectly happy, at again less than half the price of an IQ back.
Phase is a camera brand like any other. We should compliment them on the bulletproof quality of their software/hardware integration, just like Hasselblad has exceptionally good focus for an MF camera, and Leica has peerless lenses. We are allowed to joke about their pricing, and we can tease their unconditionally supportive and passionate... dealers :)
Edmund
-
That is not necessarily because C1 *wants to* make SLRs look bad, it may be because they actually have more information about their own sensors via their calibration process. In particular there is substantial variation among sensors; a boutique manufacturer like Phase can afford to characterize each unit.
certainly...
-
certainly...
Nikon and Canon -and certainly Sony- know a lot which they don't tell third party converters ...
In particular, they will know what the "blurring functions" are and how to deconvolve them, be it for their own lenses, be it for the sensors. They know what the geometrical deformations are at each distance. They may have a characterization of the individual lens. They know about focus settings. They know how well the sensor batch matches the ideal sensor etc.
You can see all of this at work when you match a Raw from a Sony compact to the incredible Jpeg which comes out of the camera.
Edmund
-
I can understand some things... but I can't understand what the word "redifined" [sic] in the OP's title means... or where it refers to... may he explain please?
OP here, hello…
What I meant was that during the good old film era it was extremely easy to define different formats, it simply was the size of the film which determined it. The image quality was determined by the film size, basically. Along with the film size came different focal length lens sets for each format, which in turn also defined the "look" of each format (mostly lack of DOF with bigger formats). All this was simple because film emulsions were the same for all formats, so IQ was directly tied to the film size.
Now this simple equation size = quality does not hold anymore, as a modern high quality small sensor can and does surpass many so called MF sized sensor in both resolution and especially DR. So what I was suggesting was to redefine MF to mean certain level of IQ regardless of the sensor size. Traditional MF "look" features like shallow DOF can be produced by faster lenses on smaller sensors, so the "look" remains the same.
I am approaching this more from the theoretical angle, scaling a size X sensor up or down with the correct lenses/apertures should give us optically exactly the same results (except diffraction effects). So it should be the end result which determines the "format", not just the sensor size.
Carry on...
-
I just love being able to buy medium format gear so cheap now because all the sheep are running down the "my tiny sensor is (theoretically?) better than yours" road.
The two tiny JPG images being compared aren't even the same shot.
The trouble is that when you have to print that smaller sensor image it has to be magnified a lot more than the big sensor, so it always loses. Anyway, who cares.
-
The old 35mm SLRs had those F1.4 lenses which don't really have an equivalent on cellphones and compact.
As a result, our cellphones have pretty much the resolution of 35mm but not the look, and we need to use real 24x36mm sensors to replace 35mm film.
Edmund
OP here, hello…
What I meant was that during the good old film era it was extremely easy to define different formats, it simply was the size of the film which determined it. The image quality was determined by the film size, basically. Along with the film size came different focal length lens sets for each format, which in turn also defined the "look" of each format (mostly lack of DOF with bigger formats). All this was simple because film emulsions were the same for all formats, so IQ was directly tied to the film size.
Now this simple equation size = quality does not hold anymore, as a modern high quality small sensor can and does surpass many so called MF sized sensor in both resolution and especially DR. So what I was suggesting was to redefine MF to mean certain level of IQ regardless of the sensor size. Traditional MF "look" features like shallow DOF can be produced by faster lenses on smaller sensors, so the "look" remains the same.
I am approaching this more from the theoretical angle, scaling a size X sensor up or down with the correct lenses/apertures should give us optically exactly the same results (except diffraction effects). So it should be the end result which determines the "format", not just the sensor size.
Carry on...
-
You can see all of this at work when you match a Raw from a Sony compact to the incredible Jpeg which comes out of the camera.
Edmund
Interesting, are you saying the Jpeg algorithms vary between individual cameras of the same model, being sensor dependant?
Is here any way of demonstrating that?
I thought that matching a RAW to an in camera jpeg was easier with the makers own software which, as you say, would benefit from their inside knowledge, or does that pick up markers in the RAW file to "guide" the conversion again based on individual/sensor batch calibration?
-
I'm myself still a beginner when it comes to profile design, but I have some strong indications so far that it ain't easy :)
The first part where you design from test targets or color filter responses is fairly mechanical and automatic, but then to make the subjective adjustments to make it work well with a tone curve and various subjects is the hard part. If you do it manually this means that you need to look at a lot of test pictures and hand-tune manually, test the profile for some time, do some further refinements. It can be weeks before you're satisfied.
Of what I've heard from Capture One profile design process they indeed do the adjustments manually, and they work a long time with it -- for their medium format cameras. Spending many hours or even weeks on a single set of profiles for each camera is of course not reasonable so I assume they use more automatic methods for the consumer cameras, and do their costly manual work only with their medium format cameras.
What I don't know yet is once you've designed a successful look how easily that can be re-used on a new camera. Hopefully much can indeed be re-used and then it's not so much work supporting a new camera. If you have to go through the whole process again though, then you could only focus that effort on the cameras you really care about.
If you use Canon or Nikon's own software for their own cameras you will most likely get access to more worked-through profiles than you get in Adobe Lightroom or Capture One. However there's a lot of taste involved too of course so the profile that has got the most attention may still not match your taste best.
There's a lot of myths and bias in there too. I've seen people praise the "color accuracy" of the medium format cameras, out of the box. But you certainly don't need to have "golden eyes" to see that the profiles have no intention to produce "accurate color". Simple A/B comparison between real objects and captured image in a calibrated system and you see hue shifts, lightness shifts, saturation shifts. Intentionally designed of course, they provide a canned look. The simplest demonstration of this is comparing Leaf and Phase One cameras that share the same sensors, yet with the default profiles produce quite different looks.
I haven't studied Phase One cameras that much, except for the P45+, but I did own a Leaf back. While providing a pleasing look I think they added too much of a look, but they had their oddly name "ProPhoto" mode that was quite neutral. Now I have a Hasselblad and I think it's about the best I've seen so far in the balance between realism and a look, I'd say it's about "95%" realistic, and very subtle looks added, except for high saturation colors which are strongly gamut mapped (desaturated to fit in a smaller gamut, which was how you did things before computers where able to do realtime gamut mapping).
Leaf's strong looks seems to work very effectively as a lock-in though. I've recently been in contact with a user that for external business reasons changed from Leaf Credo to Phase One IQ, and now they have very real issues to achieve the color they want as they've been used to start off with the Leaf canned looks. I think the strongest reasons to develop a workflow with your own custom profiles is simply to avoid this type of lock-in.
I haven't really figured out Adobe. They allow embedding a curve in their profiles, but if you do the colors will be garish, unless you pre-compensate with a LUT. In my own profiles this pre-compensation becomes pretty strong and requires a large LUT, but Adobe's own profiles while pre-compensating some it does it to much lesser extent. That is quite much of this curve look is left in the profile, and I think that is one of the more important reasons that Adobe's color is considered not that good by many. Capture One has a split approach, some of the curve is implemented in the LUT, and some is added on the side. The reason is to counter-act the negative color appearance side effects of the curve. I think it's quite obvious that Capture One have had a better understanding of the color appearance side effects of a curve than Adobe has, and has implemented a better solution in their profiles, at least in the profiles I've studied. There's no problem to fix in Adobe's DNG profiles if you want to, but Adobe doesn't really do it.
-
I don't think you need any "inside knowledge" of the sensor to make a great profile. What type of knowledge would that be? To accurately measure the sensor's color filter response (spectral sensitivity functions, SSFs) you need gear that cost say $10k - $20k, not consumer stuff but no problem for any of the bigger raw converter makers. Not all use SSFs though, I doubt Phase One does it based on the profile design stories I've heard, I know Hasselblad doesn't (I've asked). When a great deal of the profile design is about manual tuning it's not that important to have the full data on the sensor. You can start of with a simple target-based profile, and the rest is manual tuning.
I instead think the problem is that there are no good automatic methods to make high end general purpose profiles, and making a good manual one requires many hours from a profile designer that has a good eye for color and knows what to achieve (and has the custom software tools to do it), and the most important that this designer has a taste that the users like. I would guess that at each manufacturer there's one or two key people that has the major design impact on how the color becomes.
As there's no "objectively best" concerning color, it's a difficult business. Probably the Adobe folks thinks their color is the best, just as Phase One's. The users decide in the end, but there are varying opinions there too. I would not be surprised if Adobe's color is more popular in the broader user base.
-
This idea of best, and that is rarely the most accurate (however defined), is the problem. My best is not yours.
Simple analogy from my primary work: Hospitals have end of bed drug charts, you would think in the UK we would have a single best NHS chart, no, each hospital has it's own, which is best, for them. A DG, district general, hospital has different requirements than a specialist transplant centre. They all do the same job, there is no global best.
-
Some aspects may also seem "objective" while they're not. For example if a camera is better to separate various shades of greens in green foliage. While better color separation may seem an objective improvement, this aspect can also be designed in the profile. In the foliage example simply by stretching the gamut in the green range. The opposite, compressing the gamut, is a quite common adjustment to do in the skintone range to make skin look smoother. Of course by stretching in some range you need to compress in some other, so there's always a trade-off.
Camera hardware do differ in how good they are at separating color, but of what I've seen so far it's one of the least important factors, as any reasonably modern camera is already very good at the job.
When working with profiles I've noted that how highlight rolloff into the whitepoint is handled has quite a big impact on the look. It seemed like a minor detail at first, but the rolloff into white starts already in the midtones and there can be quite some difference in how you choose to desaturate into white. For regular reproduction work this has no meaning as you don't have curves and you don't clip stuff there, but for general-purpose profiles this is a key aspect. It also illustrates a thing about subjective profile design that's not really about altering hues into something "better" than the accurate hue, but how to model a scene's appearance with the limited dynamic range of a screen or print.
This aspect also has a great impact on how a thing like a bright saturated flower is rendered. With a sharp rolloff you may keep the true hue almost all the way to clipping, but with a longer rolloff the different shades of desaturated tones may make up a more refined look and show more detail in terms of luminance.
It's actually quite basic bits of color rendering, but since we as users have since the introduction of digital photography let the manufacturers deliver canned looks to us few have thought about how it actually works and how you could make it better. Instead users have assumed that the hardware, mostly the camera sensor, is where most color magic happens. I think this has been good for the medium format segment, as it's easier to motivate high hardware prices when the performance sits in the hardware rather than in the software (well, 10 years ago the hardware really was vastly superior), and it has always been simpler to sell expensive hardware than expensive software. If the pricing model were adapted to which part that provides the most value, digital backs would be a lot cheaper and Capture One would be a very expensive piece of software.
-
If the pricing model were adapted to which part that provides the most value, digital backs would be a lot cheaper and Capture One would be a very expensive piece of software.
That software is easier to "pirate" and hardware much harder to clone the current pricing model may be correct.
-
That software is easier to "pirate" and hardware much harder to clone the current pricing model may be correct.
Yes, I should have been clearer. I think they have a wise pricing model, I wouldn't do otherwise. Make the customers pay via hardware is an excellent model. I sell some consumer software and I know how hard it is to get paid, you need to have Adobe-style volumes before you can make any real money on it.
What I wanted to illustrate is that due to the necessity to focus on hardware, users have come to think that hardware plays a larger role and software a lesser when it comes to camera color, compared to how it actually is.
If you could sell a camera profile for say $500-$1000 I think we would see some tougher competition in the color rendering aspect. Say if a $1000 profile could lift your 135 camera to medium format quality in terms of color rendition wouldn't that be totally worth it? Probably, but it's more fun to spend money on a really sharp lens or some other hardware instead.
-
OP here, hello…
What I meant was that during the good old film era it was extremely easy to define different formats, it simply was the size of the film which determined it. The image quality was determined by the film size, basically. Along with the film size came different focal length lens sets for each format, which in turn also defined the "look" of each format (mostly lack of DOF with bigger formats). All this was simple because film emulsions were the same for all formats, so IQ was directly tied to the film size.
