Luminous Landscape Forum
Equipment & Techniques => Cameras, Lenses and Shooting gear => Topic started by: carl dw on February 10, 2016, 12:35:25 pm
-
I tried to add this to the previous 5ds or 5dsr thread but it's locked.
A couple of weeks ago I borrowed both bodies from my local camera shop to test side by side before making a decision. I was set up to do some copy work, a nice controlled environment where I could take my time and see what the difference in sharpness was for myself.
Shot with a Zeiss ZE Macro, the additional clarity of the 5dsr is quite clear in the attached copies. It gives the impression of a little extra bite of acutance. I resisted the temptation to sharpen too much so they do look like they could both stomach a tad more - they both had equal amounts.
I also tried to make some moire with both cameras but was unsuccessful with the conditions/subjects I often shoot.
I went for the 5dsr without hesitation, if you're going to have to deal with 50-70Mb RAWs and 144Mb TIFs you might as well have the best possible image from the onset. Full fat for me please waiter.
I previously spent years shooting transparency so I didn't waste time deliberating the over discussed limitations of Canon sensor dynamic range; any camera that shoots RAW has more dynamic range than I actually need to produce commercial images.
Hope this maybe helps someone else who's on the fence.
Carl
-
Bear in mind that a part (difficult to quantify but real) of the extra sharpness perceived when the AA filter is suppressed, is produced by a high frequency interference caused by the absence of proper AA filtering. In other words, a part of the extra sharpness is a form of noise, not real detail. This will not only happen when periodic patterns close to the Nyquist limit appear on the scene, but also everytime fine detail is captured, like in a landscape (tree leaves, rock or sand textures, grass,...).
The image on the right looks sharper:
(http://www.guillermoluijk.com/misc/nitidezruido.jpg)
I just added noise to the image on the left.
Regards
-
Bear in mind that a part (difficult to quantify but real) of the extra sharpness perceived when the AA filter is suppressed, is produced by a high frequency interference caused by the absence of proper AA filtering. In other words, a part of the extra sharpness is a form of noise, not real detail. This will not only happen when periodic patterns close to the Nyquist limit appear on the scene, but also everytime fine detail is captured, like in a landscape (tree leaves, rock or sand textures, grass,...).
The image on the right looks sharper:
(http://www.guillermoluijk.com/misc/nitidezruido.jpg)
I just added noise to the image on the left.
Regards
Hi Guillermo,
Thank you for your reply. I've read many of your threads and replies with great interest for quite a while and I have the upmost respect for your technical knowledge.
I have to confess that the science behind digital image making (like women) is a mystery to me. I take photos and people are kind enough to give me money for my efforts, it's the way it's been for the last 25 years in my life. I leave the development of hardware to those clever chaps predominantly in Japan who never cease to amaze me with their ability to deliver things I never knew I needed... as well as their comical marketing strategies that seem to believe I'm of no relevance.
Looking carefully at your example I'm seeing an image on the right that looks like the image on the left but with less detail because of the noise. That seems quite different from what I'm seeing in my own examples with and without the cancelling filter. I'm not sure how the addition of noise can add detail to an image, maybe perceived sharpness, but not detail.
On reflection maybe I should of used the term "additional detail" as opposed to sharper, is that what you were meaning? - my apologies for the ambiguity if that is so, I was simply trying to illustrate the difference in a controlled but practical situation.
Ignoring the clever science if I may for a second, my migration through digital imaging started with an Imacon 3020 then moved to a 10D, 1Ds2, 1Ds3, 5D3 with 1DX now complimented by the 5DsR.... the 5DsR is the first camera to re-excite my enthusiasm for stills since my 1Ds2. Stuff looks great with a 5DsR, and that's all that matters to me and the people who pay me at the end of the day.
Kindest regards,
Carl
-
Shot with a Zeiss ZE Macro, the additional clarity of the 5dsr is quite clear in the attached copies. It gives the impression of a little extra bite of acutance. I resisted the temptation to sharpen too much so they do look like they could both stomach a tad more - they both had equal amounts.
I also tried to make some moire with both cameras but was unsuccessful with the conditions/subjects I often shoot.
I went for the 5dsr without hesitation, if you're going to have to deal with 50-70Mb RAWs and 144Mb TIFs you might as well have the best possible image from the onset. Full fat for me please waiter.
I previously spent years shooting transparency so I didn't waste time deliberating the over discussed limitations of Canon sensor dynamic range; any camera that shoots RAW has more dynamic range than I actually need to produce commercial images.
Yes, the comparative advantages of AA filter or not is fairly well documented.
Your last two points are trade offs. That is, the deficiency of the Canon sensor, which extends beyond just the DR disadvantage to the Sony sensor, was accepted in trade for more pixels. In some situations that is a good trade and other not. Both the pixel advantage and DR and Noise disadvantages are not large, but reasonable. However, it really depends on what type of images you make, under what conditions, and what output size and media are targeted as to whether the trade is good (all else being equal).
