MF vs 1Ds3

[jjj]

There are many reasons for choosing a particular format, including resolution, shape of the image, depth of focus, whether or not there will be electricity available on the shoot, even the fact that subjects react differently to different types of cameras (I know I have mentioned this before).  So the vehement argument for one format over another is pointless.
[a href=\"index.php?act=findpost&pid=192478\"][{POST_SNAPBACK}][/a]

Also there is the fact that different cameras affect how you shoot. So you may choose a camera, simply for that reason alone.

[Ray]

What surprises me about threads like this where sincere questions are asked about format differences, is the apparent inability of professionals, who claim to use these different formats on a regular basis, to actually identify the causes of these perceived differences in image quality.

Nobody has yet answered my question as to why the best cameras for ultimate image quality, which are more often used by busy and demanding professionals, do not include an AA filter, whilst smaller format cameras with higher pixel densities (and therefore with presumably a lesser need for an AA filter) always have an AA filter.

The only certain advantage I see for the larger format is greater dynamic range as a result of larger photosites, (which should translate to smoother shadows and smother lower midtones), and the lack of an AA filter which should contribute to a slightly crisper result and perhaps also a smoother result due to the need for less sharpening.

For all I know, this mystical 3-D effect attributed to the additional size of the DB sensor, might be wholly due to the lack of an AA filter and nothing to do with sensor size.

Whilst there's a general trend for smaller formats to be worse for achieving a shallow DoF, this does not appear to be the case when comparing FF 35mm with cropped MF format. There are lots of 35mm lenses which are at least one stop wider/faster than the equivalent MF focal length.

My own comparison between the 40D with 50/1.4 at full aperture, and the 5D with TSE 90/2.8 at full aperture, resulted in a preference for the results from the larger format 5D. However, when I later repeated the test in different lighting (brighter lighting), I wasn't so sure. The results from both cameras seemed about equal. I re-examined the first results and concluded there was a very slight difference in focussing which, by chance, favoured the 5D shot in its general appeal. There was also noticeably greater contrast in the 5D image which gave the impression of greater accutance. This impression was not there in the second test in brighter lighting. I would deduce that the TSE 90/2.8 is better at bringing out lower contrast detail. In brighter lighting, the same detail on the same subject was of higher contrast and the advantage of the better lens was not so obvious.

This effect, by the way, can be very obvious when using a teleconverter, which inevitably reduces the quality of the lens. Photograph a low contrast subject, and the teleconverter often seems to offer little or no advantage. But photograph a brightly lit subject with high contrast detail, and that extra detail can be captured.

The other issue with my comparison was the 2 stop difference I used, partly because the focal lengths were not precisely matched and partly because the format size difference between the 40D and 5D is slightly greater than the difference between a DB and FF 35mm. Most lenses are noticeably better when stopped down one stop, and I'd expect the Canon 50/1.4 to produce better contrast at F2 than at F1.4.

At the risk of causing Jack Flesher to vomit, I might do the test again using the 50/1.4 at f2 and the TSE 90/2.8 at F2.8, although I would prefer to use a circumcised 5D which I think would provide a more accurate simulation of the benefits of the DB.

[Jack Flesher]

Perfectly stated John...  Where is that "bowing down" smiley

[Ray]

I think the major issue here is that you only seem interested in testing and a scientific approach to 'why this' or 'why that.'

John,
That's simply not true. Only a small fraction of a percentage of my photos were taken for testing purposes. Since acquiring digital cameras, I've taken over 100,000 photos, but only a few hundred to test for lens sharpness, DoF, noise and dynamic range issues etc.

None of my testing is scientific. I don't use IMATEST or DCRAW and all my testing is directed towards getting useful information that's relevant in the field.

Most working photographers want to create great photos, meaning the content, style, creativity, and having that captured with great gear is certainly a plus.

Same for every photography enthusiast I would think, whether amateur or professional.

I am actually shooting a lot of jobs with my Aptus 22, not because the client needs big files but because the default color is more accurate, the tones are lovely, skin looks better without tons of post work and I just prefer the pace MF digital forces me into instead of slamming the shutter down on a Canon.