Now this simple equation size = quality does not hold anymore, as a modern high quality small sensor can and does surpass many so called MF sized sensor in both resolution and especially DR. So what I was suggesting was to redefine MF to mean certain level of IQ regardless of the sensor size. Traditional MF "look" features like shallow DOF can be produced by faster lenses on smaller sensors, so the "look" remains the same.
I am approaching this more from the theoretical angle, scaling a size X sensor up or down with the correct lenses/apertures should give us optically exactly the same results (except diffraction effects). So it should be the end result which determines the "format", not just the sensor size.
Carry on...
HI,
I think that MF bodies do follow the traditional path that was introduced with film... There are "modular" cameras that take MFDBs (instead of film backs) and there are DSLR type cameras like Pentax 67 or Exakta 66 was... There are some "modern" features added in their specs (like AF for instance) but this is to be expected as things advance with time... It's not really different than it was with the later film cameras (Rollei, Contax) than the first generation of MF cameras during the film era...
What has really changed is the light sensitive areas, where with film one could have the same emulsion/technology used with a 35mm camera, and the same again with an MF camera the difference only being in size... But I have a feeling that there is a tendency for this to also change and return back to the roots, I believe that we won't see anymore sensors coming of the CCD type, IMO all future sensors will be CMos and although they won't be upsized versions of FF DSLR sensors, they will be of very similar technology and performance characteristics... In fact I believe that the later introductions (Sony's MF sensor, Leica S007) prove much of the case...
-
Yes, I should have been clearer. I think they have a wise pricing model, I wouldn't do otherwise. Make the customers pay via hardware is an excellent model. I sell some consumer software and I know how hard it is to get paid, you need to have Adobe-style volumes before you can make any real money on it.
Even Microsoft are giving it away free now ;D
-
snake oil is alive and well... not 6 stops of DR advantage, but 3 stops difference in aperture for a MF which is at best 2.5 times in sensor size vs FF camera... for all other systems (P&S vs m43 vs APS-C vs FF) 2 times difference = 1 stop... but MFs has some magical dragon urine sprayed over it so that does not work anymore ;D
I'm was posting about DOF (depth of field). Not sure where or how you got to DR. But its one of those things - either you see it or you don't.
-
I'm was posting about DOF (depth of field).
me too... "6 stops of DR advantage" is a local meme from the days of "16 bit raw files", so to say
-
I got to the 3 stops difference in aperture by direct comparison in studio with AFi-ii 12 compared to Nikon d800e. It's really quite amazing the difference. If you shoot a model at f/16 with the AFi-ii 12 you can't even get their whole body from front to back within the envelop of DOF, but at f/8 you can get that and more with the same framing. There probably is a mathematical relationship that expresses this but I'll just go by experience. Stopping down FF DSLR 3 more aperture stops seems about right to get to the same DOF.
-
The math of DoF:
Say if we prefer 4:3 format, then we get 32x24mm from 135 and 54x41 from 645, which is 40mm diagonal vs 67mm, factor 1.7. To match the look of an 80/2.8 medium format lens would require a 80/1.7 = 47mm and 2.8/1.7 = f/1.6, closest typical real lens would be a 50/1.4.
The 135 format has lenses with very short depth of field, so I don't think anyone uses MFD to get shorter, as you probably can get even shorter with 135. To match Canon's 85mm f/1.2 you'd need 150mm f/2 and few (any?) medium format systems have that.
Ultra-short DoF is rather gimmicky though and doesn't make good portraits, so it's not a loss. I've heard that many prefer the bokeh of MF lenses, and that there are some finer smoother quality in the transition from in focus to out of focus. I don't think that has to do with the format, but possibly due to that you have simpler lens designs. The high res ultra-wide aperture 135 lenses are certainly not simple designs, and maybe some aspects of the bokeh suffers from that.
A big drawback with most MFD lenses as I see it is that they use apertures that are not round, leading to ugly bokeh highlights. It's a problem on my tech camera too, due to the Copal shutter, fortunately I don't use bokeh often in my shooting style but if I would I would not be happy about that. All high end 135 lenses have round apertures.
With the newest high end 135 lenses designed for high res sensors like Otus and Canon's new 35mm/1.4 I've noted a real improvement in bokeh, bokeh fringing is almost completely gone. While 135 format is still overall lagging behind in resolving power due to the lenses, these new lenses is changing the landscape.
I think that the choice of MFD of the SLR type will be less about resolution or depth of field in the future, but more about if you like the camera ergonomics, and the raw converter workflow and the rendering (color etc) you get in that workflow. It remains to be seen how well the current pricing model will work for the new landscape. I think we will see the system come closer also in price, the high res 135 is already notably more expensive than earlier (due to increased lens complexity), and we have seen lower MFD prices here and there, like the new Leica S.
-
Another way to see it, given a specific output size (e.g. an 8 x 10 print) and a specific angle of view, then the DOF is determined by the diameter of the entrance pupil of the lens used, regardless of the format of the capture medium.
In the example given by Torger, that value would be = 80/2.8 = 28.57 mm, which will be the same entrance pupil of a 47 mm @ 1.6. Just in case, those numbers have small rounding errors, the exact figures will be a 47.407 mm lens at an aperture of 1.6592
This gives slightly less than 2 f-stop difference between medium format and 35 mm in terms of DOF.
I agree with torger that other factors such as sharpness, optical distortions, bokeh characteristics and so on influence the perception of DOF
-
Although the numbers say one thing... One must consider the aspect ratio too, with 4:3 aspect ratio, one works nearer to the subject than if the ratio was 3:2...
-
The calculations (both Torger's and mine) were based on a 4:3 aspect ratio
32 x 24 mm for 35 mm
54 x 40.5 mm for medium format
Aspect ratio, angle of view and output size have to be the same, otherwise you are comparing apples to oranges.
-
OK... but what about the position of the entrance pupil... the distance of it from the Light sensitive area also affects DOF... And since it is not a constant distance it is a parameter that hasn't been taken into account with the calculations.
-
For Theo and Francisco, this is for you (http://ocw.mit.edu/courses/mechanical-engineering/2-71-optics-spring-2009/video-lectures/lecture-6-terms-apertures-stops-pupils-and-windows-single-lens-camera/MIT2_71S09_lec06.pdf)
For the rest of us, Torger's perfectly sufficient (and accurate) approximation will do just fine.
Pythagoras anyone ?
The math of DoF:
Say if we prefer 4:3 format, then we get 32x24mm from 135 and 54x41 from 645, which is 40mm diagonal vs 67mm, factor 1.7. To match the look of an 80/2.8 medium format lens would require a 80/1.7 = 47mm and 2.8/1.7 = f/1.6, closest typical real lens would be a 50/1.4.
-
Thanks Manoli for the presentation.
Just to clarify, there is indeed an influence of focal lenght and pupil factor in the DOF. This affect close range focus, especially in the macro region.
For distances above 1 m. the differences are negligible.
OK... but what about the position of the entrance pupil... the distance of it from the Light sensitive area also affects DOF... And since it is not a constant distance it is a parameter that hasn't been taken into account with the calculations.
That is why you need to consider a specific output size and ange of view. The size of the Circle of Confusion CoC will be smaller in the 24*32 sensor than in the 40.5*54, but the required enlargement from the 24*32 sensor to get the output size will be greater. So yes, it has been taken into account in the calculation (provided we are considering object distances above 1 m.)
-
Hi Anders,
Thanks for talking about the math… I tried to post on it but I felt it was a to extensive subject to discuss at depth.
My guess is that with really good lenses, like Otus and Sigma Art, smaller formats can be quite competitive with larger formats. I just bought a Sony A7rII and I am pretty sure it can perform on par with my Hasselblad 555/ELD and P45+ combo.
The reason I keep the P45+ is that it is fun to shoot with, I would miss it for sure. But, very clearly, when it matters I will go with my Sony A7rII or the still pretty good Alpha 99.
Best regards
Erik
The math of DoF:
Say if we prefer 4:3 format, then we get 32x24mm from 135 and 54x41 from 645, which is 40mm diagonal vs 67mm, factor 1.7. To match the look of an 80/2.8 medium format lens would require a 80/1.7 = 47mm and 2.8/1.7 = f/1.6, closest typical real lens would be a 50/1.4.
The 135 format has lenses with very short depth of field, so I don't think anyone uses MFD to get shorter, as you probably can get even shorter with 135. To match Canon's 85mm f/1.2 you'd need 150mm f/2 and few (any?) medium format systems have that.
Ultra-short DoF is rather gimmicky though and doesn't make good portraits, so it's not a loss. I've heard that many prefer the bokeh of MF lenses, and that there are some finer smoother quality in the transition from in focus to out of focus. I don't think that has to do with the format, but possibly due to that you have simpler lens designs. The high res ultra-wide aperture 135 lenses are certainly not simple designs, and maybe some aspects of the bokeh suffers from that.
A big drawback with most MFD lenses as I see it is that they use apertures that are not round, leading to ugly bokeh highlights. It's a problem on my tech camera too, due to the Copal shutter, fortunately I don't use bokeh often in my shooting style but if I would I would not be happy about that. All high end 135 lenses have round apertures.
With the newest high end 135 lenses designed for high res sensors like Otus and Canon's new 35mm/1.4 I've noted a real improvement in bokeh, bokeh fringing is almost completely gone. While 135 format is still overall lagging behind in resolving power due to the lenses, these new lenses is changing the landscape.
I think that the choice of MFD of the SLR type will be less about resolution or depth of field in the future, but more about if you like the camera ergonomics, and the raw converter workflow and the rendering (color etc) you get in that workflow. It remains to be seen how well the current pricing model will work for the new landscape. I think we will see the system come closer also in price, the high res 135 is already notably more expensive than earlier (due to increased lens complexity), and we have seen lower MFD prices here and there, like the new Leica S.
-
Aspect ratio, angle of view and output size have to be the same, otherwise you are comparing apples to oranges.
Francisco, you raise a pertinent point regarding output size.
I went to do some calculations with Bart VdW's 'Depth of Field output quality planner' and was surprised to find it off-line.
[ http://bvdwolf.home.xs4all.nl/main/foto/dofplan/dofplan.html ].
Does anyone have the new URL ?
-
It is long before I've been into a technical conversation and I surely don't want to argue or challenge any others knowledge which I'm sure it more or less presents the real story... Torger's and Fransisco's calculations are correct and they do say most of the story or are enough for one to estimate (about) the DOF difference between MF and FF sensor... However, it is a general truth which may vary a little and usually ends up with MF having even more shallower DOF than the two stops suggested... Not that it makes a real difference that matters so much that one should absolutely consider, but DOF does vary with the position of the entrance pupil and the entrance pupil doesn't have a constant (virtual) position in a lens, but its position varies with parameters... (e.g focusing distance is one parameter) practically, one would find that usually MF lenses have more than two stops shallower DOF and that varies between MF lenses too... (eg. a lens made for a 6x6 camera wiil be usually found to have (slightly) shallower DOF than a lens dedicated for a 6X4.5 camera etc...). There are also differences that may be found between lenses that have a leaf shutter when compared with same focal lengths but with no leaf shutter... It is common with DSLR cameras (either 35mm or MF) the more the mounting distance of a lens, the less the DOF (without this being a rule) that's why Nikkor lenses have usually (slightly) less DOF than an Canon lenses meant for Eos... That said, I don't know anyone that choose Nikon over Canon (or vise versa) because DOF would have a very slight difference...
