Is the Canon EOS 5DS series to be replaced by a mirrorless camera? [CR1]

[dtaylor]

I agree. But in talking about the future, we have to acknowledge that DSLRs are older technology. Replacing the optical system with electronics is inevitable. It will also be cheaper. Maybe not now, but certainly in a few years. It will also be better. Some people want to believe that DSLRs should always remain, and that’s just not going to happen.

The market will decide what happens. That said, I agree that over a long enough time scale EVFs will likely replace OVFs. I just get annoyed at the gleeful proclamations that "the DSLR is dead!" and baseless product line predictions based on same.

And I hope that the SLR form factor remains because holding big glass with most mirrorless bodies is a literal pain.

 

[AlanF]

Yes I can. Absent an AA filter resolved detail is going to be governed by Nyquist given the prime lenses and apertures used. If the AA filtered sensor hits as close to its Nyquist limit as the unfiltered sensor then it's safe to say that AA filters do not generally impact extinction resolution.
Add to this that I have yet to see an image where the 5DsR resolved more detail than the 5Ds, and I looked for them. I downloaded, I pixel peeped, I printed, I studied...all things I'm guessing you did not do.
I would love to through IR's resolution tests at you but both 5Ds cameras out resolve the IR target.
You have cited neither reliable measurements nor any relevant laws of physics. Don't claim they support you unless and until you do.

First of all, I did cite reliable evidence - from lensrentals: here is the link to measuring the increased resolution of the 5DSR over the 5DS - https://www.lensrentals.com/blog/2015/06/canon-5ds-and-5ds-r-initial-resolution-tests/ and there is more to come.

Secondly, you claimed that there are no other pairs of cameras with and with and without AA-filters. Well, I am afraid there are gaps in your knowledge, because there are well-documented examples that provide more evidence. First, there is the Nikon D800, which was followed by the D800E, which is the same camera minus the AA-filter. There are several reports of the increased resolution of the D800E. Here are 3.
Nikon D800 review updated with D800E side-by-side testing
https://www.dpreview.com/articles/1879419250/nikon-d800e-detail-added-to-nikon-d800-review
Discerning the Differences between the Nikon D800 and D800E
https://www.bhphotovideo.com/explor...ning-differences-between-nikon-d800-and-d800e
Lensrentals
https://www.lensrentals.com/rent/nikon-d800e
In addition, the Sony DSC-RX1R II has a switchable off, standard and high AA-filter (LPF)
https://www.sony.co.uk/electronics/cyber-shot-compact-cameras/dsc-rx1rm2
AP has measured the MTFs with the LPF off, compromise medium and off:
"5,600l/ph is gained at ISO 100 when the LPF is turned off, and although it has obvious signs of colour moiré and grid-like aliasing, it boasts extremely good resolution. In the LPF standard mode, and at the same ISO, a score of 5,200l/ph is achieved; LPF high gives a 4,800l/ph score."
The Sony has a very high resolution lens.
You quote the Nyquist limit. But, that is not the only factor. The difficulty in resolving two overlapping lines or points caused by blurring is the same in principle as resolving two lines that are blurred by other physical factors such as those behind the Rayleigh Criterion - once you have blurred the detail, simple sharpening doesn't help. USM etc can sharpen edges, but not restore blurred resolution. I hope that those links I have given you convince you that you I did have supportive evidence.
It would be good if Canon had a switchable on/off AA-filter on the 5DSRII.

 

[I Simonius]

I don’t believe the average 5DS user would be all that bothered by adapting EF lenses,

I don't know whether I'm the average 5DS (R) user or not but I would definitely not buy or use any adapted lenses. So I do not concur with your comment as I would most decidedly be bothered.

Nor, to my mind, is there any advantage to a mirrorless version of the 5DS; the opposite in fact.

However if I am the average 5DS user then Canon has a problem, as for me, this camera only has one possible area of improvement, that being greater DR.

I have yet to see any EVF that gives me anything like the same experience as looking through the lovely OVF of the 5DS. This is a very important part of the photography experience for me.

Everything else about the 5DS is perfect (any more pixels and f-stop choice decreases and DOF suffers due to circle of confusion issues, assuming you want the absolute maximum out of this camera) I simply cannot fault it.

I'm a stills photographer with little or no interest in video so I assume the idea that some have that the 5DS is outdated comes from Videographers; or those that need to shoot hairs on clangers' knees on the moon at f1.2.