I understand that. Some people might prefer the Nikon D3 because it produces great in-camera jpegs. Other might prefer a particular RAW converter because it produces more pleasing results with less work. Apparently some photographers believe the 5D produces better skin tones than the more expensive 1Ds2, yet they are both the same format.

These are different issues to the one of fundamental quality related to format size.

You may enjoy all the testing and methodology of thinking about finding answers to many not very important questions in the grand scheme of creating photographic images, but most of the rest of us simply don't care. We shoot because the gear is only a means to an end not the end.

What makes you think I'm a gear head? In addition to some old film camera gear, I've bought only 4 inexpensive DSLRs in my life and own about 10 lenses that fit those DSLRs. The amount of gear I own and use is trivial compared with most professional photographers.

You might have noticed that I recently got beaten up in a thread on LL because I tried to argue that the 'camera doesn't matter'.

[Anders_HK]

Hi

I have not posted for about a month, due relocation to Hong Kong and frank it is tiring posting much! I was also on travels in February and March, including photographing in India.

As an AMATEUR I have since about a year choosen medium format and given up on DSLRs. Why?
1. I wanted higher image quality than DSLRs (my experience D200, D50, ZD, Aptus 65). This does not only refer to pixels, but COLORS, CONTRAST, SHARPNESS.
2. I wanted to advance in FORMATS.
3. I do not like all high tech automation of DSLRs. Before digital I shot F100 and Fuji Velvia which was a slower than DSLRs to work with and which gave me incredibly beautiful image quality, although small.

1Ds Mk3 has equal numbers of pixels to Leaf Aptus 22 and ZD, but... there is difference in the quality of what you get in those pixels.

I went with the Leaf Aptus 65 (prior to it had ZD, which has problems). The Aptus 65 is actually lower priced than the 22, but a 44x33mm sensor. It is also newer technology sensor. Yes, it cost more than 1Ds Mk3, but I do not have a car...

Compared DSLRs I experience following photographic advantages with the Aptus 65;
1.  The 6x7cm screen with its superb histogram and gray card capabilities is excellent tool.
2. Images that I jusdge as good on the 6x7 screen consistent look better (often very good!) when I open up at default settings in CS3 Camera RAW.
3. Colors, contrast and sharpness of files are superb compared to files from ANY other DSLR files (inlcluding of course D200 and other I have ever downloaded, also RAWs, from 1Ds Mk3 etc).
4. The fact that files look good when I open up with defaults in CS3 not only saves me time, but it is a great reward for my photography making me feel as I am back to joy of photography again.
5. With the Aptus 65 I feel confident to use also up to ISO 800. The little trick is that for an around a 30MP sensor the noise consists of small sized noise artifacts compared to an around 20MP camera, and by printing at reduced size the noise is not significant to disturb, actually film like.

I confess I did not read all posts in this thread in detail... Why get caught up in comparison DSLR vs. MFDB???

First and foremost DSLRs and MFDBs are different format, but also with MFDBs having some advantages from an amateurs eyes as listed by me in above. However, MFDBs are expensive and may not be the suited tools for all. It depends on our photoraphy and budget/affordability/life priorities. Clear though, apart from in very low light and sports, the state of art in MFDBs very clearly exceed DSLRs in image quality. Yes, more 3D both due to lack of IR and because of the LARGER FORMAT. How to quantify quality of MFDBs?? Not sure, how to quantify 35mm film vs. medium format film? Except... research lots of files and try out before you set your mind.

Would I be interested in getting a DSLR? In future maybe, **IF** they finally make one  with simple interface and less automation, and with which I will can get really good colors, contrast, sharpness etc with compared to current Nikon (Sony) and Canon, e.g. a Fuji S8??... and one that not weigh too much. The IR filter is only one part in equation, but... frank at around 20MP what is point of an low pass filter???

MFDB helps me focus more on photography because it is slower and makes me plan better. Less automations also helps in that. The results are higher sucess ratio of shots and less time in computer after shots to correct them   .

