Camera Raw/Lightroom: 5D Mark IV, wrong dcp profile from adobe

  • 10
  • Problem
  • Updated 7 months ago
  • (Edited)
Raw Files : 5D Mark IV
Software : Camera Raw 9.10.1
Issue : lights and colors rendering

While opening a raw in Digital Photo Professional (DPP), you can set up your space colors to sRGB or adobe98.

- When I open a raw taken by a 5D Mark III, in sRGB/sRGB, i have a normal rendering (lights / colors).
- If I set sRGB/adobe98 by changing the space colors, the light are more strong/contrasty, and the red are more saturated.
- If I open the 5D Mark III raw in Camera Raw, the rendering is like sRGB/sRGB in DPP.


Now, same operation but with 5D Mark IV raw files...

- When I open a raw taken by a 5D Mark IV, in sRGB/sRGB, I have a normal rendering (lights / colors)
- If I set sRGB / adobe98 byu changing the space colors, the light are more strong/contrasty, and the red are more saturated.
- If I open the 5D Mark IV raw in Camera Raw, the rendering is like sRGB/adobe98 in DPP, and that's NOT good.

To see differences between files and rendering I put jpeg files :
1) 5DMarkIII in DPP with sRGB/sRGB, 5DMarkIII with sRGB/adobe98, 5DMarkIII with Camera Raw
2) 5DMarkIV in DPP with sRGB/sRGB, 5DMarkIV with sRGB/adobe98, 5DMarkIV with Camera Raw


5DMARKIII :







5DMARKIV :





As you can see, the problem with the 5DMarkIV : to much contrast in camera raw, causing wrong colors (red saturation, light saturation with less saturation in the highlights etc.).

So, Is there a way for ADOBE, to make dcp profile camera standard to a sRGB look instead of an adobe98 look ?

All put all these capture and raw files on this link for ppl who want to check :

https://1drv.ms/f/s!Ak5vIZqwJCnXjK14AXupfcB8wEkclw


Crossing my finger for the giant Adobe hearing this post !

Thank you !
Photo of benjamin vingrief

benjamin vingrief

  • 30 Posts
  • 2 Reply Likes

Posted 2 years ago

  • 10
Photo of andrewthedigitaldog

andrewthedigitaldog

  • 6 Posts
  • 5 Reply Likes
Ambient light surrounding the display cannot be too low! It can be too high. Any ambient light striking the display affects perception of black. 
http://digitaldog.net/files/BlackisBack.pdf

Not many displays can natively target low cd/m^2 values lower than say 100-120 cd/m^2 so trying to drive them that low only introduces banding in the previews. The value is only correct when the display and print next to the display, ideally being illuminated match and everyone's mileage can and will vary. 
Ignore anyone who makes a specific suggestion about display backlight intensity unless they have taken print viewing conditions into account and can control them. A picture is worth 1000 words:
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
Thanks, Andrew.  I do think my prints match at 75 Luminance, and I tested with a Solux light.  I don't have a monitor booth set up like the picture, but it looks easy enough to create one.  I'm trying my monitor out at 100 L so I can be confident in my decision one way or the other after comparing it with prints edited on the brighter monitor.  I know I can use test prints from when I was at 75 L, but I want to see if my editing changes (gets darker).    

One question- in the image above, is that picture on the left supposed to be matching the monitors? Photographing monitors isn't probably accurate, so maybe that has something to do with it.  But if other people think that picture matches the monitor sufficiently, I'm WAY overthinking this stuff.  
(Edited)
Photo of Andrew Rodney

Andrew Rodney

  • 701 Posts
  • 140 Reply Likes
They do match NOT in the photo!
Photo of Todd Shaner

Todd Shaner, Champion

  • 1266 Posts
  • 419 Reply Likes
This is starting to go off-topic, but useful information for people who spend a lot of time editing images.