Given the list of Canon cameras you rolled off though, I more inclined to believe your choice was based on the inability to overcome inertia than any actual comparison between available alternatives. Don't get me wrong, we all have that same consideration when selecting our NEXT camera body as opposed to our FIRST camera body. But, it would be less disingenuous to have just said, I wanted the highest resolution Canon camera I could buy because it meets my needs.
-
An investment in lenses is a serious consideration when contemplating a switch. O for the days of M42 standard mount, used by many brands of camera.... ;) (not really, screw mounts were a pain to change if you were in a real hurry - plus I doubt that screw mounts could work for electrical connections).
-
An investment in lenses is a serious consideration when contemplating a switch. O for the days of M42 standard mount, used by many brands of camera.... ;) (not really, screw mounts were a pain to change if you were in a real hurry - plus I doubt that screw mounts could work for electrical connections).
I agree. And not just lenses, but accessories and the invested time climbing the learning curve. I know I should be using Lightroom, but have some much time invested in Photoshop that I just haven't been willing to switch. I continued to shoot Nikon even while advising people to buy Canon for a long stretch before Nikon figured it out. My point was that we should all be clear and factual about our decisions and motivations, not only to others whom we may be influencing, but to ourselves.
-
Yes, the comparative advantages of AA filter or not is fairly well documented.
Your last two points are trade offs. That is, the deficiency of the Canon sensor, which extends beyond just the DR disadvantage to the Sony sensor, was accepted in trade for more pixels. In some situations that is a good trade and other not. Both the pixel advantage and DR and Noise disadvantages are not large, but reasonable. However, it really depends on what type of images you make, under what conditions, and what output size and media are targeted as to whether the trade is good (all else being equal).
Given the list of Canon cameras you rolled off though, I more inclined to believe your choice was based on the inability to overcome inertia than any actual comparison between available alternatives. Don't get me wrong, we all have that same consideration when selecting our NEXT camera body as opposed to our FIRST camera body. But, it would be less disingenuous to have just said, I wanted the highest resolution Canon camera I could buy because it meets my needs.
A short while ago someone on this forum asked for thoughts and advice regarding the choice between a 5ds and a 5dsr; in starting this thread I’d simply intended to illustrate my own personal experience of these two almost identical Canon cameras as it might help someone else since the previous thread had been locked.
I’m quite bemused and disappointed by you deciding to morph it into another “lets slag off Canon” soap box… as well as rudely questioning my own ability to make considered decisions about my choice of camera equipment.
To use your terminology, let me be "clear and factual" about my "decisions and motivations”…. as you later suggested.
I’ve attached another image for you to look at. It is an untouched (apart from the numberplate) “snap” I took with my 1DX last year on returning to my car after photographing a friend throwing an awesome racing motorcycle into some corners on a road that passes my home as it makes it’s way across the moors of Derbyshire, UK. The sky, the heather, the reflections in the car all looked too pretty to be missed.
I haven’t uploaded this image to demonstrate the dynamic range, sharpness or indeed any aspect of the 1DX whatsoever. No, I’ve uploaded it because I’d recently bought the car for £39K from my local BMW garage with the excess profit I made with my “DR disadvantaged" Canon camera choices and my “inability to overcome inertia”. My account advised my to spend some cash so I bought a coupé, cute isn’t it? I could of bought a PhaseOne back but that would have been an unnecessary millstone and nothing like as much fun as whooshing about the country lanes in a 4 Series.
I’m lucky (or is it astute?) enough to make my entire living from taking photographs with equipment I’ve carefully chosen; I don’t care if it’s a box brownie, a Hasselblad or a pin-hole camera if it gets the job done to the required standard, puts food on my table and gives me the life I want.
To go back to your obsession, todays increasing dynamic range is a godsend to weekend warriors, but it doesn’t make up for skill and craft when a jobs to be done properly. I have no interest in comparing spec sheets with Sony, Nikon or otherwise, why should I? - it’s boring and life is too short… take photographs instead, chill out and enjoy life my friend.
Is that "less disingenuous" enough for you?
Carl
-
A short while ago someone on this forum asked for thoughts and advice regarding the choice between a 5ds and a 5dsr; in starting this thread I’d simply intended to illustrate my own personal experience of these two almost identical Canon cameras as it might help someone else since the previous thread had been locked.
I’m quite bemused and disappointed by you deciding to morph it into another “lets slag off Canon” soap box… as well as rudely questioning my own ability to make considered decisions about my choice of camera equipment.
To use your terminology, let me be "clear and factual" about my "decisions and motivations”…. as you later suggested.
