The Canon EOS R5 scores Canon’s best sensor score at DxO

nikkito

Argentine Photojournalist
haha yes I did. With my EOS R I was happy to use Adobe Camera Raw but with the R5 I felt DPP was the only choice. I only like to make minimal changes to RAW files and Adobe was proving too finicky to get results I liked. Luckily I only process 20-30 photos per shoot which makes DPP workable.

I do. My Instagram is @Chris.Chapterten how about yourself?

Hey Chris, i followed you already. Pretty cool photos! Lucky dude :-D

mine is @nicolaszonvi

cheers :)
 
  • Like
Reactions: Chris.Chapterten
Feb 15, 2020
532
364
That is weird; is it on Windows or Mac?
I mostly use it on the Mac side and I've never had any problems, except for the fact that rendering the 100% preview can be really slow.
It never crashed though...
This is on Windows. Though my computer is 7 years old, so that may have something to do with it. When it crashes the window stays open but I can’t open or edit any files. Then once I force the program to close all of my previous edits are wiped clean.
 

privatebydesign

I post too Much on Here!!
CR Pro
Jan 29, 2011
10,502
5,745
In DXO, all cameras are normalized to 8 MP for comparison.
Only for the 'Print' option in the Dynamic Range measurement. If you look at the 'Screen' measurement it is not normalized.

Screen Shot 2020-12-18 at 15.22.49.png
Screen Shot 2020-12-18 at 15.25.09.png
 
  • Like
Reactions: usern4cr

Fischer

EOS RP
Mar 17, 2020
301
216
Thanks for your answer. Yes. I have it calibrated with Spyder Pro or something. A new one. Nevertheless, the camera matching profiles do make a difference. When I have many photos to retouch in want the starting point to be good already. The Spyder helps me to be sure that the colours I see are the right colours. But those right colours are not that pretty without the Camera matching profiles. You know what I mean?

I'm interested in this thing you say "it makes it very easy to color correct your camera processing profile". What colour profile do you use in lightroom? And what space colour do you use in Photoshop?

Cheers mate
Not sure if you entirely understood what I meant.

In LR and other RAW prcessing software there is a specific color processing file for each camera. With your SpyderPro and a pro color chart you can develop your own color processing file - to your taste - that exactly fits your specific camera sensor. This will give you superior results i.e. when it comes to shadow lifting and avoiding blown highlights. I personally do not understand why not everyone interested in their colors does this.

In LR - after you create your camera specific color profile - you can go to /Develop/Calibration and choose your cutom made color profile from the drop down menu there.

From Adobe:
"Install a color profile

Color profiles are often installed when a device is added to your system. The accuracy of these profiles (often called generic profiles or canned profiles) varies from manufacturer to manufacturer. You can also obtain profiles from a custom profile service, download profiles from the web, or create custom profiles using professional profiling equipment.
1. To install a color profile, copy it to one of the following locations:
  • Windows 7, 8: \Windows\system32\spool\drivers\color
  • Mac OS: /Library/ColorSync/Profiles or /Users/[user name]/Library/ColorSync/Profiles
    Tip: By default on Mac OS 10.7 (Lion), the user Library folder is hidden. If you don’t see it in the Finder, press Option and click the Go menu. Then, choose Library. See Access hidden user library files | Mac OS 10.7 and later.
2. Restart Lightroom Classic."

Using RGB all the way (and so should you).
 

Fischer

EOS RP
Mar 17, 2020
301
216
Only for the 'Print' option in the Dynamic Range measurement. If you look at the 'Screen' measurement it is not normalized.

View attachment 194620 View attachment 194622
Not sure how you could normalise screen viewing. My understanding is that high resolution TV screens and monitors do not actually show the total amount of pixels their specs are rated to, even if they have them, but interpolate their final results. As I remember it it's something like 1,080 being the current highest "real" output with the rest done via interpolation of the input signal.
 

usern4cr

R5
CR Pro
Sep 2, 2018
1,210
2,073
Kentucky, USA
Use a color calibration device such as the Spyder or Disply1. They cost a little. Last for years, help your monitor show colors correctly and makes it very easy to color correct your camera processing profile so it shows "true" colors in whatever program you prefer to use for RAW editing such as Adobe LR etc. You can also tweak the colors to your delight. And you will certainly get much better dynamic range results with less blown highlights and less noisy shadows.
Thanks for the post, Fischer. I've never had one of these devices. Do you have a suggestion for which one you think is best to buy at the moment?
 