Increased dynamic range is the only reason I'd buy another 5DS. I wouldn't buy any other FF DSLR because for my purposes Canon has made the perfect camera in the 5DS (R)

All the talk of one camera outperforming another is irrelevant to me now I have the 5DS (R) simply because for my purposes I finally have the tool that does exactly what I want and need (barring greater DR as I frequently shoot very high contrast scenes)

I realise I must be failing in my duty as a good consumer in not wanting perpetually to improve my gear; maybe I'm not the average 5DS (R) user after all

 

[Sporgon]

Secondly, you claimed that there are no other pairs of cameras with and with and without AA-filters. Well, I am afraid there are gaps in your knowledge, because there are well-documented examples that provide more evidence. First, there is the Nikon D800, which was followed by the D800E, which is the same camera minus the AA-filter. There are several reports of the increased resolution of the D800E. Here are 3.
Nikon D800 review updated with D800E side-by-side testing
https://www.dpreview.com/articles/1879419250/nikon-d800e-detail-added-to-nikon-d800-review
Discerning the Differences between the Nikon D800 and D800E
https://www.bhphotovideo.com/explor...ning-differences-between-nikon-d800-and-d800e

Typical dpreview - begging a difference. Fancy showing the 800 and 800E with the same amount of sharpening, and 100% of .6 pixel is quite significant. Of course the 800E is going to look shaper - it looks brittle in fact.

The way to test the two side by side is for the 800E to have no sharpening, then subtle sharpening on the 800 to equal the perceived sharpness and contrast of the 800E. Then view the difference.

Incidentally that 100% crop shows what minutiae we are discussing anyway: look how small the detail is in the full image.

I wholeheartedly agree with dtaylor's comments. However before I bought the 5DS I borrowed one and rented a 5DSR, and I tested them together. With appropriate subtle sharpening on the 5DS ( about 100% of 0.3 pixel) there is naff all difference, even at 200%. I could have a good laugh on here by putting up the images that I took and see how no one would be able to tell which was from which camera, despite all this nonsense about 10% more resolution. What might give the game away however is the false colour showing in the 5DSR files, just like in the dpreview samples.

I can understand those who want to crop very tightly and base their (cropped) picture on very fine detail liking the non AA filtered camera: just don't wish non AA filtered cameras onto those of us who prefer them with, and harping on about how much detail we are missing.

Given the limitations of a bayer array sensor, technically these should have an appropriate AA filter over them. I hope that when the next generation of 5DS / SR cameras come out Canon will, unlike Nikon, continue to give us the option.

 

[AlanF]

The best option is that provided by Sony: to be able to toggle from no AA-filter, to weak AA-filter to strong. In that way, those who can see the benefits of lacking an AA-filter will be happy as will fashion photographers who need to avoid Moire like the plague. It will also take care of those akin to climate warming deniers and flat earthists who deny measurements and scientific evidence and choose to follow their gut instincts, come what may.

 

[3kramd5]

The best option is that provided by Sony: to be able to toggle from no AA-filter, to weak AA-filter to strong.

Which Sony?

Edit never mind I found it above. I wonder if canceling a filter is really the same thing as having no filter. Surely there is some loss involved.

 

[AlanF]

Which Sony?

Edit never mind I found it above. I wonder if canceling a filter is really the same thing as having no filter. Surely there is some loss involved.

https://www.sony.co.uk/electronics/cyber-shot-compact-cameras/dsc-rx1rm2
Whatever the losses, if any, not activating the AA-filter increases resolution by 16.7% These MTF measurements on the same same camera with the filter either on or off are the best direct demonstration of the effects of the filter on resolution - it is absolutely indisputable that the filter lowers resolution.
I doubt if cancelling the filters results in any significant loss relative to the filter not being there at all, just second-order effects. It's a neat idea giving you the choice of an AA-filter and I really hope Canon adopts it.

 

[RGF]

Interesting that Canon did not take this opportunity to increase (perhaps only slightly) the size of the sensor. Not going to MF, but perhaps by 10 to 25%. That would help with S/N though old EF lens would be limited to 36 - equivalent area on the sensor.

 

[tmc784]

5DS-R without lens adapter for EF lens

 

[melgross]

Or, maybe it will find a niche between the 6DII and the 5DIV, as its pricing suggests.

That’s possible too.

 

[Sporgon]

https://www.sony.co.uk/electronics/cyber-shot-compact-cameras/dsc-rx1rm2
Whatever the losses, if any, not activating the AA-filter increases resolution by 16.7% These MTF measurements on the same same camera with the filter either on or off are the best direct demonstration of the effects of the filter on resolution - it is absolutely indisputable that the filter lowers resolution.
I doubt if cancelling the filters results in any significant loss relative to the filter not being there at all, just second-order effects. It's a neat idea giving you the choice of an AA-filter and I really hope Canon adopts it.