-
usually ends up with MF having even more shallower DOF than the two stops suggested...
first of all even here, few posts up it was less than 2 stops and then - can be way less because not all MF sensors are that big and not always you reduce the size of 3:2 FF sensor by cropping it to favor the aspect of MF... how about cropping a smaller MF sensor to 3:2 orientation of FF
-
first of all even here, few posts up it was less than 2 stops and then - can be way less because not all MF sensors are that big and not always you reduce the size of 3:2 FF sensor by cropping it to favor the aspect of MF... how about cropping a smaller MF sensor to 3:2 orientation of FF
I don't like what I say being altered and what you say is irrelevant to what I say (as usual with you)... To be more understandable (even from you), the AOV will be determined by the size of the image area and this leads from first view to compare it with another lens that also covers the same image area, but the lens (any lens) has a certain image circle and its DOF depends on the distance from the image area to exhibit this DOF... further more the entrance pupil is a virtual "hole" (but very real) that doesn't have constant position in a lens... In other words, for whatever calculation to be completely (and accurately) applicable, it also requires the same mounting distance for the two lenses and the same position for the entrance pupil... I hope this advances your knowledge a bit...
-
It is common with DSLR cameras (either 35mm or MF) the more the mounting distance of a lens, the less the DOF (without this being a rule) that's why Nikkor lenses have usually (slightly) less DOF than an Canon lenses meant for Eos...
Theodoros
Is this the reason why my Schneider Digitars (35XL + 24 XL) have the longer DOF than the Rodenstocks (32 + 23)?
I never compared them directly, but I always had the impression that the Rodies have much less DOF than the Schneiders,
maybe because of the retrofocus design?
Even the 24 PC-Nikkor (with a D800) seems for me to have less DOF at the same F-Stop as the 35 XL Digitar with a 44 x 33 sensor.
Best regards,
Ben
-
Hi,
Is that observation based on prints or viewing on screen. If on screen at what scale?
At actual pixels the DoF will seem to be thinner with an 80 MP-sensor (as a smaller crop would be shown). But, that would normalise in a print.
Best regards
Erik
I got to the 3 stops difference in aperture by direct comparison in studio with AFi-ii 12 compared to Nikon d800e. It's really quite amazing the difference. If you shoot a model at f/16 with the AFi-ii 12 you can't even get their whole body from front to back within the envelop of DOF, but at f/8 you can get that and more with the same framing. There probably is a mathematical relationship that expresses this but I'll just go by experience. Stopping down FF DSLR 3 more aperture stops seems about right to get to the same DOF.
-
Theodoros
Is this the reason why my Schneider Digitars (35XL + 24 XL) have the longer DOF than the Rodenstocks (32 + 23)?
I never compared them directly, but I always had the impression that the Rodies have much less DOF than the Schneiders,
maybe because of the retrofocus design?
Even the 24 PC-Nikkor (with a D800) seems for me to have less DOF at the same F-Stop as the 35 XL Digitar with a 44 x 33 sensor.
Best regards,
Ben
Ben, I'm not familiar with modern view camera lenses as I've never owned one with a digital back... I do have access to a P3 with 75H back that a friend owns, but the wider he uses is a 47mm lens... The lenses I used to have in the past for the P2, where older lenses aimed for 4x5 sheet film and I've sold them with the camera as I now use a Fuji GX-680 with my two digital backs alongside my Contax 645 system. However, I would expect the lens that has a wider image circle, to also show a narrower DOF since the value of the circle of confusion (COC) that the lens exhibits should increase.
-
Agree Synn
I appreciate your real world comparisons. Thank you.
I see this all the time with corporate portraiture skin tones. The smaller format has this flat red smear and cutting in cyan makes a different problem. In the past with only dslr's available to me, Celtic skin required matting down with green tinted foundation in the makeup room.
MF file shows lots of different colours within the skin colour. The clarity and depth of colour shows through even at a 500 pixel webpage size.
I'll put some samples up when I get back to the studio. They are chalk and cheese.
On another note, As far as DR goes, I was surprised after switching to MF, even with my studio portraiture, I have found I am barely within the DR capabilities of the camera - H5D-50. I have to expose as accurately as I did Transparency film. The MF CMOS solution would fix that, I just don't like the smaller size of the sensor and the puny image size through the viewfinder that the Blad and Pentax CMOS offer. I like to manual focus sometimes.
-
I would expect the lens that has a wider image circle, to also show a narrower DOF since the value of the circle of confusion (COC) that the lens exhibits should increase.
Not so, why would it? Only the focal length and relative aperture matters here.
You seem to have some quite funny ideas about optics.
-
Not so, why would it? Only the focal length and relative aperture matters here.
You seem to have some quite funny ideas about optics.
I don't like people to call "funny ideas" what is common knowledge among optics engineering... in the mean times, this is a very general approach about the relationship between COC, image format (that the lens is able to cover) and DOF... I hope you'll find it helpful as to correct the funny ideas about optics that you have... https://en.wikipedia.org/wiki/Circle_of_confusion
-
I don't like people to call "funny ideas" what is common knowledge among optics engineering... in the mean times, this is a very general approach about the relationship between COC, image format (that the lens is able to cover) and DOF... I hope you'll find it helpful as to correct the funny ideas about optics that you have... https://en.wikipedia.org/wiki/Circle_of_confusion
I did not find any references to "image circle" in that article and it does not feature in any of the formulas for circle of confusion or DOF, so I must assume it does not factor in at all. And please enlighten me how it even could.
Until then I hold on to my viewpoint about "funny ideas" about optics...
-
I did not find any references to "image circle" in that article and it does not feature in any of the formulas for circle of confusion or DOF, so I must assume it does not factor in at all. And please enlighten me how it even could.
Until then I hold on to my viewpoint about "funny ideas" about optics...
You did find image format size didn't you? ??? There is even a table shown that gives you the COC value increase depending on the image area that the lens is designed for ;)... The image circle of the lens under question is supposed to exactly cover that same image area... In other words, if you use another (smaller) size image area, it doesn't change the lens characteristics ... it's quite easy, try rethink of it... :'(
-
You did find image format size didn't you? ??? There is even a table shown that gives you the COC value increase depending on the image area that the lens is designed for ;)... The image circle of the lens under question is supposed to exactly cover that same image area... In other words, if you use another (smaller) size image area, it doesn't change the lens characteristics ... it's quite easy, try rethink of it... :'(
No and no. Film format affects the circle of confusion, because it dictates how much the picture is to be enlarged for the print os a given size. What kind of a lens you use is fully up to you: if you use a lens with overly large image circle you just wasting your money on a lens designed for a bigger format, gaining nothing, or if you use a lens which has a too small image circle you vignette the corners badly. How could this affect the DOF?
That table you mention shows the effect of the image size, not the lens image circle. Does a 135 system 24 mm T&S lens have a different COC from a regular 24mm prime, because the image circle is bigger? Hah, no it does not. Your misunderstanding probably stems from the fact that you mistakenly think this table has something to do with the image circle, when it really just is the image size and the image circle has nothing to do with it (except that in practice you need to use a lens with big enough IC to cover the format).
Like I said, if the image circle had anything to do with COC or DOF, if would be part of the formulas, or it could be derived from the formulas. Please show this instead of typing funny faces on your posts.
-
The T/S lens can't be used as an example because it shares the same mounting distance... if the larger format lenses would mount at the same distance as smaller format lenses the COC value wouldn't change... what increases the COC value and thus reduces DOF is precisely the mounting distance as the COC widens with the mounting distance... as I said before, think... ??? ...or read one of the references that are on the appendix of the link that I provided... ;)
It's all common logic really... the more the mounting distance the larger the COC... Another way to think of it is as a percentage of the image area that the lens covers... Assumming that glass is the same and because COC should be the same in analogy, the larger the image area that can be covered by the lens, you get in proportion the increased COC value... and thus DOF decreases... As simple as that... ;D 8) :P
-
Petrus (and others) keep forgetting that if one uses a larger format lens for a shorter format image area, what one sees is a crop of the original image area of that larger format... But, OTOH, he admits that the larger image area does have increased COC and thus reduced DOF... Then he claims that if cropping a part of that area, DOF will be increased as if a smaller format lens (of the same focal length) would be used (which shows smaller COC)... I know no magic and testing proves that it doesn't... The increased COC value, still affects the whole image area that the lens is designed to cover and thus, the cropped image area retains, (of course!!!) the DOF that was recorded on the projected image circle... (which of course is reduced because of the larger COC value)!
Obviously he confuses the fact that the result print, if being the same in size, is still a cropped proportion of the one that was projected by the lens (minus what was not recorded on the image area) and thus what one sees on a print of the same size is a result of a bigger enlargment... Simple, isn't it?
-
The T/S lens can't be used as an example because it shares the same mounting distance... if the larger format lenses would mount at the same distance as smaller format lenses the COC value wouldn't change... what increases the COC value and thus reduces DOF is precisely the mounting distance as the COC widens with the mounting distance... as I said before, think... ??? ...or read one of the references that are on the appendix of the link that I provided... ;)
I think I understand now enough to step out of the "discussion"…
But I wonder if a, say, Sigma Art 50mm lens really has different COC and DOF when mounted on Canon, Nikon and Sigma 135 format bodies, which all have different mounting distances?
(sorry, just could not resist after all…)
-
I think I understand now enough to step out of the "discussion"…
But I wonder if a, say, Sigma Art 50mm lens really has different COC and DOF when mounted on Canon, Nikon and Sigma 135 format bodies, which all have different mounting distances?
(sorry, just could not resist after all…)
Yes... In fact I've mentioned that a bit earlier (look above)... Nikon has a (very) little narrower DOF than Canon... most people would never notice as the mounting distance differs by only 2.5%... With 6x4.5 DSLRs where the mounting distance can differ up to 30% it is quite visible though and with 6X6 even more so... Try A P-6 mount (Exakta 66) lens on a DSLR and it becomes quite obvious (the 180/2.8 is a very good example - no wonder why many still consider it as the bokeh king)... Eric Weiss mentioned that he finds a three stops difference earlier from MF to his DSLRs... This may sound larger than it should to some, but don't forget that Eric uses Rollei which is a 6X6 camera...
-
Hi,
There is a factor called pupil enlargement, but that is dependent lens design and not on mounting distance, per sé. Anyway, it is clearly second order.
So, as long as the lens has an identical design it will yield identical defocus and bokeh for different focusing distances.
Best regards
Erik
I think I understand now enough to step out of the "discussion"…
But I wonder if a, say, Sigma Art 50mm lens really has different COC and DOF when mounted on Canon, Nikon and Sigma 135 format bodies, which all have different mounting distances?
(sorry, just could not resist after all…)
-
The entrance pupil position is another aspect I also mentioned, but although it does affect DOF, it makes a small difference... The main thing that can affect DOF by a considerable value is the COC...
-
I think I understand now enough to step out of the "discussion"…
+1
-
Yeah, good point.
I think much of the discussion is about lack of objectivity much more than any demonstrable facts.
Best regards
Erik
I think I understand now enough to step out of the "discussion"…
But I wonder if a, say, Sigma Art 50mm lens really has different COC and DOF when mounted on Canon, Nikon and Sigma 135 format bodies, which all have different mounting distances?
(sorry, just could not resist after all…)
-
Hi,
There is a factor called pupil enlargement, but that is dependent lens design and not on mounting distance, per sé. Anyway, it is clearly second order.
So, as long as the lens has an identical design it will yield identical defocus and bokeh for different focusing distances.
Best regards
Erik
Naturally the universal lens makers like Sigma or Tamron, make their lenses to mount for a certain distance from the censor, but since the mounting distance between DSLRs differs by a little, one may notice that the lens frame is made longer for Canon than for a Nikon, this obviously means that extra length has been added on the frame to act as a spacer so that the (active) design of the lens is always at constant distantce from the sensor despite the different DSLR that it is mounted on. Hence the mounting distance stays constant and so does the COC and the DOF, but one may notice that a respective Canon lens and a Sigma lens made for Canon may have a (very) small difference between them as far as DOF is concerned....
-
The T/S lens can't be used as an example because it shares the same mounting distance... if the larger format lenses would mount at the same distance as smaller format lenses the COC value wouldn't change...
Mounting distance is irrelevant. It doesn't show up in the optical equations at all. Lenses focus rays to a particular distance, and one designs the mount to hold the glass elements at that distance. But the actual mount distance (flange focal length) is irrelevant, except that it means you cannot mount certain lenses on certain cameras.