Fischer

EOS RP
Mar 17, 2020
301
216
Thanks for the post, Fischer. I've never had one of these devices. Do you have a suggestion for which one you think is best to buy at the moment?
I use display one but have not reason to think there is any real difference between the leading brands. Get the one that fits your purse. And the basic ones are often good enough for photography. What you want check is that the model you buy has software you find easy to work with. You can find videos on Youtube that show how they work.

There's a pretty dramatic example of how important this can be here: https://www.dpreview.com/forums/post/57062274 that shows how big a difference it can make having your own profile instead of just relying on Adobe's (or Canon's DPP profile for that matter). What is critical here is that there is no way to post-process the Adobe profile to save the blown highlights or recover the shadow details - it entirely depends on using a custom RAW color profile for your camera as your first step.
 
  • Like
Reactions: usern4cr

Dragon

EF 800L
May 29, 2019
504
494
Only for the 'Print' option in the Dynamic Range measurement. If you look at the 'Screen' measurement it is not normalized.

View attachment 194620 View attachment 194622
True, but the print number is the one they put forward and the one they use in their overall scoring. The Screen numbers are useful to see the difference in pixel level noise, but otherwise not all that meaningful, since the quality of the final total image (or equal angle of view area for a crop) is the objective. I have also found that if you do some smart noise reduction on the higher density image before normalizing the improvement in DR is even greater than simply averaging the raw pixels. In any case, there is no doubt that the R6 is impressive at higher ISOs. It is interesting that the R6's gain over the R5 at higher ISO's is not confirmed by the Photons to Photos data (which would favor the R6 even less if corrected for actual ISO based on DXO's measurements). P to P numbers are also normalized similar to the DXO print numbers, but the spread between the two methods shows almost a full stop of difference at higher ISOs. I have an R5 and am regularly impressed by how usable ISO 12800 images are.
 
Last edited:

privatebydesign

I post too Much on Here!!
CR Pro
Jan 29, 2011
10,502
5,745
True, but the print number is the one they put forward and the one they use in their overall scoring. The Screen numbers are useful to see the difference in pixel level noise, but otherwise not all that meaningful, since the quality of the final total image (or equal angle of view area for a crop) is the objective. I have also found that if you do some smart noise reduction on the higher density image before normalizing the improvement in DR is even greater than simply averaging the raw pixels. In any case, there is no doubt that the R6 is impressive at higher ISOs. It is interesting that the R6's gain over the R5 at higher ISO's is not confirmed by the Photons to Photos data (which would favor the R6 even less if corrected for actual ISO based on DXO's measurements). P to P numbers are also normalized similar to the DXO print numbers, but the spread between the two methods shows almost a full stop of difference at higher ISOs. I have an R5 and am regularly impressed by how usable ISO 12800 images are.

Who cares about the DxO overall score? Nobody knows how they derive it or what figures they use to do it, given that surely nobody pays any attention to that bullshit number derived from a secret source of other made up numbers. A couple of people have gotten close to reverse engineering their Overall score algorithm but they never seem to fit all the examples.

I was merely pointing out that DxO do supply normalized and non-normalized figures. What people actually derive from them is up to them, if I am ever interested I go to DPReview and download the relevant studio RAW files and play with them. The I know what I can realistically do with the files and if they are useful for my needs.
 

LSXPhotog

Automotive, Motorsports, Commerical, & Real Estate
CR Pro
Apr 2, 2015
534
548
www.diossiphotography.com
DxO is truly a joke with it's monopoly on camera sensor scoring. If they purely collected data and allowed users to sort through it, I would like them a lot more. But then they go ahead and project with a truly arbitrary overall rating score on lenses and camera sensors.

DxO...the same company that claimed the exact same lens will transmit more light when mounted on a different camera. DxO...the same company that reduces the sharpness rating on a lens if it happens to stop down to a larger aperture that is more heavily impacted by diffraction. They will seriously lower a lens' sharpness score if a sharp lens can shoot at f/32 while a softer lens that can only stop down to f/22 will benefit - they claim it has a lower sharpness score, contrary to their own data. DxO...the same company that will score a lens based on the performance of the camera sensor behind it rather than the performance of the lens itself.

Their declaration of ISO performance being called "Sports" is a demonstration of complete ignorance to camera use.

DxO is also unrealistically obsessive over the weight it places on "color depth".

Yeah, just not a fan of them at all and I wish more people in positions of vocal power would start speaking out against them to dilute their credibility. They have a lot of influence in the camera market I don't feel they deserve at all.
 
Feb 15, 2020
532
364
DxO is truly a joke with it's monopoly on camera sensor scoring. If they purely collected data and allowed users to sort through it, I would like them a lot more. But then they go ahead and project with a truly arbitrary overall rating score on lenses and camera sensors.