Ah, one more important advantage with medium format: THIS FORUM. Help and replies direct from professionals with much knowledge, not only on technical, but also extremely skilled at photography. Much thanks to all of you on my prior posts in this forum. The Aptus 65 works for me, the P30 is very excellent competitor, and I feel the quality I get in files from my 65 will keep me content for many many years   .

Regards
Anders

[csp]

I choose what I chose (MF) because it helps me to achieve my visions in a very natural way to me. My work is just connected to the way mf works and look. That's all.

[a href=\"index.php?act=findpost&pid=192522\"][{POST_SNAPBACK}][/a]

sorry i cant see a vision or style in your images all i see are snapshots nothing connects them  whether technical nor artistic but maybe this comes because we don't share the same faith.

[samuel_js]

sorry i cant see a vision or style in your images all i see are snapshots nothing connects them  whether technical nor artistic but maybe this comes because we don't share the same faith.
[a href=\"index.php?act=findpost&pid=192618\"][{POST_SNAPBACK}][/a]

You don't have to be sorry, for me...


P.S. You are very right. Like some said, a photograph is a "snapshot" of a moment in the history of humanity. I'm very glad you see my pictures as that.  

[NBP]

So, any chance of us seeing any of your work csp?  - Just so we can all see where we are going wrong, like....















...no, thought not.

[AndrewDyer]

Well.... it looks like we have a 'flamer' here with the initials csp...
Maybe if we ignore it... like a turd floating down a river... it will go away.
Here's hoping.
A

[thsinar]

That's ways too much testing shots, Ray!

The real tests take place when you actually shoot images, for your pleasure, for clients, for other purposes, not by testing for testing.

 

Thierry

I've taken over 100,000 photos, but only a few hundred to test etc...
[a href=\"index.php?act=findpost&pid=192596\"][{POST_SNAPBACK}][/a]

[Quentin]

What surprises me about threads like this where sincere questions are asked about format differences, is the apparent inability of professionals, who claim to use these different formats on a regular basis, to actually identify the causes of these perceived differences in image quality.

Nobody has yet answered my question as to why the best cameras for ultimate image quality, which are more often used by busy and demanding professionals, do not include an AA filter, whilst smaller format cameras with higher pixel densities (and therefore with presumably a lesser need for an AA filter) always have an AA filter.

Not quite always.  Partly for reasons of cost, the Kodak SLR/n and SLR/c had no AA filter, and when operated at low ISO within their comfort zone, this did produce visibly sharper files but with quite a lot of color aliasing that could take a while to remove.  I still use my old Kodak occasionally and it has somewhat sold me on non-AA filter cameras, as I upgraded to a Mamiya ZD.

Smaller format cameras are more likely to be purchased by less specialized users (ie joe public) for whom a weak AA filter is desirable.

Quentin

[Ray]

That's ways too much testing shots, Ray!

The real tests take place when you actually shoot images, for your pleasure, for clients, for other purposes, not by testing for testing.

 

Thierry
[a href=\"index.php?act=findpost&pid=192638\"][{POST_SNAPBACK}][/a]

Thierry,
I know it is and I'm getting pissed off. I've got 10 Canon lenses. I could have had 11, but the 400/5.6 prime proved not to be even as sharp as my 100-400 IS zoom which has a reputation for being a bit soft at 400mm. Before getting a refund, it was sent in for calibration. After testing it again, I returned it for a refund.

My first copy of the EF-S 10-22mm, bought for my 20D, would not autofocus accurately at close distances. I returned it.

My Canon 50/1.4 proved not to be as sharp as the much cheaper Canon 50/1.8, but I learned about this only after thorough testing, after the warranty had expired, for the purpose of determining if there was any noticeable difference in sharpness at F5.6 and F8. I tested 3 lenses, all at 50mm. I figured I needed to know at which aperture my lenses are sharpest; a reasonable concern, wouldn't you say?

I also tested that 50/1.4 shortly after I bought it, at full aperture, at infinity, using autofocus. I now realise that I wrongly attributed the softness to a low MTF response. After recently sending in the lens for calibration, it's now reasonably sharp at F1.4 autofocussed at infinity, but way out when autofocussed at close distances. It looks as though this lens might not be capable of accurate autofocussing at both close distances and infinity.