-
Yes... In fact I've mentioned that a bit earlier (look above)... Nikon has a (very) little narrower DOF than Canon...
Uh, no.
You can use Nikon lenses on Canon cameras with a mount adapter, and they will show exactly the same DOF. Or, of course, use the same Sigma lens design on both - and both will have the same DOF.
-
Mounting distance is irrelevant. It doesn't show up in the optical equations at all. Lenses focus rays to a particular distance, and one designs the mount to hold the glass elements at that distance. But the actual mount distance (flange focal length) is irrelevant, except that it means you cannot mount certain lenses on certain cameras.
Matthew, both your previous comments are irrelevant to what I say and what I say is totally different. Mounting distance would be irrelevant only if the image circle was the same, it's the larger image circle of larger format lenses that causes the larger value of the COC and DOF is directly related to the COC value...
The mounting distance is not relevant to theory... it only happens that with DSLRs the larger the image circle, the bigger the mounting distance... So, a 6x6 lens mounts further away that 6x4.5 which in return the later mounts further away from the light sensitive area than a 35mm DSLR lens... As a consequence, because larger image area lenses have increased COC and thus reduced DOF, this reduced DOF is then (of course) retained no matter what (cropped) size image area is used behind the lens, but the lens still has to be mounted where it should be for the image area it was originally designed for... and that's why and adapter is used to cover the extra distance...
So as a conclusion, the mounting distance is relevant because all lenses that can be used via an adapter for smaller image areas, have been designed for larger ones, that have increased COC value and thus reduced DOF... which has the direct conclusion that DOF decreases with the mounting distance increase... Does this means that if one uses a 6x4.5 lens on a DSLR it will present reduced DOF for the same focal length and aperture value than another lens that was designed for this DSLR? Yes it does! All that use larger format lenses via an adapter on their DSLRs too can confirm that! I personally use all the Contax 645 series of lenses (but the 350) on my Nikons via a JAS adapter...
-
Does this means that if one uses a 6x4.5 lens on a DSLR it will present reduced DOF for the same focal length and aperture value than another lens that was designed for this DSLR? Yes it does!
How come DOF calculation formulas do not take this into account at all? They only need focal length, relative aperture, distance and film/sensor format to calculate DOF.
-
How come DOF calculation formulas do not take this into account at all? They only need focal length, relative aperture, distance and film/sensor format to calculate DOF.
They do.
(https://upload.wikimedia.org/math/7/7/d/77d46a8b41790040f6615b2a62a887ee.png)
f = Focal length
s = Subject distance (or focus distance)
N = Lens f-number
c = COC
You put the numbers in and get a pretty accurate result. It's not going to be perfect because there are always design, engineering and operating errors at play in the real world, but this is the complete formula to calculate how big will your depth of field be at a given focusing distance.
The formula changes a bit for very close distances (macro).
P.S. COC is calculated separately, and as stated above by others it depends on various factors. It's not directly related to sensor size, but some systems may use that as a quick way to calculate COC based on standard values, thus having a single factor for the main sensor formats for quick selection and calculations. It's not going to be accurate for all cameras with a certain sensor size, but it's a simplification that is usually close enough. Maybe that is why you think they don't take other factors into account.
-
How come DOF calculation formulas do not take this into account at all? They only need focal length, relative aperture, distance and film/sensor format to calculate DOF.
They do take COC into account, in fact the existance of COC is why DOF exists, so it couldn't be otherwise... But I thought you have jumped out of the conversation, which is a very good thing to do instead of you (and some others) misinforming people... ;) :P
-
I still do not see the mount distance in the calculations. This is getting weird. Focal length: yes. Image circle size: never.
If I put a lens shade which is too long on a lens, it will make the image circle smaller. Is the DOF going to change?
-
I still do not see the mount distance in the calculations. This is getting weird. Focal length: yes. Image circle size: never.
It's because you don't want to see it... It has been explained to you (and others) at elementary school levels many times above... ??? Once more then... ;D To have larger COC value (which directly affects DOF) the lens must be able to project a larger image area... But on DSLRs (of what ever format) for this to happen the mounting distance increases respectively... ;) I hope now that it has gone below elementary school level you (and some "usual" others) get it... :P
EDIT: It's like using a 24x36mm old MFDB on an MF camera, if you use a (say) 180/2.8 lens on it at full aperture, you have "thinner" DOF than using a same focal length and speed lens on a Nikon or a Canon.... Why? ...Because COC value has changed! ...simple isn't it? (real elementary school level) 8)
-
To have larger COC value (which directly affects DOF) the lens must be able to project a larger image area.
Lenses do not have COC values.
-
Lenses do not have COC values.
Size has a value (so that it can be measured)... but I think this is assumed even at elementary schools and thus it is not taught... :o :'( :P :-*
-
Hi,
CoC, as used in DoF calculations depends only on viewing distance and magnification. It is the largest disk of unsharpness that seems acceptable to the eye at a given viewing distance.
But, the sensor image is normally much smaller than any image viewed at a realistic distance, so the CoC value needs to be multiplied the image magnification.
Say we want to have an 7x5" print viewed at 10" and let's assume that it needs a CoC of 0.15 mm to appea sharp. Now, a 24x36 image would need to be enlarged around 5 times for 5"x7", so we would need a CoC of 0.15 / 5 = 0.03 mm for a "sharp image".
Now, assume that we use a P45+ sensor instead which is 49x37 mm. Looking at the 37 mm dimension we would need around 3.5x magnification, so the needed CoC on the P45+ would be 0.15/3.5 around 0.43 mm.
This has nothing to do with flange distance, just image scaling and human vision.
Now, with a larger sensor we need a longer lens for same field of view. The longer lens will give a shorter DoF at similar aperture.
On the other hand, short flange mounting distances allow for more symmetric lenses, which have less aperture enlargement. But aperture enlargement plays a small role in DoF calculations.
So neither flange distance or CoC are really related to lenses and both are red herrings in this context.
Best regards
Erik
Lenses do not have COC values.
-
They do.
(https://upload.wikimedia.org/math/7/7/d/77d46a8b41790040f6615b2a62a887ee.png)
f = Focal length
s = Subject distance (or focus distance)
N = Lens f-number
c = COC
You put the numbers in and get a pretty accurate result. It's not going to be perfect because there are always design, engineering and operating errors at play in the real world, but this is the complete formula to calculate how big will your depth of field be at a given focusing distance.
The formula changes a bit for very close distances (macro).
P.S. COC is calculated separately, and as stated above by others it depends on various factors. It's not directly related to sensor size, but some systems may use that as a quick way to calculate COC based on standard values, thus having a single factor for the main sensor formats for quick selection and calculations. It's not going to be accurate for all cameras with a certain sensor size, but it's a simplification that is usually close enough. Maybe that is why you think they don't take other factors into account.
That's essentially correct. It's only about focal length, (effective) aperture, subject distance (or magnification factor), and COC (fixed value determined by viewing constraints). Pupil factor can be added to the equation for increased accuracy with asymmetrical lens designs. Image circle doesn't play a role as long as it's larger than the sensor.
Cheers,
Bart
-
Lenses do not have COC values.
Correct. COC is determined by viewing distance and acceptable amount of magnified blur at that distance.
Cheers,
Bart
-
Please, CoC is not a structural characteristic of a lens, it is the maximum amount of defocusing or unsharpness that will be indistinguishable from a perfectly focused object, based on OUTPUT (will I be able to stress this enough?) size and viewing distance. It is a concept based on the angular resolving power of the human eye. IT HAS NOTING TO DO WITH IMAGE CIRCLE.
Now, in relation to mounting distance, I think there is an indirect relation between it and DOF, but the real factor is pupil magnification. I will try to explain
Pupil is referred to the apparent size of the aperture. Entrance pupil is viewed from the front, exit is viewed from the back of the lens
Pupil magnification is the ratio between exit pupil and entrance pupil
Think of three common lens designs:
- Symmetrical - used typically in normal lenses and most large format lenses, the pupil magnification is 1
- Retro-focus - Typical of wide angle lenses, to allow greater distance from the sensor/film plane and clear the mirror box in DSLR/SLR , pupil magnification is greater than 1
- Tele - Allow for a lens that is shorter than its focal length and so it can be mounted closed to the sensor plane and pupil magnification is less than 1
What is the effect of this?
The Pupil magnification is a factor that does affect DOF and it is included in the complete equations, but it is usually discarded because its effects are only relevant at close focusing distances
The effect of pupil magnification is Inverse to DOF, which means, the greater the pupil magnification (as in retrofocus designs) the smaller the DOF. Since the retrofocus lens will have a greater mounting distance than a symmetrical lens, you can erroneously assume that the change in DOF is due to mounting distance, but it is really due to pupil magnification.
Another effect at close focusing distances: Lens focal length, it turns out that for a given field of view, the longer the lens, the shorter the DOF for the same entrance pupil diameter.
What to consider as a close focusing distance? A rough approximation is 10 times the focal length.
-
Lenses do not have COC values.
Correct... none ever said that they do! The COC has a (measurable) size though, depending on the viewing distance... Still the size is determined with respect to the whole projected image area... "value" is a number used in the equation that razrbick posted... Now we can all play with words to "cover" Petrus trolling all this time...
-
Please, CoC is not a structural characteristic of a lens, it is the maximum amount of defocusing or unsharpness that will be indistinguishable from a perfectly focused object, based on OUTPUT (will I be able to stress this enough?) size and viewing distance. It is a concept based on the angular resolving power of the human eye. IT HAS NOTING TO DO WITH IMAGE CIRCLE.
Now, in relation to mounting distance, I think there is an indirect relation between it and DOF, but the real factor is pupil magnification. I will try to explain
Pupil is referred to the apparent size of the aperture. Entrance pupil is viewed from the front, exit is viewed from the back of the lens
Pupil magnification is the ratio between exit pupil and entrance pupil
Think of three common lens designs:
- Symmetrical - used typically in normal lenses and most large format lenses, the pupil magnification is 1
- Retro-focus - Typical of wide angle lenses, to allow greater distance from the sensor/film plane and clear the mirror box in DSLR/SLR , pupil magnification is greater than 1
- Tele - Allow for a lens that is shorter than its focal length and so it can be mounted closed to the sensor plane and pupil magnification is less than 1
What is the effect of this?
The Pupil magnification is a factor that does affect DOF and it is included in the complete equations, but it is usually discarded because its effects are only relevant at close focusing distances
The effect of pupil magnification is Inverse to DOF, which means, the greater the pupil magnification (as in retrofocus designs) the smaller the DOF. Since the retrofocus lens will have a greater mounting distance than a symmetrical lens, you can erroneously assume that the change in DOF is due to mounting distance, but it is really due to pupil magnification.
Another effect at close focusing distances: Lens focal length, it turns out that for a given field of view, the longer the lens, the shorter the DOF for the same entrance pupil diameter.
What to consider as a close focusing distance? A rough approximation is 10 times the focal length.
Does this (irrelevant) mean that your calculation that you posted earlier is correct? ...In other words, can DOF differ more than two stops between a FF DSLR and an MF camera for the same AOV or not? ....Can the difference be up to three stops (as Eric Weiss posted earlier) or not? ...by the way... wasn't you (as well) leaving the conversation earlier?
-
I propose to some people, to do some more (homework) reading and come back next week to play the "Einstein"...
-
Does this (irrelevant) mean that your calculation that you posted earlier is correct? ...In other words, can DOF differ more than two stops between a FF DSLR and an MF camera for the same AOV or not? ....Can the difference be up to three stops (as Eric Weiss posted earlier) or not? ...by the way... wasn't you (as well) leaving the conversation earlier?
Yes, it can be more than two stops at close focusing distances. The equation I posted earlier is the simplified version that works for distances greater than 10 X focal length (which is correct). I have no issue in reviewing and correcting myself if proven wrong.