I’ve attached another image for you to look at. It is an untouched (apart from the numberplate) “snap” I took with my 1DX last year on returning to my car after photographing a friend throwing an awesome racing motorcycle into some corners on a road that passes my home as it makes it’s way across the moors of Derbyshire, UK. The sky, the heather, the reflections in the car all looked too pretty to be missed.
I haven’t uploaded this image to demonstrate the dynamic range, sharpness or indeed any aspect of the 1DX whatsoever. No, I’ve uploaded it because I’d recently bought the car for £39K from my local BMW garage with the excess profit I made with my “DR disadvantaged" Canon camera choices and my “inability to overcome inertia”. My account advised my to spend some cash so I bought a coupé, cute isn’t it? I could of bought a PhaseOne back but that would have been an unnecessary millstone and nothing like as much fun as whooshing about the country lanes in a 4 Series.
I’m lucky (or is it astute?) enough to make my entire living from taking photographs with equipment I’ve carefully chosen; I don’t care if it’s a box brownie, a Hasselblad or a pin-hole camera if it gets the job done to the required standard, puts food on my table and gives me the life I want.
To go back to your obsession, todays increasing dynamic range is a godsend to weekend warriors, but it doesn’t make up for skill and craft when a jobs to be done properly. I have no interest in comparing spec sheets with Sony, Nikon or otherwise, why should I? - it’s boring and life is too short… take photographs instead, chill out and enjoy life my friend.
Is that "less disingenuous" enough for you?
Carl
Great car! But why is the steering wheel on the wrong side. 😃😃. From one guy who chose the new Canon 5 series to another. I like my bimmers a bit older...e39 style.
-
Great car! But why is the steering wheel on the wrong side. 😃😃. From one guy who chose the new Canon 5 series to another.
Hey Craig, cheers.. it's almost as much fun as the new camera! - I had the steering wheel moved over so the wife can drive it.
What do you think to your “DR disadvantaged" investment? - you've been a user longer than me ;)
Carl
-
Great car! But why is the steering wheel on the wrong side. 😃😃. From one guy who chose the new Canon 5 series to another. I like my bimmers a bit older...e39 style.
That's why you can afford a plane to go with it!
-
Hey Craig, cheers.. it's almost as much fun as the new camera! - I had the steering wheel moved over so the wife can drive it.
What do you think to your “DR disadvantaged" investment? - you've been a user longer than me ;)
Carl
I'm quite happy with it, and I have the "s". I don't find extended DR to be of much value for my work anyway. It's amazing what proper lighting can do, and after decades of shooting reversal film I'm pretty much used to limited DR. Plus I lacked the inertia to change systems after being with Canon since the 1Ds.
-
I’m quite bemused and disappointed by you deciding to morph it into another “lets slag off Canon” soap box… as well as rudely questioning my own ability to make considered decisions about my choice of camera equipment.
I was not "slagging off" Canon. I was merely pointing out that the rationalization for your choice was "it was the highest resolution Canon camera I could buy." There is nothing wrong with that. I did the same thing when I bought the Nikon D810. It was the highest resolution Nikon available. The fact that it was the highest resolution DSLR at that time and had the best performing sensor system at the time were bonuses.
The sensor subsystem is not the only characteristic to consider when purchasing a camera. But it is one characteristic that should be considered. Rationalizing that it doesn't matter is foolish and does a great disservice to those looking for information. We can discuss under what conditions various performance characteristics are advantage and how prevalent one might encounter them. Just like a car, horsepower, lateral stability and braking don't matter, except when they do!
-
On reflection maybe I should of used the term "additional detail" as opposed to sharper, is that what you were meaning? - my apologies for the ambiguity if that is so, I was simply trying to illustrate the difference in a controlled but practical situation.
Yes and no.
If the AA filter on the Canon 5DS were ideal, i.e. perfect in real world language, the image captured with the 5DS R would never have any more detail than the one from the 5DS since a perfect AA filter would never eliminate or attenuate any valid information. In addition to that, the perfect AA filter would get rid of all of the interference caused by fine detail present in the scene. In such a case the image from the 5DS and perfect AA filter would always be better than the one from the 5DS R (same detail, none of the interference).
But real AA filters are not perfect, so in the real world the 5DS AA filter will not eliminate all of the interference (that is why the 5DS CAN still produce aliasing/moire), and will probably cut some of the fine detail or at least will attenuate it making the valid information look less sharp. This will require extra sharpening that will also enhance the presence of noise in the 5DS.
An exercise to find out how close to a perfect AA filter the 5DS is, is applying an unsharp mask to try to reach the same level of sharpness of the 5DS R, and then judge how much detail is present. The better the AA filter is, the more indistinguisable the images will be:
(http://www.guillermoluijk.com/misc/A.jpg)
(http://www.guillermoluijk.com/misc/B.jpg)
I used a coin to randomize which is A and which is B, the 5DS R or the 5DS. Which one has more detail? is the difference big?.
