DPReview Post R5 and R6 in comparison tool

Keith_Reeder

I really don't mind offending trolls.
Feb 8, 2014
960
477
63
Blyth, NE England
to me its more like moire effect rather than sensor banding.
Me too - I was around during the Great Nikon D200 Sensor Banding Wars, and the famous Battle Of The Canon 7D, and this isn't like any "sensor banding" they ever manifested. In fact, at 300% view (100 ISO again), it's definitely moire:

moire.jpg

Moire is easily dealt with. And it's the price you pay for light anti-aliasing. A finely cross-hatched area of a lino-cut style print like this is just where you'd expect to see it.

We're seeing the same kind of false colour (to a far lesser extent) that's shown in the second image example on this page:

 
Last edited:
  • Like
Reactions: 1 users
Upvote 0
Jan 29, 2011
10,675
6,121
and the reason is (likely): left and right files are identical, the same file being compared..due to a human error :) I found a spot in the image that is purely noise pattern and it is identical in both samples... I am not going to bid my house on it. but... it is a miracle otherwise.. not what TDP samples are showing. there is a difference and noticeable in the level of details available at ISO6400 specifically. R6 is slightly more detailed.
But when they are not normalized the images are different sizes, the file names are different and the file sizes different. At this point though I'd still be wary of minuscule differences, nobody has a well developed RAW converter and that will only get better, but initial results show no practical differences at high iso between the two when normalized, which is exactly what should be expected.
 
  • Like
Reactions: 3 users
Upvote 0
Jan 29, 2011
10,675
6,121
and the reason is (likely): left and right files are identical, the same file being compared..due to a human error :) I found a spot in the image that is purely noise pattern and it is identical in both samples... I am not going to bid my house on it. but... it is a miracle otherwise.. not what TDP samples are showing. there is a difference and noticeable in the level of details available at ISO6400 specifically. R6 is slightly more detailed.
I was interested in this possibility so I screenshot the two as that is the only way to compare what you think the issue might be (as any link might go to a different but correct file). Anyway I opened it in PS copied it to another layer and put it in 'difference' mode. When moved to cover the other picture it should show as black but it doesn't, the two files even as displayed are not the same file.

1595691263965.png
 
  • Like
Reactions: 1 user
Upvote 0
Jan 29, 2011
10,675
6,121
but back to your request. Here is a screen shot. (R5 and R6)View attachment 191536

Same thing done with this screenshot to get the best (darkest) result from the patch, they are not the same file.

1595691517650.png

To be clear, if the files are the same the result is pure black, it might look like they are but they aren't. Here is what it should look like if they are the same file, this is the screenshot with a copy on top turned to 'difference' mode. pure black every pixel.

1595691673852.png
 
Upvote 0

SecureGSM

2 x 5D IV
Feb 26, 2017
2,360
1,231
Same thing done with this screenshot to get the best (darkest) result from the patch, they are not the same file.

View attachment 191541

To be clear, if the files are the same the result is pure black, it might look like they are but they aren't. Here is what it should look like if they are the same file, this is the screenshot with a copy on top turned to 'difference' mode. pure black every pixel.

View attachment 191542
Yup, I stand corrected. files are indeed not the same. apologies. .
 
Upvote 0
Jan 29, 2011
10,675
6,121
moire is unlikely. colour banding it is.. and regardless what you call this : it's crap... I am not particularly concerned about Nikon or Sony... However if that's what it is then obviously I am not going for R5.. If this is not an issue on DPR side, I hope that Canon will fix the issue sooner or later..I do not want any of this rubbish in my high ISO shots. how one is going to fix that?
No it looks like moire, it is on the full sized image too. where it has rendered the grain in the wood beam as a mesh. it is the bain of high resolution sensors and interference pattern sized detail.

1595692785867.png
 
  • Like
Reactions: 1 users
Upvote 0

Michael Clark

Now we see through a glass, darkly...
Apr 5, 2016
4,722
2,655
1. +++With the claim about the R6 having more detail, you are referring to the word 'green' having higher contrast edges on the R6? That is not necessarily an indication of detail, especially since small differences such as these can also originate from the scaling methods used.

A.M.: it well may be that you are correct and I am incorrect. I looked at a number of methods that Bryan used: P.S., DPP... there is a difference. Is a small difference but visible to my eye. And it is not only a higher contrast edges... not only an acutance... There is a small difference in how details were rendered.
2. DPR tool:

there is absolutely no difference in details level or noise characteristics between R5, R6 or.... 5D4. I just looked at R5 vs 5D4 at ISO 6400. There is a bit of difference in colour reproduction. And that’s all.
We all know however that 5D4 is not as great as 1Dx3 at ISO 6400. Slightly less great. However the tool does not present any tangible difference. At all. I am looking at a very large screen now.

but back to your request. Here is a screen shot. (R5 and R6)View attachment 191536

I'm going to throw out a WAG here and say those two spots are dust on the surface of the test target?
 
  • Like
Reactions: 1 user
Upvote 0

SecureGSM

2 x 5D IV
Feb 26, 2017
2,360
1,231
I'm going to throw out a WAG here and say those two spots are dust on the surface of the test target?
Yup, PBD has worked this out already so. ISO 6400 quality is indeed Undistinguishable except that nasty moire / false colour effect in R5 shot image. Wondering how easy would be to get rid of that in post.
 