I really don't care what you (Theodoros) think, my concern is with the vast audience that uses LuLa as reference, that might get the wrong concepts and ideas about this subject
-
Yes, it can be more than two stops at close focusing distances. The equation I posted earlier is the simplified version that works for distances greater than 10 X focal length (which is correct). I have no issue in reviewing and correcting myself if proven wrong.
I really don't care what you (Theodoros) think, my concern is with the vast audience that uses LuLa as reference, that might get the wrong concepts and ideas about this subject
This (your later comment above) is certainly better than supporting any troll earlier with a "+1" comment...
EDIT: By the way... can you define "close focus distances"? I ask this because if I'm not mistaken, where shallow DOF maters, it's always "close focus distances"... I mean, who cares if DOF covers infinity at a long distance landscape anyway?
-
Hi,
Petrus has indicated that pupil magnification stated to be significant at around 10x focal length.
But, that is not really relevant as macro lenses are normally used at closes distances and those used to be quite symmetric designs. As macro lenses are often on the longer side, there is little need to inverted telephoto designs. Indeed, most of the classic macro lenses are symmetric double gauss designs, perhaps combined with a field flattening group and variable airspace.
Modern 135 macro lenses can be quite complex designs, however, so they may be a bit different.
When Eric Hiss talks about 3 stops of difference in DoF he compares an 80 mp back to a 36 MP 135 DSLR. Not clear from his posting if he compares prints of similar size or actual pixels on screen. This is an example of publishing the findings but not disclosing the methods can give confusing results.
Best regards
Erik
PS
Just to say, some one sharing insights and knowledge isn't exactly what I would call trolling.
This (your later comment above) is certainly better than supporting any troll earlier with a "+1" comment...
EDIT: By the way... can you define "close focus distances"? I ask this because if I'm not mistaken, where shallow DOF maters, it's always "close focus distances"... I mean, who cares if DOF covers infinity at a long distance landscape anyway?
-
It is all in the equations. Einstein, who was mentioned earlier, played around with his equations, and to his surprise, E=mc2 popped out. As the calculations were correct, this result had to be taken at the face value, and it has held ever since.
There is no way to squeeze the size of the lens image circle (or mounting/flange distance, which just a mechanical engineering decision, not optical) out of any DOF equation known to man.
That is all there is to it.
-
(http://m.quickmeme.com/img/ab/ab54ab431b60377a1e19c1b597607651514a52166670e22e13825749a3cea5c3.jpg)
-
When Eric Hiss talks about 3 stops of difference in DoF he compares an 80 mp back to a 36 MP 135 DSLR. Not clear from his posting if he compares prints of similar size or actual pixels on screen. This is an example of publishing the findings but not disclosing the methods can give confusing results.
Yes quite right. When Eric Hiss posts about what he found by actually testing in real life with real cameras and real models in the studio, he knows that people who are doing same will find the information useful and carry on, and conversely those who enjoy to spend time in forum discussion will continue on in an endless circle endlessly anyhow. Eric Hiss also understands the difference between publishing a paper in a respected academic environment with peer review and posting a few empirical observations to a forum full of old men who have nothing better to do than pick at each other.
-
EricWhiss, I'm genuinely interested in finding what is going on. You are making a claim which does not agree with theory, which is fine, as long as you provide the evidence (if you have experience publishing scientific papers you should be used to that, at least I am)
Nobody is forcing you to read this thread Synn, If you and many more are not interested just skip it, I don't have any issue.
Since I agree that this thread is not going anywhere if we don't provide any evidence, I made a test, which I will explain in detail.
If my methodology is wrong I'm the first one interested in rectifying it.
Test: comparison of DOF of images at same OUTPUT MAGNIFICATION using the following lenses:
1.- Hasselblad C 80mm Planar f2.8 (medium format)
2.- Nikon AIS 50mm f1.4
The camera used was a nikon D800. (the test is based on a different crops for each lens, so that OUTPUT MAGNIFICATION of the crop is the same, which would be equivalent to have the same angle of view in both formats).
Camera and subject was constant
Focusing distance was the closest possible with the Hasselblad 80mm (aprox 95 cm) without extension tubes
All files processed with default settings in LR, only the White balance was adjusted
Here are the results:
For the first test, the image to the left is the Hassy 80 @ f5.6, the center is the nikon @ f2.8 (2 stops compared to the 80mm) and the right one is the nikon @ f2 (3 stops compared to the 80mm)
(http://www.frankdisilvestro.com.au/img/s1/v46/p1468335347-4.jpg)
The second test is the Hassy 80 @ f4.0, then the nikon @ f2 (2 stops compared to the 80mm) and finally at the right is the nikon @ f1.4 (3 stops compared to the 80mm)
(http://www.frankdisilvestro.com.au/img/s2/v61/p1468335340-4.jpg)
My observation is that the difference between the two lenses which translates to same output size with equivalent fields of view is 2 stops.
I would really like (seriously) if you can share the results that show 3 f stops difference
-
Since I agree that this thread is not going anywhere if we don't provide any evidence, I made a test, which I will explain in detail.
If my methodology is wrong I'm the first one interested in rectifying it.
The camera used was a nikon D800. (the test is based on a different crops for each lens, so that OUTPUT MAGNIFICATION of the crop is the same, which would be equivalent to have the same angle of view in both formats).
My observation is that the difference between the two lenses which translates to same output size with equivalent fields of view is 2 stops.
I would really like (seriously) if you can share the results that show 3 f stops difference
If I may Francisco... Where the test suffers is that you used crop (and thus magnified the FF lens result which lead to a perception of shallower DOF)... To do a proper test, you should have either used two lenses of the same focal length (one MF and one 35mm) and keep the image area of the 35mm DSLR constant... (where the difference would be of about a stop - the stop that you are missing) or to use 6x6 film with the Hasselblad and aim for an AOV that would be the same on the diagonal of the square vs. the 3:2 ratios (put the scale on the diagonal)...
-
Test: comparison of DOF of images at same OUTPUT MAGNIFICATION [...]
Correct, that's the only sensible thing to do. That will factor in the need for different output magnification due to physical sensor size differences.
using the following lenses:
1.- Hasselblad C 80mm Planar f2.8 (medium format)
2.- Nikon AIS 50mm f1.4
The camera used was a nikon D800. (the test is based on a different crops for each lens, so that OUTPUT MAGNIFICATION of the crop is the same, which would be equivalent to have the same angle of view in both formats).
Yes, both are relatively symmetrical designs, and the focal lengths roughly offer a comparable field of view (as far as the aspect ratio differences allow to compare). To satisfy my curiosity, and to run the DOF calculation numbers on, on which Hasselblad was that?
My observation is that the difference between the two lenses which translates to same output size with equivalent fields of view is 2 stops.
Yes, 2 stops results in pretty close similarity, as expected.
Cheers,
Bart
-
Hi,
Thanks for sharing. I can do a demo when I get home with a P45+ and a Macro Planar 120 and a Sony 90/2.8 macro on the A7rII. These systems are well matched regarding resolution.
Main main issues with Eric Hiss statement are that they contradict theory, which is a serious thing in science/egineering, and that he doesn't report on how the observations were made. If his observations were made at actual pixels, criteria would differ between systems having different resolutions.
Personally, I have been shooting MFD , P45+ and Sony DSLR, mostly 24MP. I have not seen any magic differences. But I have not done a lot of detailed tests. But all tests I have made the systems are quite similar, except the P45+ having much higher resolution.
Best regards
Erik
EricWhiss, I'm genuinely interested in finding what is going on. You are making a claim which does not agree with theory, which is fine, as long as you provide the evidence (if you have experience publishing scientific papers you should be used to that, at least I am)
Nobody is forcing you to read this thread Synn, If you and many more are not interested just skip it, I don't have any issue.
Since I agree that this thread is not going anywhere if we don't provide any evidence, I made a test, which I will explain in detail.
If my methodology is wrong I'm the first one interested in rectifying it.
Test: comparison of DOF of images at same OUTPUT MAGNIFICATION using the following lenses:
1.- Hasselblad C 80mm Planar f2.8 (medium format)
2.- Nikon AIS 50mm f1.4
The camera used was a nikon D800. (the test is based on a different crops for each lens, so that OUTPUT MAGNIFICATION of the crop is the same, which would be equivalent to have the same angle of view in both formats).
Camera and subject was constant
Focusing distance was the closest possible with the Hasselblad 80mm (aprox 95 cm) without extension tubes
All files processed with default settings in LR, only the White balance was adjusted
Here are the results:
For the first test, the image to the left is the Hassy 80 @ f5.6, the center is the nikon @ f2.8 (2 stops compared to the 80mm) and the right one is the nikon @ f2 (3 stops compared to the 80mm)
(http://www.frankdisilvestro.com.au/img/s1/v46/p1468335347-4.jpg)
The second test is the Hassy 80 @ f4.0, then the nikon @ f2 (2 stops compared to the 80mm) and finally at the right is the nikon @ f1.4 (3 stops compared to the 80mm)
(http://www.frankdisilvestro.com.au/img/s2/v61/p1468335340-4.jpg)
My observation is that the difference between the two lenses which translates to same output size with equivalent fields of view is 2 stops.
I would really like (seriously) if you can share the results that show 3 f stops difference
-
Hi,
Thanks for sharing. I can do a demo when I get home with a P45+ and a Macro Planar 120 and a Sony 90/2.8 macro on the A7rII. These systems are well matched regarding resolution.
And more importantly for a DOF comparison, their Angles of View are quite closely comparable:
A7rII & 90mm (angle of view: h:20.5° × v:13.8°, d:24.5°) (with close-up shooting , they get wider at infinity)
P45+ & 120mm (angle of view: h:20.3° × v:15.3°, d:25.2°) (with close-up shooting , they get wider at infinity)
So then only the output magnification becomes the variable for equal size output, and total DOF should be relatively easy to compare.
Shooting close-up also allows to more accurately determine the magnification factor (on sensor), which allows to use the more accurate DOF calculation formula (attached) that applies at any distance, with the additional possibility to factor in the pupil factor variable (P in the formulas).
Cheers,
Bart
P.S. The attached formulas also allow to compensate for the change in apparent or effective aperture at non-infinity focus distances, if the magnification factor is known or calculated from shooting distance (from first principal plane, which is less accurately known at close distances). All variables in the same units, f=focal length, s=shooting or subject distance, M=magnification factor, N=(effective) aperture, P=pupil factor (the Exit/Entrance pupil diameter ratio at the given focus setting).
-
Correct, that's the only sensible thing to do. That will factor in the need for different output magnification due to physical sensor size difference.
Cheers,
Bart
I don't see how the output magnification in the test that Fransisco did factors for the different image circles that the two lenses project... The 6x6 lens projects an image circle that is 4 times the area of the image circle that a 35mm lens projects... (assuming that both lenses are designed to "just" cover the respective light sensitive areas of their respective cameras).
EDIT: Also, in the test that Erik proposes, the P45+ requires an image circle of (about) 65% the area that the lenses of the V system are able to project... So in both cases, one compares DOF for different FOV than the one that the MF lens (really) projects... Surely, if one will use a "full frame" MF back, (like Eric did) results will be different (and even more so if one uses the full image area that a 6x6 camera has).
-
I don't see how the output magnification in the test that Fransisco did factors for the different image circles that the two lenses project...
The size of (the part of) the image circle (used for composition) only matters for composing the shot, it is irrelevant for DOF. The magnification to (equal sized) output is relevant for the COC assumption, because the (acceptable) optical blur gets magnified when resizing for output.
The 6x6 lens projects an image circle that is 4 times the area of the image circle that a 35mm lens projects... (assuming that both lenses are designed to "just" cover the respective light sensitive areas of their respective cameras).
Because the physically smaller sensor needs more magnification than the larger sensor to achieve the same output size, it needs a different COC assumption, or a different aperture, for the DOF calculation. If the output size is equalized (which dictates the observable COC in output), then only aperture remains to make a difference.