I also have an autofocussing problem with my latest lens acquisition, the EF-S 17-55/2.8, but I'm not sure if it can be corrected.

There's no doubt in my mind, if I had any MFDB system that was capable of accurate focussing, perhaps in tethered mode, and I compared it with my 35mm gear at wide apertures in autofocus mode, I'd get a very strong impression that the MFDB was vastly superior in the 'shallow DoF' department.

I might even kid myself the superior results were due to the sensor size rather than more accurate focussing.

[Ray]

Smaller format cameras are more likely to be purchased by less specialized users (ie joe public) for whom a weak AA filter is desirable.
[a href=\"index.php?act=findpost&pid=192639\"][{POST_SNAPBACK}][/a]

Quentin,
The 1Ds3 is not a camera likely to be bought by Joe Public, yet it seems to have a rather strong AA filter.

I don't recall hearing complaints from DB users along the lines that they were dissatisfied with their back because of time wasted trying to correct for aliasing artifacts, and had therefore sold their MF equipment and bought 35mm DSLRs.

But maybe there are a few dissatisfied DB users who are keeping quiet about this.

[Sean H]

Samuel, I love the first landscape shot!
[a href=\"index.php?act=findpost&pid=192528\"][{POST_SNAPBACK}][/a]

Isn't it beautiful? We've seen some really good photography on this thread; also check out the 'recent works' -- Graham has some stunning shots.

[Paul2660]

Ray,  a bit of thoughts on the AA filter usage

You are totally correct that it seems that most of not all 35mm DSLR's have a built in AA filter (low pass).  It's built into the sensor array at the time of manufacturing.  Some  of the Digital backs offer you a solution where you can either add an AA filter (such as Mamiya's ZD) or offer a software solution to offset the effect of not having an AA filter (such as Phase One's Morie plug-in).  

Canon, Nikon, Olympus, Sony etc. all are using some form of AA filter on their chips.  This IMO is because they would rather do this than have thousands' of users complain about issues like Morie and color aliasing (Christmas tree lights effect) on many if not all of their images.  This would cause most of these users to then have to use post processing software and that's not what they intended the use of the these cameras for. I believe it's also an issue of simplification of manufacturing.  You only need one chip for all the make of one camera, not a set of chips, one with AA on without.

Canon on the 1DS and above (can't speak to Nikon as I haven't used them for many years but will assume they are the same), all use a AA filter also.  The effect on the overall image is the same as on the other types of cameras.  The AA filter slightly blurs the image to reduce morie and color aliasing.  These issues have been the main two concerns with Digital for years.

Morie, tends to show in all fabrics, and whenever a repeatable pattern is being shot.
Color Aliasing (Christmas Tree lights), not as well known about but can be just as bad. This is an effect mainly seen in finer detail parts of images.  In outdoor work, very often shows up in tree limbs, areas of dense leaves, grass.  It's my understanding that this is due to the design of the bayer pattern chips.  This is why there was so much (and I hope still is) design around the Foveon chip.

Canon has a predetermined market that they are intending to sell the 1ds MKIII to. IMO after shooting the MKIII, they put a very intense AA filter on the chip.  There has already been alot written on this.  Some believe that the details can be brought back other don't.  I personally would prefer to have them at time of capture and not have to work with software tools to bring the details back.  Many also wrote early on that it would have been great if Canon had made two versions of this camera, one with a heavy AA filter for fashion/studio work and one with a lighter version of the filter for outdoor landscape work.  

I feel that Until Foveon or someone like them comes out with a chip that reads the R, G, B on different chips, then color aliasing will be an issue that all companies have to deal with in some method.   Morie is a different issue and I assume it will take either new chip technology or camera processing ability to resolve at time of capture.

The other solution is to remove the low pass filter from the chip by a 3rd party like MaxMax and the hotrod conversion.  They are currently doing this on the 5D and you can go to their website and see the before and after images.  

This will hopefully not been seen as a which is better MFDB or DSLR, just a bit more info to help answer your question as to why you see AA filters on certain types of cameras.