Regards
-
Apropos Guillermo's posts, I found this article by Tim Parkin interesting as it explores (among other things) the positive impact of sharpening the 5Ds files compared with those of the un-sharpened 5DsR files.
https://www.onlandscape.co.uk/2015/07/testing-the-canon-5dsr/
As an "early adopter" I chose the 5Ds over the 5DsR to avoid the risk of moire: hard to recover from in my line of work. :)
-
Prefer "B", looks more like a real artwork.
-
This is an interesting thread, and I thank the contributors.
I started the 5ds v 5dsr thread, and I locked it because it degenerated into a Nikon v Canon spat, which bored me. There were very few responses to my question (though a lot more to what was supposed to be my concluding remarks, having bought the camera), and as it turned out, money made the choice pretty easy: £200 might be worth paying for the small difference (and it is small, in reality, I think) but £500 certainly wasn't. IMHO, of course.
I suspect both cameras out-resolve my abilities.
Jeremy
-
No disrespect to any test...but the ones I have seen done well, vs ones that leave me scratching my head....
Any time you do a test for sharpness, I highly recommend using a very high contrast yet complex items that are machined, as subject matter.
Natural items, or shapes drawn or painted make poor subjects and really makes it hard to see any true difference.
Further more, the lack of AA filter give the image a more 3D look in my observations. At one time I used a Canon 1Ds vs a Kodak SLRc camera with lecia R glass, and the Kodak without a filter was much sharper. I know that the 1Ds may have had this or that, but the Kodak file was much comparable to my PhaseOne file. This may have little value to those that are looking for something more properly and scientifically done, but this was many years ago, and it was when I became strongly against a AA filter forced on the sensor. Much of those talks are here on Lula.
But the 5DSR, says the filter is there but not activated or something...something funny in their verbage. It looks like they kept it, but somehow it is canceled out at some point??
-
No disrespect to any test...but the ones I have seen done well, vs ones that leave me scratching my head....
Any time you do a test for sharpness, I highly recommend using a very high contrast yet complex items that are machined, as subject matter.
Natural items, or shapes drawn or painted make poor subjects and really makes it hard to see any true difference.
Further more, the lack of AA filter give the image a more 3D look in my observations. At one time I used a Canon 1Ds vs a Kodak SLRc camera with lecia R glass, and the Kodak without a filter was much sharper. I know that the 1Ds may have had this or that, but the Kodak file was much comparable to my PhaseOne file. This may have little value to those that are looking for something more properly and scientifically done, but this was many years ago, and it was when I became strongly against a AA filter forced on the sensor. Much of those talks are here on Lula.
But the 5DSR, says the filter is there but not activated or something...something funny in their verbage. It looks like they kept it, but somehow it is canceled out at some point??
There seems to be technical reasons for that, probably to keep the total thickness and refraction index of the sensor layering constant among the different cameras in the system so that the specifically designed lenses perform at their best.
Regarding the so called by many "3D effect" of larger formats, I have never read a convinving explanation of it beyond subjective adjectives. Do you think increased sharpness is the objective reason that achieves the "3D look"?.
Regarding the artificial subjects with high contrast edges, for the same reason they are the ones that more clearly show increased sharpness, they are also more prone to display aliasing artifacts (hard staircase edges, moire colour patterns in near-Nyquist repetitive structures,...).
Regards.
Enviado desde mi GT-I9195 mediante Tapatalk
-
Yes and no.
If the AA filter on the Canon 5DS were ideal, i.e. perfect in real world language, the image captured with the 5DS R would never have any more detail than the one from the 5DS since a perfect AA filter would never eliminate or attenuate any valid information. In addition to that, the perfect AA filter would get rid of all of the interference caused by fine detail present in the scene. In such a case the image from the 5DS and perfect AA filter would always be better than the one from the 5DS R (same detail, none of the interference).
But real AA filters are not perfect, so in the real world the 5DS AA filter will not eliminate all of the interference (that is why the 5DS CAN still produce aliasing/moire), and will probably cut some of the fine detail or at least will attenuate it making the valid information look less sharp. This will require extra sharpening that will also enhance the presence of noise in the 5DS.
An exercise to find out how close to a perfect AA filter the 5DS is, is applying an unsharp mask to try to reach the same level of sharpness of the 5DS R, and then judge how much detail is present. The better the AA filter is, the more indistinguisable the images will be:
(http://www.guillermoluijk.com/misc/A.jpg)
(http://www.guillermoluijk.com/misc/B.jpg)
I used a coin to randomize which is A and which is B, the 5DS R or the 5DS. Which one has more detail? is the difference big?.