Well another piece of irrefutable evidence that you've provided - Sony believes that even a 42 mp sensor requires an AA filter option

 

[AlanF]

Well another piece of irrefutable evidence that you've provided - Sony believes that even a 42 mp sensor requires an AA filter option

I see your emoji. But, the irrefutable evidence is that you have a choice between highest resolution with the possibility of more aliasing artefacts and lower resolution with lower probability of aliasing artefacts. The higher the resolution sensors, the more the Moire is pushed away. 50 mpx sensors are far less susceptible to Moire than 24 mpx.

 

[dtaylor]

First of all, I did cite reliable evidence - from lensrentals:

Lens Rentals was not testing extinction resolution. Do you understand the difference between testing at a particular contrast level (i.e. MTF50) and testing for extinction resolution?

Secondly, you claimed that there are no other pairs of cameras with and with and without AA-filters.

No, I said it's hard to compare a lot of Canon cameras with non-AA filter cameras from other manufacturers because Canon often has 'odd' MP sizes.

Well, I am afraid there are gaps in your knowledge, because there are well-documented examples that provide more evidence. First, there is the Nikon D800, which was followed by the D800E, which is the same camera minus the AA-filter. There are several reports of the increased resolution of the D800E. Here are 3.

Let's talk about those three articles...

The dpreview article did not claim higher extinction resolution on the D800E. They couldn't because they use the same chart as Imaging Resource and the D800/D800E both out resolve the chart in terms of extinction resolution.

The bhphoto article did not talk about extinction resolution at all and had no testing of any kind, resolution or otherwise. Why would you cite it?

The LensRentals.com link consisted of a comment from Roger which almost certainly was in reference to MTF50 resolution based off their lens tests. No tests, no discussion of extinction resolution...why would you cite it?

In addition, the Sony DSC-RX1R II has a switchable off, standard and high AA-filter (LPF)
https://www.sony.co.uk/electronics/cyber-shot-compact-cameras/dsc-rx1rm2
AP has measured the MTFs with the LPF off, compromise medium and off:
"5,600l/ph is gained at ISO 100 when the LPF is turned off, and although it has obvious signs of colour moiré and grid-like aliasing, it boasts extremely good resolution. In the LPF standard mode, and at the same ISO, a score of 5,200l/ph is achieved; LPF high gives a 4,800l/ph score."

That's interesting and the only relevant piece of information you've cited. But it's also a one-off camera with a unique LPF design. That their LPF impacts extinction resolution does not mean regular LPFs will do the same.

The difficulty in resolving two overlapping lines or points caused by blurring is the same in principle as resolving two lines that are blurred by other physical factors such as those behind the Rayleigh Criterion - once you have blurred the detail, simple sharpening doesn't help.

That depends entirely on the amount of blur.

USM etc can sharpen edges, but not restore blurred resolution.

Resolved detail that is slightly blurred is still resolved detail. And if it's resolved but blurred then we're discussing sharpness, not resolution.

 

[AlanF]

dtaylor, if you want to claim that controlled measurements are irrelevant compared with your eyeballing and scientific logic, then so be it - it won't affect any of my decisions. Talking of Lensrentals, are you the Daniel Taylor who is getting attacked on https://www.lensrentals.com/blog/2018/10/teardown-of-the-canon-eos-r-mirrorless-camera/ for claiming his eyeballs are better than eyeAF etc and what he doesn't use others don't need?

 

[3kramd5]

No, I said it's hard to compare a lot of Canon cameras with non-AA filter cameras from other manufacturers because Canon often has 'odd' MP sizes.

It’s more than hard; it’s maybe impossible . To really make a determination, you need all else equal. Not just the same sensor; you need the same downstream electronics running the same code (to insure is no cooking of the file, or if there is cooking that is the consistent), same set up, same lighting, same charts, same converter, etc. Crossing brands, all bets are off.

 

[Hector1970]

dtaylor, if you want to claim that controlled measurements are irrelevant compared with your eyeballing and scientific logic, then so be it - it won't affect any of my decisions. Talking of Lensrentals, are you the Daniel Taylor who is getting attacked on https://www.lensrentals.com/blog/2018/10/teardown-of-the-canon-eos-r-mirrorless-camera/ for claiming his eyeballs are better than eyeAF etc and what he doesn't use others don't need?

No need to make it personal. Maybe he is better than eyeAF (which while I've find its good isn't necessarily earth shattering - its certainly far less deliberate than a good photographer). I can see eyeAF is useful but I think it makes for poorer photographs as the photographer is less in control of what is being photographed.
DTaylor has proven than in any arguement about cameras two opposing opinions can dig up plenty of website information that supports their claims.
In the end its the photographers personal experience that matters.
You've written an encyclopedia at this start on the greatness of the 5DSR. Overall I find the 5DIV a far superior all round camera and that's my experience taking hundreds of thousands of photographs with both camera often side by side. It's not that the 5DSR is bad. You completely disagree which is fine.