Low Ambient Room Light = Higher Potential for Eye Strain

This is true for both computer work and large screen TV viewing. Fortunately there is an easy solution that doesn't affect the perceived screen black level. It's called Bias Lighting:

https://www.howtogeek.com/213464/how-to-decrease-eye-fatigue-while-watching-tv-and-gaming-with-bias-...
Photo of Andrew Rodney

Andrew Rodney

  • 701 Posts
  • 140 Reply Likes
Ambient light is emitting from the display! Again, any other ambient light not emitted cannot be TOO LOW! It can be TOO HIGH. Less is always better.
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
For anyone else struggling with the 5d Mark iv camera profiles, I created some that are working great for all types of pictures-candids and portraits.  All were made in DPE.  There are 3 tone curve variations with and without dual illuminant color adjustments for tungsten and daylight (using a Color Checker Passport).  I included the recipes so you can tweak them as you'd like.  The legacy curve is from Todd's Legacy Mark iii Profile and is pretty subtle with it's contrast, but starts off lots better than Adobe Standard.  The contrast increases with the other two (adjust exposure as needed).  They don't all work for all my pictures, but there has been one that is best for all of the pictures I've gone through.  

Thanks again Todd for all your help!  I have never been as excited to click a camera profile as I am now!  My pictures are instantly pretty and will just get better with editing.  FINALLY!  

In case you want a preview first, I added 3 screenshots of a Lightroom image comparing it to Adobe Standard and each of the profiles.  All settings were the same, except a -10 Highlight adjustment.
https://www.dropbox.com/sh/o97ud0ffevoajdt/AAB3mnzszKRB9sSBR0cvDSsKa?dl=0
(Edited)
Photo of Todd Shaner

Todd Shaner, Champion

  • 1265 Posts
  • 419 Reply Likes
Laura, I finally had a chance to look at your custom DNG profiles. They look good and if it helps set a better "starting point" for your images it will reduce your editing time. The point is to not go too crazy switching between different profiles. As you work with them you'll learn what works best with specific image types. Bravo!
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
Without your Legacy curve, I wouldn't have anything good to go off of.  A good profile seems dependent on a good, general curve and yours is the one that withstood every single image.  Slight enough to be able to add more contrast to it when needed (and use as a base for other profiles) and yet interesting enough to stand alone on the most delicate of my newborn session pictures.   

I cannot thank you enough for putting in the time to help!!
Photo of benjamin vingrief

benjamin vingrief

  • 30 Posts
  • 2 Reply Likes
There is an image where the problem is obvious :

In order :

Image 1 : DPP, Camera Standard, view set in sRGB / sRGB
Image 2 : DPP, Camera Standard, view set in AdobeRGB / sRGB (contrast + red saturations broken)
Image 3 : CR, Camera Standard Profile (contrast + red saturations broken)

As you can see, dcp profiles from Adobe are completely mimicing the Camera Standard from DPP with AdobeRGB rendering, and NOT sRGB rendering.

See pictures to understand.

OneDrive link to DL jpg + RAW. Try it yourself in DPP / CR. That kind of picture shows exactly the mistakes. No problem with other camera (5DMarkIII Nikon camera etc).

https://1drv.ms/f/s!Ak5vIZqwJCnXjMkeEdVLr8aAL5VqbA

The picture from DPP adobeRGB and CR are exactly the same (only shape differ due to the lense correction in CR). Colors and contrasts are exactly the same. The problem is, her skin color is not like this... Seriously... 




Photo of benjamin vingrief

benjamin vingrief

  • 30 Posts
  • 2 Reply Likes
Adobe knows how to exactly copy the colors from DPP.
But they must take as reference the sRGB view in DPP, and not the AdobeRGB view.

That's the whole point of all of this.
(Edited)
Photo of Todd Shaner

Todd Shaner, Champion

  • 1164 Posts
  • 380 Reply Likes
Adobe knows how to exactly copy the colors from DPP.
But they must take as reference the sRGB view in DPP, and not the AdobeRGB view.
When the working color space is set to sRGB in DPP you are restricting the output to a smaller gamut compared to Adobe RGB and ProPhoto RGB used in LR/ACR. Virtually all print output and wide gamut displays have gamut outside the sRGB color space. Because of this using DPP or LR with an sRGB working color space is unacceptable for professional or serious amateur photographers.

You can accomplish the same objective with LR/ACR using its native controls or DPE to modify the camera profile's Tone Curve. NO restriction on the working color space is required. I'm confused–Is that not working for you?
Photo of Andrew Rodney

Andrew Rodney

  • 660 Posts
  • 126 Reply Likes
>>Virtually all print output and wide gamut displays have gamut outside the sRGB color space.