Regards
I'm not sure where this is leading in my world. The simple test I did originally was to see if I could observe a difference under conditions I could replicate doing a task I'm sometimes asked to do. The fact is I could see a difference, more detail coming from the R camera.
I wasn't going to travel down this road but I have a bit of spare time so I'll indulge myself, also as you have gone to some trouble to try and explain things to me. I just tried to replicate your suggestion. As I mentioned in my first post, I didn't over-do the sharpening in my original image.
So, I added extra unsharp mask to the 5ds image to bring as close to the 5dsr as I can... but I'm still not convinced they are equal in detail. However, I then added the same unsharp mask to the 5dsr and it took another step away from the 5ds with regard to visible detail (see attached)..... adding even more to the 5ds makes the file fall apart.
I'm a practical person who earns his crust taking photographs, not a scientist! My logic was as follows...
- Two cameras are almost identical.
- Camera A reproduces more detail than Camera B.
- I can't reproduce moire under conditions and with subjects I normally shoot with either.
- Both cameras are within my budget.
- Buy camera A.
Result - happiness!
I'm sorry to disappoint those who have more time and inclination to dig deeper, peep at their pixels and procrastinate their way to an early grave. I'm loving the results from my 5dsr, personally.
Carl
-
Apropos Guillermo's posts, I found this article by Tim Parkin interesting as it explores (among other things) the positive impact of sharpening the 5Ds files compared with those of the un-sharpened 5DsR files.
https://www.onlandscape.co.uk/2015/07/testing-the-canon-5dsr/
As an "early adopter" I chose the 5Ds over the 5DsR to avoid the risk of moire: hard to recover from in my line of work. :)
I did take a look at the link to Tim Parkin's test but as soon as I read "More recently, Canon provided me with both the 5DS and the 5DSr and a bunch of lenses 16-35 f/4 IS, 24-70 f/2.8, 24 f/2.8 with which I was to do some in-depth testing." I personally realised it wasn't relevant to me.
Testing high resolution sensors should be done with the very best lenses, otherwise it's more a test of the lens. Using zoom lenses and a medium quality wide angle cannot bring anything conclusive to the table in my opinion.
Don't get me wrong, I own a 16-35 f/4 and a 24-70 f/2.8II and they are good lenses for their purpose, flexibility and saving time.
It's only my opinion though. I think you made a great choice of camera, I think I did too. We're happy bunnies! :)
-
I did take a look at the link to Tim Parkin's test but as soon as I read "More recently, Canon provided me with both the 5DS and the 5DSr and a bunch of lenses 16-35 f/4 IS, 24-70 f/2.8, 24 f/2.8 with which I was to do some in-depth testing." I personally realised it wasn't relevant to me.
Testing high resolution sensors should be done with the very best lenses, otherwise it's more a test of the lens. Using zoom lenses and a medium quality wide angle cannot bring anything conclusive to the table in my opinion.
Don't get me wrong, I own a 16-35 f/4 and a 24-70 f/2.8II and they are good lenses for their purpose, flexibility and saving time.
It's only my opinion though. I think you made a great choice of camera, I think I did too. We're happy bunnies! :)
They (who ever they are) say that a camera is the thing you attach to a lens...but that is just their opinion.
I found that Canon's humble 40mm 2.8 STM works very well with the 5Ds attached. :)
-
This is an interesting thread, and I thank the contributors.
I started the 5ds v 5dsr thread, and I locked it because it degenerated into a Nikon v Canon spat, which bored me. There were very few responses to my question (though a lot more to what was supposed to be my concluding remarks, having bought the camera), and as it turned out, money made the choice pretty easy: £200 might be worth paying for the small difference (and it is small, in reality, I think) but £500 certainly wasn't. IMHO, of course.
I suspect both cameras out-resolve my abilities.
Jeremy
Hi Jeremy, I'm really glad you chimed in... I'm sorry I saw your thread too late. This also quickly turned into a Canon knocking debate.... seems to be back on track now, it's just a non anti-alising filter knocking debate! :)
I hope you're enjoying your camera, that the bit many people seem to miss.
Regards,
Carl
-
They (who ever they are) say that a camera is the thing you attach to a lens...but that is just their opinion.
I found that Canon's humble 40mm 2.8 STM works very well with the 5Ds attached. :)
I agree, all lenses work well in the eyes of the beholder, I've had some great results "lens whacking" my 1DX with an old Contax prime, but I wouldn't assess the potential of a sensor using that technique. ;)
-
I also prefer the 5DsR, after working on hundreds of RAW files in Capture One Pro, Camera RAW, and RawTherapee. It gives an extra pop and texture that I've never seen before in any other camera, including medium format. Whether the advantage is just an illusion or not, I really don't care!