 

[AlanF]

How you rate a camera as all round depends on how you weight the individual features in your scores - a bit like DPR giving an overall score for a camera or DxOmark for a lens - which is bound to provoke both discussion and indignation! What matters to you are your the features that you value and you use the most, and those will determine the all round value to you. I also routinely use both the 5DIV and 5DSR and choose which one depending on the circumstances: for higher resolution, the 5DSR is my go-to body; for snappy AF, especially, BIF, I prefer the 5DIV. On our safari beginning next week, I'll be using the 5DIV with the 400mm DO II plus extenders and my wife the 5DSR with the 100-400mm II. In those circumstances, she will have the more useful all round combination being able to have good reach and the ability for close ups without changing extenders but I'll get the better long distance performance with the 2xTC. (Mind you, I will miss having the the 5DSR + 400mm DO II + 1.4xTC, but we have to make sacrifices...)

Although it might be heresy to say so in this forum, the Nikon D850 beats both as an all round stills camera because it combines a high mpx sensor without an AA-filter, very fast AF and reasonable fps. But, I am confident that Canon will match it or overtake it.

 

[Hector1970]

How you rate a camera as all round depends on how you weight the individual features in your scores - a bit like DPR giving an overall score for a camera or DxOmark for a lens - which is bound to provoke both discussion and indignation! What matters to you are your the features that you value and you use the most, and those will determine the all round value to you. I also routinely use both the 5DIV and 5DSR and choose which one depending on the circumstances: for higher resolution, the 5DSR is my go-to body; for snappy AF, especially, BIF, I prefer the 5DIV. On our safari beginning next week, I'll be using the 5DIV with the 400mm DO II plus extenders and my wife the 5DSR with the 100-400mm II. In those circumstances, she will have the more useful all round combination being able to have good reach and the ability for close ups without changing extenders but I'll get the better long distance performance with the 2xTC. (Mind you, I will miss having the the 5DSR + 400mm DO II + 1.4xTC, but we have to make sacrifices...)

Although it might be heresy to say so in this forum, the Nikon D850 beats both as an all round stills camera because it combines a high mpx sensor without an AA-filter, very fast AF and reasonable fps. But, I am confident that Canon will match it or overtake it.

Enjoy your Safari. I’m not long back from Kenya and it was a great experience. I was using mainly a 600 F4 II + 5DIV which was great if a little awkward (length wise and it’s admittedly heavy) and. 100-400 II which was fine but noticeably inferior (it was also stuck with the 5DSR ) but more flexible.
Tough trip all round but very rewarding. You could do a safari there ignore the animals and focus only on birds and be kept busy all day. An amazing cultural experience. I was in Tanzania before close to 20 years ago. It really emphasized to me how good Cameras and lens have got. Differences between 5DSR and 5DIV are nothing in the bigger picture

 

[Mt Spokane Photography]

How you rate a camera as all round depends on how you weight the individual features in your scores - a bit like DPR giving an overall score for a camera or DxOmark for a lens -....

Exactly. A camera rating site could, perhaps come up with a list of attributes that a viewer can check off or put a priority on, then show the cameras that score highest in those categories. That would be unlikely, because ratings for things like autofocus are really tough to pin down, many factors affect it, light intensity, light color, contrast, even air temperature, so any score is going to be biased toward testing conditions and not towards your specific use.

If you don't check video or 4K video, then the video rating would not be factored into your results. If you want 40+ mp, than that would be factored in, it might be another way of helping a buyer find what works best for their usage. I get the feeling that many people buy capabilities they don't use, and such a person is sitting right here!

I do use corded tethering, and occasionally, wireless tethering, so performance in those areas might be a higher priority for me than for most. Generally, Canon cameras score very high in tethering, so that's one of my likes. I also like the touch screen, because with my loss of feeling in my fingers, I find it much faster and easier most of the time. I have a big problem feeling the shutter button, and have glued a raised button on my cameras to help me feel it. I'm not aware of any camera that would be outstanding for me as far as feeling the shutter button. Ergonomics would be another impossible to pin down rating.

I think that different cultures have different expectations as well, perhaps the terminology on the screen might be better or worse, depending on translation?

 

[dtaylor]

I thought I replied to this already but I don't see the post?

dtaylor, if you want to claim that controlled measurements are irrelevant compared with your eyeballing and scientific logic, then so be it -

You haven't cited any relevant controlled tests. Opinion pieces with no testing are not controlled tests. And tests for MTF50 resolution tell you nothing about extinction resolution.

Talking of Lensrentals, are you the Daniel Taylor who is getting attacked on https://www.lensrentals.com/blog/2018/10/teardown-of-the-canon-eos-r-mirrorless-camera/ for claiming his eyeballs are better than eyeAF etc and what he doesn't use others don't need?

Do I have a stalker now?