Cheers,
Bart
-
Hi Bart,
Modern lenses often use internal focusing and floating lens groups, both affecting focal length. So how do we determine effective aperture?
Testing at medium magnifications may be a workaround for that problem?
Best regards
Erik
And more importantly for a DOF comparison, their Angles of View are quite closely comparable:
A7rII & 90mm (angle of view: h:20.5° × v:13.8°, d:24.5°) (with close-up shooting , they get wider at infinity)
P45+ & 120mm (angle of view: h:20.3° × v:15.3°, d:25.2°) (with close-up shooting , they get wider at infinity)
So then only the output magnification becomes the variable for equal size output, and total DOF should be relatively easy to compare.
Shooting close-up also allows to more accurately determine the magnification factor (on sensor), which allows to use the more accurate DOF calculation formula (attached) that applies at any distance, with the additional possibility to factor in the pupil factor variable (P in the formulas).
Cheers,
Bart
P.S. The attached formulas also allow to compensate for the change in apparent or effective aperture at non-infinity focus distances, if the magnification factor is known or calculated from shooting distance (from first principal plane, which is less accurately known at close distances). All variables in the same units, f=focal length, s=shooting distance, M=magnification factor, N=(effective) aperture, P=pupil factor (the Exit/Entrance pupil diameter ratio at the given focus setting).
-
The size of (the part of) the image circle (used for composition) only matters for composing the shot, it is irrelevant for DOF. The magnification to (equal sized) output is relevant for the COC assumption, because the (acceptable) optical blur gets magnified when resizing for output.
Because the physically smaller sensor needs more magnification than the larger sensor to achieve the same output size, it needs a different COC assumption, or a different aperture, for the DOF calculation. If the output size is equalized (which dictates the observable COC in output), then only aperture remains to make a difference.
Cheers,
Bart
I believe that for theory (and the formulas) to be applicable, one must adjust the AOV (in the diagonal of the image circle) for the whole image area that the lens has been designed for... The other (and easier) way to do an accurate test (for certain lenses) is to use two lenses of the same focal lengths at the same aperture setting, shoot both on the same sensor with the scale on the diagonal and then magnify the larger format lens image by the respective difference of its larger image circle area... (four times in this case)... IMO, that would give a (pretty) accurate approach...
-
Hi Bart,
Modern lenses often use internal focusing and floating lens groups, both affecting focal length. So how do we determine effective aperture?
Hi Erik,
You are right that focal length will remain a somewhat uncertain (limited precision) parameter in all DOF calculations.
Because M is determined by focal length and shooting distance (my first formula attachment earlier), the exact focal length is an important factor. However, it is also not obvious where the first principal plane of a lens is situated. So there we have to deal with two uncertainties/imprecisions.
That's why it's IMHO easier to use the effective result of both, the magnification factor itself, and calculate with that. A closer shooting distances we can quite accurately measure the magnification factor. Just shoot a ruler or other object of known size, and measure the size it produces on the sensor (pixels x pitch, or percentage of sensor width). Then divide that sensor projection size by the original subject size, and we have out magnification factor.
That will only leave us with a single uncertainty, the real focal length.
We do have the pupil factor as an additional correction though, and it can be approximated by looking/measuring the apparent entrance and exit pupil of the lens when looking though it.
Testing at medium magnifications may be a workaround for that problem?
Yes, up tot a point. With the magnification factor as a known certainty, the effective aperture can be factored in (which helps accuracy), and the focal length 'only' adds an uncertainty to the total DOF that's roughly twice the uncertainty of the focal length. IOW, a 1% change in assumed focal length, will result in a ~2% change in DOF. When we shoot from a closer distance, the magnification factor becomes larger and weighs more in the overall calculation, so the effect of the focal length uncertainty gets slightly smaller (but hardly significantly so).
I just prefer to have only one uncertain parameter (focal length) instead of two (focal length and distance). It also becomes harder to set the distance to an exact position (or measure it without a laser distance finder) as the distance increases, and it becomes harder to eyeball the visual DOF difference in output with increased distance.
So a medium distance (if that's what we normally shoot), or a bit closer (say 1 - 2 metres), seems good for testing to me. Of course DOF becomes more critical at even closer distances (and easier to see the acceptable limits), so it helps if the formula is more accurate for those distances as well.
Cheers,
Bart
-
EricWhiss, I'm genuinely interested in finding what is going on. You are making a claim which does not agree with theory, which is fine, as long as you provide the evidence (if you have experience publishing scientific papers you should be used to that, at least I am)
First off, your Hasselblad probably has a smaller sensor than my AFi-ii 12 - if you wrote what sensor I didn't see it, secondly I'm not making a claim, rather I shared an observation, and theres a difference between observation and claim. Lastly, a lot of this DOF is arbitrary and depends on usage and interpretation which is subjective. Results will likely depend on a number of other factors too such as equipment used, technique, and so on. Trying to pin down the difference between your two stops and my three stops will be difficult.
-
I believe that for theory (and the formulas) to be applicable, one must adjust the AOV (in the diagonal of the image circle) for the whole image area that the lens has been designed for...
If I understand what you are suggesting, then no. The only thing that matters is the diameter of the cone of rays that is intersected by the sensor, the cone that is focused slightly in front or behind of the sensor. The intersection is a sort of circular disc, depending on the shape of the aperture, and possible vignetting (by the lens barrel) in the shape of an oval on one or two sides, and it is the blurred image of our subject. The diameter of it is the COC, at the limit of acceptable blur. So it's the exit pupil, the size of which is determined by the aperture / the entrance pupil, that sets the angular limits of the cone of rays, not the designed image circle of the lens. A narrower cone can be intersected further away from the plane of best focus, before it exceeds the COC limit we need for sharp enough output after magnification.
Cheers,
Bart
-
Hi,
Just my closing (maybe) comment since I don't want to leave unanswered questions.
To satisfy my curiosity, and to run the DOF calculation numbers on, on which Hasselblad was that?
First off, your Hasselblad probably has a smaller sensor than my AFi-ii 12
It seems that I was not specific enough in my first post. Both lenses were mounted on a Nikon D800. The Hassy (zeiss) 80mm was mounted using a fotodiox adapter. The rationale here is that If I enlarge the image taken with the 50mm 1.6X with respect to the image taken with the 80mm (the relation between the lenses), I get the same output magnification, which would be the same as having used a medium format sensor (1.6x larger than 24x36) and producing an output image of the same size.
In case it is not clear, I used crops that show the same scene object. These crops have different size in pixels but show the same scene object, when you enlarge the crop of the 50mm lens by1.6 you get the same physical size as the one taken from the 80mm.
I guess that Theodoros and EricWHiss will not agree with this, maybe Bart and Erik Kafehr will, but that is my understanding on how to compare DOF across different formats, so I will not continue arguing about it.
I'm not making a claim, rather I shared an observation, and theres a difference between observation and claim. Lastly, a lot of this DOF is arbitrary and depends on usage and interpretation which is subjective. Results will likely depend on a number of other factors too such as equipment used, technique, and so on. Trying to pin down the difference between your two stops and my three stops will be difficult.
All I was asking is if you can share some of the images / evidence from which you based your observations
------------------------------------------------------------------------------------------
Anyway, Playing with the detailed equations and doing the tests I learned many things that affect DOF at close ranges or less than 10X the focal length, especially the pupil magnification and the focal length. Another learning is that those variations are not symmetrical with respect the point of exact focus (based on CoC size).
Finally the following image is a comparison of a Hasselblad/Zeiss C 50mm F/4 Distagon (medium format, old type, no internal moving elements) and the Nikon 50mm f/1.4 using the same setup as before, at the closest possible distance without using extension tubes (ca. 45 cm.)
The left image is the Hassy @ 4.0, the middle is the Nikon 50 @ 4.0 and the right one is the nikon @ 2.8
(http://www.frankdisilvestro.com.au/img/s4/v67/p1470675520-4.jpg)
My observation: DOF with the medium format lens is slightly less than the nikon 50 using the same aperture, with more difference in the near region than in the far region but not enough as 1 stop
Regards,
-
Francisco,
My observation: DOF with the medium format lens is slightly less than the nikon 50 using the same aperture, with more difference in the near region than in the far region...
From this (http://www.cambridgeincolour.com/tutorials/depth-of-field.htm) article (refer also to the table approximately half way down):
"Even though the total depth of field is virtually constant, the fraction of the depth of field which is in front of and behind the focus distance does change with focal length".
Could this explain your real-world difference in the near- and far-region?
-
Francisco,
From this (http://www.cambridgeincolour.com/tutorials/depth-of-field.htm) article (refer also to the table approximately half way down):
"Even though the total depth of field is virtually constant, the fraction of the depth of field which is in front of and behind the focus distance does change with focal length".
And it changes with distance. At macro distances it's equal in front and to the rear, so a front:rear ratio of 1:1, while at the hyperfocal distance it's a ratio of 1:infinity. This and many more details can be found here (http://toothwalker.org/optics/dof.html) (one of the, if not the, best resources for such matters).
Could this explain your real-world difference in the near- and far-region?
Yes, but to quantify one would need to do separate calculations for front DOF and rear DOF, instead of total DOF. A longer focal length has a hyperfocus point that is more distant, so it changes between 1:1 to a 1:infinity DOF ratio at a different pace.
The total DOF formula is also useful for close-up photography that uses Focus stacking with a focus rail, which keeps the magnification constant (but changes perspective), unlike refocusing which changes the total DOF for each shot because the magnification changes. One can use the sensel pitch as COC limit, and calculate how large the step distance for all shots must be. For refocusing one could assume the worst, and use the total DOF for the closest distance (largest magnification) and gradually have surplus DOF as one focuses further away.
Cheers,
Bart
-
Hi Bart and Rob,
I haven't had the time to check in detail the information referred by Bart and so far my preliminary results are theoretical, using the formulas available here (http://toothwalker.org/optics/dofderivation.html)
The effect I am talking about is due to pupil magnification only. Given the same focal lenght and same subject distance, the greater the pupil magnification the lower the depth of field, but this change is different in the near and far regions (this is what I refer as non-symetrical). Anyway I need to review this subject further to be sure it is right.
-
Anyway, Playing with the detailed equations and doing the tests
Just a warning: the equations you are playing with ignore all optical aberrations of the lenses (unless you are using ray tracing software). These aberrations are essential to characterise how "unsharp" the out of focus areas of a given lens are.
-
Just a warning: the equations you are playing with ignore all optical aberrations of the lenses (unless you are using ray tracing software). These aberrations are essential to characterise how "unsharp" the out of focus areas of a given lens are.
Absolutely! The simplified models used are just to start to understand what is going on.
-
The perception of depth of field is influenced by many factors, for example the grain of the image or lack of grain, the resolution of the screen or print and so forth. I wouldn't be surprised if different raw conversion and parameters used will also impacted a persons perception of DOF. What looks like its within DOF on screen may not look like its still inside the DOF when printed. And of course there is the print size variable, and variation in peoples eye sight as well. 9274
If you have a very good lens and sensor, and a quantitative testing method, you may be able to observe the focus peak. I saw this with my Rolleiflex Schneider lenses using imatest and an 8ft chart shooting tethered. I was shooting at 3-4 meters and could notice a jump in sharpness at the peak of focus. I don't think I would have seen this with normal focusing, but I had my camera mounted on a geared rail and was using the rail to tune the focus following the numbers on Imatest to see the max. Taking steps smaller than 1mm you can see this jump in sharpness right at the peak - move 2 mm steps and you might miss it. This was observed with several lenses and with different aperture settings.
