Pages: 1 [2]   Go Down

Author Topic: 5ds or 5dsr - my findings  (Read 13446 times)

one iota

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 74
Re: 5ds or 5dsr - my findings
« Reply #20 on: February 12, 2016, 05:54:12 am »

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. :)
Logged
Mahn England

carl dw

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 80
Re: 5ds or 5dsr - my findings
« Reply #21 on: February 12, 2016, 05:54:35 am »

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





Logged

carl dw

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 80
Re: 5ds or 5dsr - my findings
« Reply #22 on: February 12, 2016, 06:02:59 am »

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.  ;)
Logged

RobertJ

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 706
Re: 5ds or 5dsr - my findings
« Reply #23 on: February 12, 2016, 06:35:23 am »

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!
Logged

carl dw

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 80
Re: 5ds or 5dsr - my findings
« Reply #24 on: February 12, 2016, 06:59:56 am »

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



Logged

RobertJ

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 706
Re: 5ds or 5dsr - my findings
« Reply #25 on: February 12, 2016, 09:33:32 am »

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.
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: 5ds or 5dsr - my findings
« Reply #26 on: February 12, 2016, 12:11:07 pm »

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
« Last Edit: February 13, 2016, 07:32:06 am by BartvanderWolf »
Logged
== If you do what you did, you'll get what you got. ==

dwswager

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1375
Re: 5ds or 5dsr - my findings
« Reply #27 on: February 12, 2016, 01:13:01 pm »

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.
Logged

Slobodan Blagojevic

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 18090
  • When everyone thinks the same, nobody thinks
    • My website
Re: 5ds or 5dsr - my findings
« Reply #28 on: February 12, 2016, 02:02:41 pm »

... 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...?) ;)

NancyP

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2513
Re: 5ds or 5dsr - my findings
« Reply #29 on: February 12, 2016, 02:22:02 pm »

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.
Logged

Guillermo Luijk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2005
    • http://www.guillermoluijk.com
Re: 5ds or 5dsr - my findings
« Reply #30 on: February 13, 2016, 06:52:04 am »

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
« Last Edit: February 13, 2016, 11:16:03 am by Guillermo Luijk »
Logged

carl dw

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 80
Re: 5ds or 5dsr - my findings
« Reply #31 on: February 13, 2016, 09:34:37 am »

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

Logged

carl dw

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 80
Re: 5ds or 5dsr - my findings
« Reply #32 on: February 13, 2016, 10:00:26 am »

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
Logged

Craig Lamson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3264
    • Craig Lamson Photo Homepage
Re: 5ds or 5dsr - my findings
« Reply #33 on: February 13, 2016, 11:17:11 am »

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.
Logged
Craig Lamson Photo

nemophoto

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1021
    • Nemo Niemann Photography
Re: 5ds or 5dsr - my findings
« Reply #34 on: February 23, 2016, 09:23:31 am »

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.  ;)
Logged

dwswager

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1375
Re: 5ds or 5dsr - my findings
« Reply #35 on: February 23, 2016, 09:40:49 am »

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.
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: 5ds or 5dsr - my findings
« Reply #36 on: February 23, 2016, 09:49:33 am »

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.

Quote
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
Logged
== If you do what you did, you'll get what you got. ==

dwswager

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1375
Re: 5ds or 5dsr - my findings
« Reply #37 on: February 23, 2016, 10:10:48 am »

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.
Logged

Phil Indeblanc

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2017
Re: 5ds or 5dsr - my findings
« Reply #38 on: February 23, 2016, 07:59:20 pm »

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.
Logged
If you buy a camera, you're a photographer...
Pages: 1 [2]   Go Up