I'm sure the 5DsR Mark II (and the upcoming 5D4 and 6D2) will have on-chip ADC, and Lloyd Chambers and the weekend warriors will finally be able to do their +4 shadow lifting!
-
I also prefer the 5DsR, after working on hundreds of RAW files in Capture One Pro, Camera RAW, and RawTherapee. It gives an extra pop and texture that I've never seen before in any other camera, including medium format. Whether the advantage is just an illusion or not, I really don't care!
I'm sure the 5DsR Mark II (and the upcoming 5D4 and 6D2) will have on-chip ADC, and Lloyd Chambers and the weekend warriors will finally be able to do their +4 shadow lifting!
Robert, I haven't run 5dsr raws through anything except Lightroom to date - though they still blast my previous cameras out of the water.
How did you find Capture Pro and RawTherapee deals with the camera? - do you have a preference?
I have to confess I haven't even considered the Canon Raw software, it's always been so rubbish in the past I didn't want to waist my time.
Cheers,
Carl
-
The most important thing that I discovered with the 5DsR and these different programs is that I've seen jagged edges on some 5DsR files in LR/ACR (even with no sharpening), and it really horrified me. Then, with the same files, I discovered that in C1 and RT, the jaggies were gone, smooth, and perfect. Very interesting (and kind of disappointing for Adobe).
I find that I use multiple programs. Like, process in C1, color noise reduction in LR/ACR, deconvolution sharpening in RT. Sometimes I'll use C1 for everything or RT for everything.
I would say that I prefer C1Pro to ACR, but maybe because I've developed my own settings that I like, and I can get things the way I want a lot faster than any other program. I believe you can download a free trial if you haven't done so already. Make sure you don't run the "DB" version. https://www.phaseone.com/en/Download.aspx
RT really takes some practice, but it is (potentially) by far the sharpest result because of the deconvolution sharpening, the "Contrast by Detail Levels" is really cool too, and it has better algorithms.
There's been a few times when a file has slight moire in LR/ACR, but then it's gone when you open in RT.
Comparing the programs over, and over, and over again, I realized that it's probably impossible to choose a winner. Everyone just has to test their own workflow, and use their eyes to find out what they like. Luckily you can download/test all the programs and see what you prefer.
-
I'm certain you're aware that deep down all us Canon uses know we made a huge mistake; we're full of regret and disappointed with the performance of our cameras and ultimately disheartened by the photographs they produce. Nikon and Sony are much, much better cameras in every way and you guys really know your stuff... can we carry on with the reason for the thread now "5ds or 5dsr"?
Indeed.
Carl, I think it has a lot to do with the subject matter one shoots. Technically, a discrete image sensor has to be bandwidth limited (by an AA-filter), it's Digital Signal Processing 101. The simple reason is that it will otherwise generate aliasing artifacts, such as stairstepping, coarser(!) or fake moiré detail, False color aliasing, and it is harder to sharpen without exaggerating the aforementioned artifacts. However, that will only happen if there is more/finer detail than the sensor can resolve. Having a high resolution sensor with dense sampling helps, as does getting the subject as large as possible in the frame, with little cropping.
That means that the subject matter has to be in the plane (a DOF zone) of optimal focus, otherwise defocus will kill the potentially troublesome subject detail. Also narrow apertures that cause diffraction blur will reduce the chance of fine detail getting through with enough modulation to cause aliasing issues. And obviously the lens must be reasonably good. So, one may get away with the negative effects of the lacking AA-filter, but it's more risky because when it strikes, there is little to do about it (except time consuming postprocessing).
The 5DS does not totally eliminate the chance of Aliasing artifacts from damaging the subjects (surface) structure, because it is underdimensioned, but it does reduce common risks. Proper deconvolution Capture sharpening will restore a lot of the detail/contrast that looked like it had been lost. Sharpening 5DS R shots risk bringing out the aliasing, if there is any. So with more sharpening potential for the 5DS and less (or only targeted) for the 5DS R, the effective visual resolution differences will be small. There will be differences, but they will be small.
There is one additional factor that makes a difference, and that is the Raw converter. Capture One for instance does a very good job handling the aliasing artifacts that both cameras can generate. It can't unscramble the mess that aliasing is, but it does not over-accentuate it either, while still extracting higher levels of detail than e.g. Lightroom does.
Cheers,
Bart
-
I'm certain you're aware that deep down all us Canon uses know we made a huge mistake; we're full of regret and disappointed with the performance of our cameras and ultimately disheartened by the photographs they produce. Nikon and Sony are much, much better cameras in every way and you guys really know your stuff... can we carry on with the reason for the thread now "5ds or 5dsr"?
Sarcasm aside, there are Nikon/Sony shooters that think having better sensor performance magically makes everything better and on the other side, Canon shooters who don't argue higher levels of performance don't matter. Both are wrong. Every photographic opportunity is different and requires a specific level of performance of one type or another. As long as the tool meets those performance requirements, all else equal, then having extra performance doesn't matter.