It would be interesting to graph the sharpness over distance to see how the sharpness rolls off from the focus peak in either direction. I don't have time for this now, but I think it would be interesting to compare the roll off curve between formats. My hunch is that the rate is faster with larger formats. Using this method you could also make some better study of the differences in DOF and learn how much of the DOF falls behind the peak or in front. If you wanted to measure the difference between say your Hasselblad and your DSLR then you could take lenses with equivalent lenses for each format and arbitrarily decide on a % of peak sharpness for the border of the DOF. Then you could measure the DOF as you defined for the DSLR at say f/6.8 and then see what f/stop the MFDB needed to be set at to get the same DOF as defined by a drop from Peak sharpness to a % you chose. If you had lots of time you could also repeat this with other different formats. I guess if I were doing this test myself, I'd also want to process out the files the best way I knew how for each system - meaning using the best RAW converter for each camera. Imatest tells you whether you have over sharpened or not.
-
Just a warning: the equations you are playing with ignore all optical aberrations of the lenses (unless you are using ray tracing software). These aberrations are essential to characterise how "unsharp" the out of focus areas of a given lens are.
you can use the same MF back with the same lens and just consider a crop from it matching the area of FF sensor... that shall be better.
-
If you have a very good lens and sensor, and a quantitative testing method, you may be able to observe the focus peak. I saw this with my Rolleiflex Schneider lenses using imatest and an 8ft chart shooting tethered. I was shooting at 3-4 meters and could notice a jump in sharpness at the peak of focus. I don't think I would have seen this with normal focusing, but I had my camera mounted on a geared rail and was using the rail to tune the focus following the numbers on Imatest to see the max. Taking steps smaller than 1mm you can see this jump in sharpness right at the peak - move 2 mm steps and you might miss it. This was observed with several lenses and with different aperture settings.
It would be interesting to graph the sharpness over distance to see how the sharpness rolls off from the focus peak in either direction. I don't have time for this now, but I think it would be interesting to compare the roll off curve between formats.
Hi Eric,
I also did similar tests (see attachment), using a Stackshot rail at some 3.34 metres distance on a slanted edge target, to find the best/highest focus/resolution possible at various apertures, as a prelude for deconvolution sharpening. In my findings I did not observe a (sharp) peak, but a gradual optimum (which is what I expected because defocus blur does not change focus abruptly, but rather more gradual).
My hunch is that the rate is faster with larger formats.
I don't see why it would. After all, all that happens when we focus, we make the focus plane intersect the exit pupil's cone of rays at slightly different positions. That cone of rays has a constant acceptance angle for each focus point/sensel, so equal steps will show a gradual increase/decrease. When the same focal length and aperture are used, then the size of the sensors makes no difference, the cones are basically the same.
Cheers,
Bart
-
Hi Bart,
My experience is more a bit like your diagram. I see some maximum in sharpness, but it is more at the 2cm level at 4.0 meters than on the mm level.
Regarding "standard sharpening" in Imatest, it may be a red herring. I wrote Norman Koren about it, asking if it was an indication for good or adequate shrapening, an he cautioned about it.
My experience is that I can apply a large amount of sharpening at small radius and still have undersharpening according to Imatest.
To understand Imatest "standard sharpeinung" it may be helpful the read the relevant information at Imatst: http://www.imatest.com/docs/sharpening/
My take on the issue is that Imatest assumes a sharpening that will result in MTF 100% at 0.3 x Nyquist frequency. When we sharpen at actual pixels we tend to sharpen with a large amount at small radius. This will enhance "micro contrast". But, that "micro contrast" will normally contribute very little to perceived sharpness in a print, as medium frequency detail will dominate.
If I push pixel level sharpening in Lightroom and analyse in Imatest I will always end up with under sharpening. On the other hand I can try some more balanced sharpening like the "high frequency/landscape" presets in Lightroom and apply some sharpening say 15% at radius 2 in Photshop unsharp mask, and that will be pretty good sharpening according to Imatest.
Using "landscape" sharpening and say 15% unsharp mask at radius = 2 will give excellent SQF values in Imatest.
Focus magic gives pretty optimal results in Imatest. No MTF above 1.0 in 0-0.3 Nyquist, MTF near 1.0 at 0.3 Nyquist and a decent level of MTF at Nyquist. Also, FocusMagic doesn't enhance noise as much as many other methods of sharpening.
As a side note, the "Orange Peel" artefacts "Diglloyd" has found may be a sharpening artefact, possibly combined with aliasing. What I may have seen is that sharpening using FocusMagick will not cause some artefacts which may be the foretold "orange peel" artefacts.
So, I would say that I really appreciate FocusMagic, it is one of the best tools I have found for sharpening. But, it doesn't fit my parametric workflow. Would I do a large print, I would probably deploy FocusMagic and some other tools, but for 90% of my work I try to get around with the tools provided with LR 6.
Now, getting back to MFD, what I have seen is that there are no magick differences between my MFD gear and my Sony Alpha 99 gear. The great advantage of MFD I see is resolution, but that advantage takes like A1-size prints to show up, unless I view my images with a loupe.
I am pretty sure that those 80MP backs, paired with excellent lenses, have a resolution advantage over present day high MP 135, even if the best lenses are used. With lesser backs, say 40-50 MP I am a bit skeptical.
One interesting observation is that a frquent poster here in LuLa, Chris Barret, shifted over his work from IQ-260 (I think) to Sony A7r. He uses it with Canon T&S lenses and Hasselblad CFi glass, pretty much what I have. He feels that the old Hasselblad/Zeiss glass works very well on the A7r.
Rainer Viertlboeck, a well known German architecture photographer is also quite happy with the A7r: http://forum.luminous-landscape.com/index.php?topic=90358.0
So, it seems that real commercial photographers see real benefits with high MP 135.
Personally, I have been shooting my A7rII for just three weeks so I don't feel I can make a lot of comments.
Best regards
Erik
Hi Eric,
I also did similar tests (see attachment), using a Stackshot rail at some 3.34 metres distance on a slanted edge target, to find the best/highest focus/resolution possible at various apertures, as a prelude for deconvolution sharpening. In my findings I did not observe a (sharp) peak, but a gradual optimum (which is what I expected because defocus blur does not change focus abruptly, but rather more gradual).
I don't see why it would. After all, all that happens when we focus, we make the focus plane intersect the exit pupil's cone of rays at slightly different positions. That cone of rays has a constant acceptance angle for each focus point/sensel, so equal steps will show a gradual increase/decrease. When the same focal length and aperture are used, then the size of the sensors makes no difference, the cones are basically the same.
Cheers,
Bart
-
Just a warning: the equations you are playing with ignore all optical aberrations of the lenses (unless you are using ray tracing software). These aberrations are essential to characterise how "unsharp" the out of focus areas of a given lens are.
you can use the same MF back with the same lens and just consider a crop from it matching the area of FF sensor... that shall be better.
By changing the sensor size, you will change the apparent depth of field on prints or on screen, so: no, you can't do that.
-
By changing the sensor size, you will change the apparent depth of field on prints or on screen, so: no, you can't do that.
Same reason why shooting on the same image area (like Fransisco did) and then crop the image is wrong... It's like having used an APS-c sensor on one of the images (the 50mm lens)... Like if he has used a D800 for the MF lens and a D7000 for the 35mm lens...
-
Try this:
Take any image from your MF DB an print it A1or A2. Now cut a square at the center of the pint 10x10 cm.
Hold the A1 (or A2) print next to the small square you cut before (in other words, look at them from the same distance). Does the complete print has the same DOF than the small square or not?
-
Try this:
Take any image from your MF DB an print it A1or A2. Now cut a square at the center of the pint 10x10 cm.
Hold the A1 (or A2) print next to the small square you cut before (in other words, look at them from the same distance). Does the complete print has the same DOF than the small square or not?
Off course it does Francisco... but that's not my point! My point is that the result you got is the same as shooting a 75mm (FF lens) on an APS-c camera...
EDIT: ...Since later you magnified it (and thus reduced the DOF to the magnification of the respective focal length - i.e. from 50mm to 75mm of a FF lens...)
-
Try this:
Take any image from your MF DB an print it A1or A2. Now cut a square at the center of the pint 10x10 cm.
Hold the A1 (or A2) print next to the small square you cut before (in other words, look at them from the same distance). Does the complete print has the same DOF than the small square or not?
I do not think this is a very relevant comparison. "Depth of field" is perceptual measure that depends on a variety of factors including viewing distance (not to mention the visual acuity of the viewer etc.), and in general people do not view prints of very different sizes from the same distance.
About the best one can do is:
(a) First, give a measure of DOF under somewhat arbitrary "reference" conditions, including
-- the focal length f
-- the aperture ratio N (or one could instead use the effective aperture diameter, A = f/N)
-- the "effective format", meaning the dimensions of the part of image circle at the focal plane that is used to produce the displayed image (so that crops are format changes)
-- the viewing distance
-- some measure of the viewer's visual acuity; maybe a proxy like "circle of confusion diameter no more than 1/3000th of the viewing distance, as I believe is used in many traditional DOF scales).
(b) Then to deal with other situations, describe the changes in DOF caused by changing any of those conditions, in relative terms.
For example:
-- Doubling the focal length at equal aperture ratio or halving the aperture ratio at equal focal length (either of which doubles the effective aperture diameter) approximately halves the DOF.
-- Doubling the viewing distance or halving the displayed image size doubles the DOF.
-- Halving the (linear) size of the image (as with cropping) and still displaying the new image at the same size as the original (so doubling the degree of enlargement) halves the DOF.
-- Aging enough to half one's visual acuity doubles the DOF of all your photos!
-
Off course it does Francisco... but that's not my point! My point is that the result you got is the same as shooting a 75mm (FF lens) on an APS-c camera...
EDIT: ...Since later you magnified it (and thus reduced the DOF to the magnification of the respective focal length - i.e. from 50mm to 75mm of a FF lens...)
I'm using an arbitrary area of the sensor for the 80mm which has a diagonal 1.6x compared to the sensor area used for the 50mm (same aspect ratio), then for the output I enlarged the smaller area (from the 50mm) 1.6x compared to the area used for the 80mm and compared them side to side.
Any object from the scene will look the same in the output.
Yes, you can use a 75mm on FF and a 50 mm on APS-C and will get a valid comparison
-
I'm using an arbitrary area of the sensor for the 80mm which has a diagonal 1.6x compared to the sensor area used for the 50mm (same aspect ratio), then for the output I enlarged the smaller area (from the 50mm) 1.6x compared to the area used for the 80mm and compared them side to side.
Any object from the scene will look the same in the output.
Yes, you can use a 75mm on FF and a 50 mm on APS-C and will get a valid comparison
It's not the same magnification factor Fransisco... you treat the MF lens like being an FF lens...
-
I do not think this is a very relevant comparison. "Depth of field" is perceptual measure that depends on a variety of factors including viewing distance (not to mention the visual acuity of the viewer etc.), and in general people do not view prints of very different sizes from the same distance.
About the best one can do is:
.......-- the "effective format", meaning the dimensions of the part of image circle at the focal plane that is used to produce the displayed image (so that crops are format changes)........
That's right!
-
Let my try again:
Assumption: Scene is at a distance > 1m.
Take an image with a FF DSLR (35mm) with a 50mm lens and another with a MF DB with an 80mm lens.
Then:
- take a crop of 10 x 15 mm out of the FF DSLR image (diagonal 18.02776mm)
- take a crop of 16 x 24 mm out of the MF DB image (diagonal 28.8441 mm)
Make a print of 20 x 30 cm out of both crops which means:
- the 10 x 15 mm from the FF DSLR will be enlarged 20x and the image from the MF DB will be enlarged 12.5X
Look at them side to side. Do they show different field of view? Do objects in the original scene look the same size in the prints or are they different?
-
Let my try again:
Assumption: Scene is at a distance > 1m.
Take an image with a FF DSLR (35mm) with a 50mm lens and another with a MF DB with an 80mm lens.
Then:
- take a crop of 10 x 15 mm out of the FF DSLR image (diagonal 18.02776mm)
- take a crop of 16 x 24 mm out of the MF DB image (diagonal 28.8441 mm)
Make a print of 20 x 30 cm out of both crops which means:
- the 10 x 15 mm from the FF DSLR will be enlarged 20x and the image from the MF DB will be enlarged 12.5X
Look at them side to side. Do they show different field of view? Do objects in the original scene look the same size in the prints or are they different?