DxO...the same company that claimed the exact same lens will transmit more light when mounted on a different camera. DxO...the same company that reduces the sharpness rating on a lens if it happens to stop down to a larger aperture that is more heavily impacted by diffraction. They will seriously lower a lens' sharpness score if a sharp lens can shoot at f/32 while a softer lens that can only stop down to f/22 will benefit - they claim it has a lower sharpness score, contrary to their own data. DxO...the same company that will score a lens based on the performance of the camera sensor behind it rather than the performance of the lens itself.

Their declaration of ISO performance being called "Sports" is a demonstration of complete ignorance to camera use.

DxO is also unrealistically obsessive over the weight it places on "color depth".

Yeah, just not a fan of them at all and I wish more people in positions of vocal power would start speaking out against them to dilute their credibility. They have a lot of influence in the camera market I don't feel they deserve at all.
100% agree!
 
  • Love
  • Like
Reactions: LSXPhotog and siav8

Dragon

EF 800L
May 29, 2019
504
494
DxO is truly a joke with it's monopoly on camera sensor scoring. If they purely collected data and allowed users to sort through it, I would like them a lot more. But then they go ahead and project with a truly arbitrary overall rating score on lenses and camera sensors.

DxO...the same company that claimed the exact same lens will transmit more light when mounted on a different camera. DxO...the same company that reduces the sharpness rating on a lens if it happens to stop down to a larger aperture that is more heavily impacted by diffraction. They will seriously lower a lens' sharpness score if a sharp lens can shoot at f/32 while a softer lens that can only stop down to f/22 will benefit - they claim it has a lower sharpness score, contrary to their own data. DxO...the same company that will score a lens based on the performance of the camera sensor behind it rather than the performance of the lens itself.

Their declaration of ISO performance being called "Sports" is a demonstration of complete ignorance to camera use.

DxO is also unrealistically obsessive over the weight it places on "color depth".

Yeah, just not a fan of them at all and I wish more people in positions of vocal power would start speaking out against them to dilute their credibility. They have a lot of influence in the camera market I don't feel they deserve at all.
Agreed. The raw data is quite useful, but the whole scoring system is a mess and to add insult to injury, the methodology for picking the numbers is thoroughly obfuscated. Sadly, whether the case or not, it looks like a way to take bribes for high scores.
 
  • Love
  • Like
Reactions: LSXPhotog and siav8

YuengLinger

EOS R5
CR Pro
Dec 20, 2012
3,516
1,961
USA
I can't compare to Nikon, Panasonic, or Sony, but I can say that the R5 produces images that not only look great RAW, but are very easy to work with--even at higher ISO's. (Shadow and highlight recovery, noise clean-up, sharpening, colors.)

What surprises me above all, and should have weighed huge in any fair assessment, is the ability to use 100% crops from the R5. I've had the 5DIII and 5DIV--they do not come close. Can the brands with higher scores do this too? If so, great for them.

Here is an example of approx. 100% cropped, ISO 640, and, imo, it is perfectly usable online. They print fine for 4x6 and 5x7. I've done so. What an amazing "cheat" this can be when we don't frame optimally!

100% Crop R5.jpg


The R5 has surpassed my wildest expectations, and I've only just begun to explore its potential.
 

usern4cr

R5
CR Pro
Sep 2, 2018
1,210
2,073
Kentucky, USA
I can't compare to Nikon, Panasonic, or Sony, but I can say that the R5 produces images that not only look great RAW, but are very easy to work with--even at higher ISO's. (Shadow and highlight recovery, noise clean-up, sharpening, colors.)

What surprises me above all, and should have weighed huge in any fair assessment, is the ability to use 100% crops from the R5. I've had the 5DIII and 5DIV--they do not come close. Can the brands with higher scores do this too? If so, great for them.

Here is an example of approx. 100% cropped, ISO 640, and, imo, it is perfectly usable online. They print fine for 4x6 and 5x7. I've done so. What an amazing "cheat" this can be when we don't frame optimally!

View attachment 194798

The R5 has surpassed my wildest expectations, and I've only just begun to explore its potential.
I'll second your praise of the R5 regarding shadow recovery, etc., even with cRaw which is now my only way to use it as I need the smaller size files. It's really great!

My previous camera was an EM1_II, and while it had a lot of benefits it had a poor shadow recovery (IMHO) which was only saved (for me) by the use of PL3 and their "prime" noise reduction. Now I get to have an already stellar R5 file and PL4 "deep prime" noise reduction for vastly better performance. It's a great day to be a photographer! :)