However, having a performance margin does allow:
1. Executing some task that is impossible otherwise
2. Executing to a higher level of quality
3. Executing more efficiently or effectively
4. Forgiveness when the execution isn't exact
If we compare winding film with our thumb, 4fps cameras and 10fps cameras. All 3 can make great photographs. However, the 4fps camera provides those 4 benefits over winding with our thumb and the 10fps provides an additional margin of those benefits over 4fps. We don't always need the extra performance, but that does not mean it is not beneficial to have at our disposal when we do.
-
... It is always helpful to stick to what is actually said, rather than introducing things that were not said.
Or not asked (like 5Ds, 5Dsr, or...?) ;)
-
Gosh, the first time I tried rapid drive (5 fps) on the 60D, my first DSLR, I thought it was an ef*-ng miracle. We really are fortunate to have all these options. I remember the thumb days well. One thing about the thumb limitations - it concentrated the mind wonderfully, because you had ONE chance at the shot.
-
I added extra unsharp mask to the 5ds image to bring as close to the 5dsr as I can... but I'm still not convinced they are equal in detail. However, I then added the same unsharp mask to the 5dsr and it took another step away from the 5ds
Never intended to offend you Carl. I know it must be difficult to have a clear preference for the images from the 5DS R, because your eyes indicate so, and then listen to someone who claims a part of the extra detail might be just a form of noise. But as Bart ellaborated much better than me, this is the truth in digital sampling systems. And this is the reason for AA filters to exist.
The good news is that the more dense a sensor is compared to the resolution of the lenses used, and the 5DS is already very dense for FF optics, the less significant is the role played by the AA filter. For this reason you should not have many aliasing/moire problems on the 5DS R that would be solved on the 5DS, and you should not have much extra detail on the 5DS R than with the 5DS and proper sharpening.
Eliminating the AA filter from a 5Mpx FF sensor would be simply suicide.
Regards.
Enviado desde mi GT-I9195 mediante Tapatalk
-
The most important thing that I discovered with the 5DsR and these different programs is that I've seen jagged edges on some 5DsR files in LR/ACR (even with no sharpening), and it really horrified me. Then, with the same files, I discovered that in C1 and RT, the jaggies were gone, smooth, and perfect. Very interesting (and kind of disappointing for Adobe).
I find that I use multiple programs. Like, process in C1, color noise reduction in LR/ACR, deconvolution sharpening in RT. Sometimes I'll use C1 for everything or RT for everything.
I would say that I prefer C1Pro to ACR, but maybe because I've developed my own settings that I like, and I can get things the way I want a lot faster than any other program. I believe you can download a free trial if you haven't done so already. Make sure you don't run the "DB" version. https://www.phaseone.com/en/Download.aspx
RT really takes some practice, but it is (potentially) by far the sharpest result because of the deconvolution sharpening, the "Contrast by Detail Levels" is really cool too, and it has better algorithms.
There's been a few times when a file has slight moire in LR/ACR, but then it's gone when you open in RT.
Comparing the programs over, and over, and over again, I realized that it's probably impossible to choose a winner. Everyone just has to test their own workflow, and use their eyes to find out what they like. Luckily you can download/test all the programs and see what you prefer.
Robert, this is gold, thank you. I did have a play with C1 a few years ago and while I liked the colour it produced I found the workflow and terminology a nightmare to get my head around, also the trial copy I downloaded kept falling over! ...disconcerting when you're trying to run a business. I think that's why I've always steered back to Adobe, it may not be the best in some respects but it never leaves me in a panic!
I'm going to give the latest C1 another chance to woo me.
Cheers, for your thoughts,
Carl
-
Indeed.
Carl, I think it has a lot to do with the subject matter one shoots. Technically, a discrete image sensor has to be bandwidth limited (by an AA-filter), it's Digital Signal Processing 101. The simple reason is that it will otherwise generate aliasing artifacts, such as stairstepping, coarser(!) or fake moiré detail, False color aliasing, and it is harder to sharpen without exaggerating the aforementioned artifacts. However, that will only happen if there is more/finer detail than the sensor can resolve. Having a high resolution sensor with dense sampling helps, as does getting the subject as large as possible in the frame, with little cropping.
That means that the subject matter has to be in the plane (a DOF zone) of optimal focus, otherwise defocus will kill the potentially troublesome subject detail. Also narrow apertures that cause diffraction blur will reduce the chance of fine detail getting through with enough modulation to cause aliasing issues. And obviously the lens must be reasonably good. So, one may get away with the negative effects of the lacking AA-filter, but it's more risky because when it strikes, there is little to do about it (except time consuming postprocessing).