You should have used the scale on the diagonal Fransisco... use lenses of the same focal length and blow (instead of cropping the FF lens) the MF lens result by 4x... (because the image area that the lens projects is 4x).... Otherwise, the magnification factor is less than it should be...
EDIT: The magnification factor you got out of your method was only of (about) 2.5x (1.6 ^2)....
-
You should have used the scale on the diagonal Fransisco... use lenses of the same focal length and blow (instead of cropping the FF lens) the MF lens result by 4x... (because the image area that the lens projects is 4x).... Otherwise, the magnification factor is less than it should be...
The scale of the diagonal is the same 1.6X, image area has nothing to do
-
The scale of the diagonal is the same 1.6X, image area has nothing to do
Obviously you posted after (before you see) my editing... That's exactly why I'm saying that you've treated the MF lens like being a FF lens... But if you was using both lenses at their full format (to have the same AOV on the diagonal), the MF lens with your method is already cropped...
The suggestion for the diagonal is only so that accuracy is maximized (since the same AOV for comparison can only be to the diagonal as formats are different)...
EDIT: You are now (with your method) missing the difference of the magnification factor (and have some small difference because of not using the scale on the diagonal) by what the difference of (about) 4x is with (about) 2.5x....
-
Obviously you posted after (before you see) my editing... That's exactly why I'm saying that you've treated the MF lens like being a FF lens... But if you was using both lenses at their full format (to have the same AOV on the diagonal), the MF lens with your method is already cropped...
The suggestion for the diagonal is only so that accuracy is maximized (since the same AOV for comparison can only be to the diagonal as formats are different)...
EDIT: You are now (with your method) missing the difference of the magnification factor (and have some small difference because of not using the scale on the diagonal) by what the difference of (about) 4x is with (about) 2.5x....
A 50mm on 24x36 mm sensor will have the same angle of view than an 80 mm on a 38.4 x 57.6 mm sensor. We have to keep the same aspect ratio otherwise it will be comparing apples to oranges
80mm is 50mm X 1.6
38.4 is 24 X 1.6
57.6 is 36 X 1.6
The diagonal of 24X36 is 43.2662mm
The diagonal of 38.4X57.6 is 69.22658
69.22658 is 43.26662 X 1.6
The area of 24X36 is 864 mm2
The area of 38.4X57.6 is 2211.84 mm2
The difference between 2211.84 and 864 is 2.56 which is the same as 1.6^2
The difference in the diagonals of the experiment I proposed is 1.6
There are no small differences in my calculations. I use as many decimal digits I need to maintain accuracy
Where do you get 4X from?
-
My reaction every time I see a new post in this thread:
(http://2.bp.blogspot.com/-UUBBVvfBUcM/T1t8Zi673mI/AAAAAAAAA0c/uEdZ-PA4IpI/s1600/homer-simpson-bush-gif.gif)
-
A 50mm on 24x36 mm sensor will have the same angle of view than an 80 mm on a 38.4 x 57.6 mm sensor. We have to keep the same aspect ratio otherwise it will be comparing apples to oranges
80mm is 50mm X 1.6
38.4 is 24 X 1.6
57.6 is 36 X 1.6
The diagonal of 24X36 is 43.2662mm
The diagonal of 38.4X57.6 is 69.22658
69.22658 is 43.26662 X 1.6
The area of 24X36 is 864 mm2
The area of 38.4X57.6 is 2211.84 mm2
The difference between 2211.84 and 864 is 2.56 which is the same as 1.6^2
The difference in the diagonals of the experiment I proposed is 1.6
There are no small differences in my calculations. I use as many decimal digits I need to maintain accuracy
Where do you get 4X from?
56x56 is close to four times the area of 24x36.... If you shoot on the diagonal, you don't have to keep the aspect ratio because the diagonal is (supposed to be) a diameter of the image circle no matter what the aspect ratio is.
EDIT: The projected area of the 6x6 is exactly 3.63 larger (not 2.56x) and the diagonal is 1.9x (not 1.6x)... If you insist to stick with the 3:2 ratio (which you don't have to because the angle of the diameter of the image circle doesn't matter) you'll have to choose a 3:2 rectangle with a 1.9x larger diagonal (by using a Pythagoras theorem) which will end up to the same result... What you now do, is treating the MF lens like if it was of a smaller format that projects an image circle of (about) 30% smaller area image circle (and thus different COC size).
-
The projected area of the 6x6 is exactly 3.63 larger (not 2.56x) and the diagonal is 1.9x (not 1.6x)... If you insist to stick with the 3:2 ratio (which you don't have to because the angle of the diameter of the image circle doesn't matter) you'll have to choose a 3:2 rectangle with a 1.9x larger diagonal (by using a Pythagoras theorem) which will end up to the same result... What you now do, is treating the MF lens like if it was of a smaller format that projects an image circle of (about) 30% smaller area image circle (and thus different COC size).
I have no problem in using a different ratio (btw, the diagonal of 56x56 is 79.196, which is 1.83X the diagonal of 24x36 or 43.267), but the essential issue is:
- You are convinced and will not change your mind that the image circle is related to CoC
- I am convinced and will not change my mind that the image circle does not have anything to do with CoC
As a result of this, I better follow Synn advice and move on.
Have a nice day
-
My reaction every time I see a new post in this thread:
(http://2.bp.blogspot.com/-UUBBVvfBUcM/T1t8Zi673mI/AAAAAAAAA0c/uEdZ-PA4IpI/s1600/homer-simpson-bush-gif.gif)
+1
-
Try this:
Take any image from your MF DB an print it A1or A2. Now cut a square at the center of the pint 10x10 cm.
Hold the A1 (or A2) print next to the small square you cut before (in other words, look at them from the same distance). Does the complete print has the same DOF than the small square or not?
The classical formulas for depth of field suppose that smaller prints are viewed at a shorter distance than larger prints. So, for this particular test, you will need to look at the crop at close distance and compare it to the whole print hung further away. You will then find out that they do not have the same DOF.
Probably, I should also disappear in a bush as Homer Simpson in the animation, but I've never liked that series anyway.
-
+1
+100
-
My reaction every time I see a new post in this thread:
(http://2.bp.blogspot.com/-UUBBVvfBUcM/T1t8Zi673mI/AAAAAAAAA0c/uEdZ-PA4IpI/s1600/homer-simpson-bush-gif.gif)
I've seen better animations of front of Depth of Field range (depth of hedge?) ... ;)
Do note that this is an Equipment & Techniques forum, so 'some' factual information is to be expected (amongst the usual distractions). One can only hope for a reasonable Signal to Noise ratio of the contributions, or even contribute to increase that signal level.
Cheers,
Bart
-
I made some comments on another photography site about the new Leica S 007 being a medium format camera (ok, it is). My point was that the old definition of Medium Format being anything larger than 135 sized sensor is not really relevant in the digital age.
During the film era things were simple: with the same or similar emulsions available in different sizes, you were able to increase the IQ simply by using bigger piece of film. With that came the effect of diminishing DOF with the same aperture lenses, etc. The MF look, if you wish. And LF look with large plates.
If we put this "scientifically" the formula which defined the "look" had only two variables: film size (which governed the IQ) and the lens aperture, larger formats having less DOF.
Now we have another variable in the equation: Sensor quality. Sensors can not be judged by surface area only, like film was (in a simplified sense). This makes the "format equation" much more complicated, because we can get the same MF quality with small but high resolution sensor with large aperture lenses. So a 135 size sensor (like Nikon D8xx, Sony & new Canon) with fast lens can give you optically the same picture as slightly larger "mid format" (?) sensor with slightly slower lens, as larger format lenses tend to be. As we know, new 135 sensors beat many MF backs and sensors both in resolution and all in DR. With a slightly faster lens all shallow DOF effects of MF cameras can be duplicated with smaller sensors also.
So are we stuck with the old film era definition of MF? How about defining digital MF as something with more than certain resolution (30 MPix?) and the availability of lenses for the system which give the same minimum DOF as f/1.4 on a 135 sensor, or f/2 on a "MF" sensor (whatever it is)? It should be the end result that matters, not the way it is achieved.
Hello Petrus,
Since digital age, marketing started to be very strong to sell everything under every name possible (if not trademarked). In my world, medium format start at 6x6 cm of film or digital sensor... So, IMHO, digital medium format are not yet proper medium format cameras because it seems to never exceed 53.7 x 40.3mm so 5x4 cm. Yet it is pretty close so lets call it medium format. Every thing under 5x4cm is called crop, it is something new in between, proper to digital age. I would not call it medium format.
For some schools and teachers, pixel count have nothing to do with what we call medium format. What count is perspective of a subject on a certain sensor surface, so roughly a kind of visual rendering. Same goes for Large Format, another rendering. It is vain to try to mimic those specific rendering between formats. It is maybe possible with photoshop (certainly but at a cost: time).
The less the enlargement, the better the picture.
Pixel count is secondary in this story. It just improve what can be done in the picture but at the end the global rendering is given by the format, not the pixel count. It is why I continue to give, even today, all credits to film, all my attention to the format and the format only (especially in digital).
-
I have no problem in using a different ratio (btw, the diagonal of 56x56 is 79.196, which is 1.83X the diagonal of 24x36 or 43.267), but the essential issue is:
- You are convinced and will not change your mind that the image circle is related to CoC
- I am convinced and will not change my mind that the image circle does not have anything to do with CoC
As a result of this, I better follow Synn advice and move on.
Have a nice day
Off course COC is related with the image circle of a lens since the magnification factor reduces with larger image areas... This is no science... it is common sense!
-
Off course COC is related with the image circle of a lens since the magnification factor reduces with larger image areas... This is no science... it is common sense!
The COC is a limit that indicates the maximum allowable size of a blurred point source (before it is observed as unsharp/blurred detail). Since the output magnification determines the size of the COC as observed by the user (from a certain distance), it is a parameter that has virtually nothing to do with the lens, but everything to do with output size/magnification.
This is science and common sense.
Cheers,
Bart
-
Off course COC is related with the image circle of a lens since the magnification factor reduces with larger image areas... This is no science... it is common sense!
So, if I place a lens shade which is too long on a lens the DOF will change. Interesting.
-
So, if I place a lens shade which is too long on a lens the DOF will change. Interesting.
In a sense it will -- if you use DOF in the long-standing sense where it also depends on factors like print size and viewing distance.
Your lens-shade vignetting scenario produces a smaller usable image covering a smaller field of view (a kind of crop), and if the available image is then printed at the same size, so increasing the degree of enlargement and giving larger circles of confusion on the print, it will have less DOF.
A more clearcut case is that un-cropped prints from my Four Thirds telephoto zoom lens at 200mm, f/3.5 have visibly less DOF than would equal size un-cropped prints from the 200mm, f/3.5 settings in 35mm format -- related to the FOV being far smaller with the Four Thirds image.
-
In a sense it will -- if you use DOF in the long-standing sense where it also depends on factors like print size and viewing distance.
Your lens-shade vignetting scenario produces a smaller usable image covering a smaller field of view (a kind of crop), and if the available image is then printed at the same size, so increasing the degree of enlargement and giving larger circles of confusion on the print, it will have less DOF.
I thought I would just go "Damn! Vignetting from the wrong lens shade!" But DOF changes...
-
I thought I would just go "Damn! Vignetting from the wrong lens shade!" But DOF changes...
;D yea that's start to sound weird isn't it ? :D I think you got your answer pages ago :p The late discussion is just weird and out of this world ...
-
I thought I would just go "Damn! Vignetting from the wrong lens shade!" But DOF changes...
Indeed, you've added a smaller aperture, and you'd need a longer focal length to achieve the full Field of View that the lens normally produces. Then as a result the output magnification also needs to change in order to maintain the same DOF in equal sized output. All in line with the calculations earlier.
Cheers,
Bart