Upvote 0

Aussie shooter

https://brettguyphotography.picfair.com/
Dec 6, 2016
1,183
1,817
brettguyphotography.picfair.com
;)

I’m sure they are within processing distance of each other, and the differences so small they would be practically impossible to see at normal output sizes.

I think we can all agree (I’m sure we won’t) that the high iso performance difference is not big enough to base a purchasing decision on.

Regarding the other long running question, I see less than a stop difference between the R5 and the R, probably 1/2 to 2/3, so anybody hoping to see big gains there will be disappointed.
It ia the price that will determine which camera one purchases. Although tbh there are a lot of people who dont want to go through all the normalizing and downscaling etc. they just want to do basic PP and export a file. I am guessing for those people there will be a difference in high ISO noise performance but if they is what they want they will have to accept the drop in resolution .
 
Upvote 0
Regarding the other long running question, I see less than a stop difference between the R5 and the R, probably 1/2 to 2/3, so anybody hoping to see big gains there will be disappointed.

Sunday crow, served cold.

Certainly not enough high ISO improvement to justify (IMO) f/11 telephotos. Larger f/8 designs would have made more sense in light (pun) of what I see here.

I'm at least glad the R5 didn't do worse than the R6.
 
  • Like
Reactions: 1 users
Upvote 0
Jan 29, 2011
10,675
6,121
Sunday crow, served cold.

Certainly not enough high ISO improvement to justify (IMO) f/11 telephotos. Larger f/8 designs would have made more sense in light (pun) of what I see here.

I'm at least glad the R5 didn't do worse than the R6.
All is good, no bitterness intended. But it has illustrated the fact that Canon can’t bend the laws of physics....

And that this generation of sensors holds up to scrutiny the same as the last several generations. Gapless micro-lenses effectively leveled the field for sensor area vs pixel size and we are well into the physical limits of background noise in non cooled sensors tech to get much more out of high iso/low light performance.
 
  • Like
Reactions: 1 user
Upvote 0

H. Jones

Photojournalist
Aug 1, 2014
803
1,637
Interesting to me is that the 5DS vs R5 comparison looks like the R5 is rendering things sharper even despite the 5 megapixel less resolution. Maybe Canon is right that the new 16-point low pass filter is making things sharper despite the lower resolution, or it's just adding more sharpness to Jpegs. The 5DS R definitely looks a little bit mushier in this comparison of the two.

1595797366259.png
 
  • Like
Reactions: 1 user
Upvote 0
Jan 29, 2011
10,675
6,121
Interesting to me is that the 5DS vs R5 comparison looks like the R5 is rendering things sharper even despite the 5 megapixel less resolution. Maybe Canon is right that the new 16-point low pass filter is making things sharper despite the lower resolution, or it's just adding more sharpness to Jpegs. The 5DS R definitely looks a little bit mushier in this comparison of the two.

View attachment 191588
I don’t think you can tell much at all from jpegs, but the file size would indicate there is more detail to be had in the 5DS R.
 
Upvote 0

Sporgon

5% of gear used 95% of the time
CR Pro
Nov 11, 2012
4,719
1,537
Yorkshire, England
Agreed, and in itself is a demonstration of the similarity. Both managing to resolve such small and non contrasting spots as well as each other.
This is what I've found in the three years or so that I've been using the 5DS. 20mp on a FF is high resolution in its own right. To see the (small amount of) extra detail resolved by the 50mp sensor you have to view it at full size. Once you reduce the output size down to that of 20mp the extra amount of tiny detail has become too small to see anyway. So for me, the moral of the story is very high mp sensor for very large output or very heavy cropping. When I started using the 5DS I thought I was getting better colour definition and fidelity with the very high mp, but now, three years on I think this was due to making the classic mistake of not comparing sensors of the same generation.
 
Upvote 0
This is what I've found in the three years or so that I've been using the 5DS. 20mp on a FF is high resolution in its own right. To see the (small amount of) extra detail resolved by the 50mp sensor you have to view it at full size. Once you reduce the output size down to that of 20mp the extra amount of tiny detail has become too small to see anyway. So for me, the moral of the story is very high mp sensor for very large output or very heavy cropping. When I started using the 5DS I thought I was getting better colour definition and fidelity with the very high mp, but now, three years on I think this was due to making the classic mistake of not comparing sensors of the same generation.

5Ds(R) was amazing when it was good and terrifying when it was bad. About the most succinct review I could give of my time with it.
 
Upvote 0

Sporgon

5% of gear used 95% of the time
CR Pro
Nov 11, 2012
4,719
1,537
Yorkshire, England
Interesting that Canon elected to use an AA filter in the R5. I guess this could be because:
A) it's the right thing to do anyway for a Bayer Array sensor.
B) the 5DSr didn't outsell the 5DS
C) Canon are holding the AA free version back in order to sell you another camera later.

My money's on C :)
 
  • Like
Reactions: 1 user
Upvote 0

brad-man

Semi-Reactive Member
Jun 6, 2012
1,673
580
S Florida
Interesting that Canon elected to use an AA filter in the R5. I guess this could be because:
A) it's the right thing to do anyway for a Bayer Array sensor.
B) the 5DSr didn't outsell the 5DS
C) Canon are holding the AA free version back in order to sell you another camera later.

My money's on C :)
Yup. It will probably be the R5sR hi rez. In continuous mode it needs to be suspended in a bucket of ice.
 
Upvote 0