The 5DS does not totally eliminate the chance of Aliasing artifacts from damaging the subjects (surface) structure, because it is underdimensioned, but it does reduce common risks. Proper deconvolution Capture sharpening will restore a lot of the detail/contrast that looked like it had been lost. Sharpening 5DS R shots risk bringing out the aliasing, if there is any. So with more sharpening potential for the 5DS and less (or only targeted) for the 5DS R, the effective visual resolution differences will be small. There will be differences, but they will be small.
There is one additional factor that makes a difference, and that is the Raw converter. Capture One for instance does a very good job handling the aliasing artifacts that both cameras can generate. It can't unscramble the mess that aliasing is, but it does not over-accentuate it either, while still extracting higher levels of detail than e.g. Lightroom does.
Cheers,
Bart
Bart and Guillermo, thank you for taking the time to explain the principles of digital imaging to me (and anyone else reading) in such detail. It is refreshing to have the science laid out in a way I can understand.
Taking it all on board, are you guys suggesting that the anti-alaising cancellation camera options are simply a marketing ploy? My real world experience of the cameras still stands up as far as I'm concerned from the testing I did that was relevant to my way of working as well as subsequent jobs I've shot, so there's no buyers remorse, just intrigue.
Also, Guillermo, you didn't offend me in the slightest!
Many thanks,
Carl
-
Nice car Carl, (just to return this thread back to sanity)
Give C1 v9 a try. I know its different to use than the Adobe products but the results are worth the effort.
-
To me, as a fashion shooter, it was a no-brainer to go with the 5Ds. Moire is a killer. I remember spending A LOT of time removing moire from my 4MP Canon 1D files. (Jeans were a bitch.) I was lucky that the early Capture One software came with a moire-removing plug-in. Now, for the few times, it's a no-brainer in Lightroom.
That said, in various images I've seen, I can't say I've ever seen that much of a difference between the two files -- 5Ds vs 5DsR. Yes, ever so slight softening of the edges, but to me, that also translates into what I see -- smoother skin tones. I've always thought, previously, that higher res cameras might show far more detail on a girl's face than I wished to see! In reality, though I have more detail, I also have smoother tones.
The funny thing now is that my Canon 1Dx files look low res! I use to have to massage the 1Dx files up for 8' in-store posters for a client. (The files are produced, by me, at 1/4 scale at 400 dpi.) With the 5Ds, they are at their native res -- much less on my part to make them look good. So the 5Ds translates into less time spent per image on digital imaging. Regardless of what some folks on the internet say, the 5Ds produces a very high quality file. Will it shoot at 1 million ISO? No. But I defy almost anyone to show me all the times they shoot at that. It's bragging rights or specialized photography. For me, I rarely shoot more than 500 ISO, and occasionally 800. I only wish the sensor had made its way into the 1Dx body. I've never really liked the ergonomics of the 5D, but have managed. ;)
-
That said, in various images I've seen, I can't say I've ever seen that much of a difference between the two files -- 5Ds vs 5DsR. Yes, ever so slight softening of the edges, but to me, that also translates into what I see -- smoother skin tones. I've always thought, previously, that higher res cameras might show far more detail on a girl's face than I wished to see! In reality, though I have more detail, I also have smoother tones.
Question: It is my understanding that the AA filter is still in place on the 5DsR and the effect is then negated. Is this in fact how the 5DsR works? Unlike say the Nikon D800 twins where one had the AA filter and the other just didn't.
-
Question: It is my understanding that the AA filter is still in place on the 5DsR and the effect is then negated. Is this in fact how the 5DsR works?
Correct, as far as the information that is know is correct.
Unlike say the Nikon D800 twins where one had the AA filter and the other just didn't.
Not correct, the D800E also used/uses an AA cancellation filter stack. The filter stack thickness is part of the optical design, so leaving parts out would negatively impact resolution/aberrations.
So nothing new under the sun with the Canon approach compared to the Nikon approach.
Cheers,
Bart
-
Not correct, the D800E also used/uses an AA cancellation filter stack. The filter stack thickness is part of the optical design, so leaving parts out would negatively impact resolution/aberrations.
So nothing new under the sun with the Canon approach compared to the Nikon approach.
Cheers,
Bart
Oh, so it wasn't till the D810 that the AA filter was removed completely. Hmm. Though I'm pretty sure that the D7100 also did not have one either.
-
Nor did my Oly C2500L :-)
I think C1 does better with the 5DsR files as its not designed to deal with AA files from its core. Not that it can't, but it has a easier time to deal with non-AA better than others such as LR.
Guillermo...
To further touch on the 3D look of the nonAA, I think without it, there is zero level of a film that causes a neutralizing effect, and sharpness surely helps this to define edge contrast and sepearation.
Its not something you will get out of reading, but once you see it, you will surely now it.