Paul Caldwell

[Ray]

Canon, Nikon, Olympus, Sony etc. all are using some form of AA filter on their chips.  This IMO is because they would rather do this than have thousands' of users complain about issues like Morie and color aliasing (Christmas tree lights effect) on many if not all of their images.  This would cause most of these users to then have to use post processing software and that's not what they intended the use of the these cameras for. I believe it's also an issue of simplification of manufacturing.  You only need one chip for all the make of one camera, not a set of chips, one with AA on without.
[a href=\"index.php?act=findpost&pid=192672\"][{POST_SNAPBACK}][/a]

Paul,
I understand this is the usual explanation, but it still doesn't sound quite right to me. I presume the new 12mp 450D also has an AA filter. Whether it's strong or weak, I don't know, but this degree of pixel density translates to 31mp on a FF 35mm sensor or 62mp on a 36x48mm DB. I believe most DBs don't even have microlenses, which also act as a mild AA filter on CMOS sensors which need microlenses.

The 5D was considered by some to be very close in resolution to the 1Ds2 because it had a weaker AA filter. The 1Ds3 with almost double the pixel count of the 5D needs, apparently, an even stronger AA filter. It doesn't quite make sense.

There is perhaps some validity to the argument that non-professionals cannot be expected to learn how to deal with moire issues. However, the argument that professionals would be more concerned about the effects of moire and more annoyed at the time and effort needed to deal with it, seems at least equally valid if not more valid.

I find it difficult to understand that anyone would be willing to spend $30,000 on a DB if moire issues were a pain in the arse.

[Frank Doorhof]

I think people that buy a MF system are not scared by moire and can solve it.

I have to be 100% honest and say that Moire can be seen on app 10 shots in each session, however it NEVER ruined a shot for me and was always easily fixed.

However on a DSLR people will start complaining online and hurting the system because they simply don't know what moire is and how it's retouched.

See it as JPEG vs RAW.
There are more and more people claiming that JPEG is as good as RAW, and I think we all know that this is not true.
You can talk what you want but those people will say they see no difference what so ever.

So take away their AA filter and the forums will be full about the strange colordots, strange rings etc. that ruin their cat and dog shots.

I see the MF system as tool for professional photographers, it can give you that extra mile but you sure have to know how to use it.
A DSLR is VERY GOOD but it's full of foolproof options that will give you a good picture in 99% of the cases without photoshop. (I have seen moire on some 5D shots actually)

[Dustbak]

Sometimes moire is a real pain in the b*tt but for me it mostly occurs in product shots of clothing (fabric) which is easily solved by spending a little more on a multishot back

When MS is no option and you have used every trick in the book to prevent moire during the shoot you have to handle it in PP (mostly fairly easy sometimes a bit harder) but indeed that does eat up my valuable time which I would have preferred to spend in other ways.

I guess most other DB users that run into moire feel pretty much the same way?

[Dansk]

On the money Paul

Color Aliasing (Christmas Tree lights), not as well known about but can be just as bad. This is an effect mainly seen in finer detail parts of images.

This effect is worse on Canon than Phase in my experience especially when back lit

[thsinar]

FYI: there are 2 types of Moiré occurring:

- Luminance Moiré, also called "Pattern" Moiré

and

- Chrominance Moiré, also called "Colour" Moiré

Luminance Moiré is the one which we are confronted with when shooting e.g. a fabric, and happens when one shoots a regular structure on top of another regular structure (CDD is a regular structure). Luminance Moiré can basically happen also in images taken with the multi-shot technology.

Chrominance Moiré is typical of single-shot technology, and does not happen in a multi-shot image, unless the registration is not perfect. Chrominance Moiré is therefore the typical Moiré seen in images taken in 1-shot mode. It happens often (not always, but often enough), that when the colour Moiré is removed, appears then a Luminance Moiré which can very unlikely be corrected or taken away: the "sharpness" information being "stored" in the luminance information, one cannot simply "smear" to get rid of the Moiré, unless one take in account a loss of sharpness.

Thierry