Yet no printer can produce all of sRGB. And there we go. 
Adobe doesn't know how to exactly copy the colors from DPP or any other proprietary converter. 
Photo of Laura Kent

Laura Kent

  • 98 Posts
  • 15 Reply Likes
I don't think Benjamin is suggesting that the color space actually changes.  And he isn't debating which color space is better.  Is it not a fact that most other camera profiles look most like the sRGB working space preview in DPP, however this one looks like the Adobe RGB rendition?  If they were able to copy the sRGB in 5d Mark iii, why is it wrong to hope they could stay consistent and try to copy the look of the sRGB preview for this camera too?  

Let's not pretend that without a decent profile you can achieve results consistently or as general as you can with a camera profile.  And certainly not efficiently.  Why do 5d Mark iv users have to settle for that?  With a good profile, your images all have a consistent base and the tone curve is free to make the minor adjustments needed for each image/session.  

So a good camera profile is really important.  The irony of making a custom camera profile in DPE is that you need a base profile from your camera.  Or you really are reinventing the wheel.  That isn't a problem for non-5d Mark iv camera users because if they want to modify their profile, they get to use one of their correct profiles as a base.  But we don't have any good options for that.  Adobe standard doesn't help since it has no contrast.  Camera Standard and Portrait are way too contrasty and the colors are still unusable even after the color chart/color checker adjustments.  So you have to come up with your own base tone curve.  Finding something that added just enough contrast and worked with all images without being too much felt near impossible.  Maybe it shouldn't have been, but I tried for a long time to try to match the curve in DPP and if I got close it NEVER worked consistently on my images like DPP's did. It's only because Todd share his tone curve coordinates from his Legacy 5d Mark iii profile that I was able to create my own, using his tone curve with adobe standard as my base profile.  I'm not sure how he came up with them, did he just try to match the tone, or did he have inside knowledge of the right curve?  Either way, his beat all my efforts.  In summary, without a correct base curve to go off of, making a custom profile isn't going to be successful.  If your camera profiles are correct, then modifying them to suit your needs IS more easily done.  And why is it OK to settle for not having access to a couple acceptable profiles for our camera?  Adjust as you want, but we should still have access to CORRECT camera profiles like everyone else.  Unless Adobe's making this a thing and going with the Adobe RGB look starting with our camera, I think they should be held accountable and adjust the profiles so they match the sRGB rendering they did in the past.  

Andrew, I can't speak for Benjamin, but I can promise you that when I am complaining about the 5d Mark iv profile not matching the DPP preview it is not because they aren't EXACT.  In fact they aren't even close.  They aren't like other camera profiles are when compared to their respective DPP preview.  I think we are fine with that.  We are not being picky, or dramatic.  The difference is HUGE and it's WRONG.  I am happy with the way Adobe has created other Canon profiles to closely match DPP in the past.  I'd settle for that.  Do you even see the difference that we are referring to (more problematic in photographs of people than of products/landscapes)?  I really do want to know if you see the problem or not.  5d Mark iv users are not getting the same useable profiles that other camera users are getting.  I think there's something about the principal behind this whole thing.  We should be able to modify camera profiles in DPP if we want to, not because we have to.  And if Adobe is wrong, they should fix it by adding corrected profiles.

https://www.dropbox.com/sh/u7yw5nh4f2gtcra/AACHR6OCZfqldR1k3LOOjDd6a?dl=0
(Edited)
Photo of Andrew Rodney

Andrew Rodney

  • 660 Posts
  • 126 Reply Likes
If you so demand a JPG rendering then shoot JPG! If you want to match DPP, use it instead. If you are unhappy with a generic .dcp profile, create a custom profile and edit it if necessary. Problem solved! If you are unhappy with the rendering, alter the rendering (there are a few tools and slides for that in LR!)!
Photo of benjamin vingrief

benjamin vingrief

  • 30 Posts
  • 2 Reply Likes
Andrew, go inside. Take a breath, but stop answering here.
You are not helping. Thank you. You don't want to understand simple things :

we are not idiots. we knows what we are talking about.

Thank you.
Photo of Todd Shaner

Todd Shaner, Champion

  • 1164 Posts
  • 380 Reply Likes
I think we're all on the same page concerning the DPP Standard Picture Style versus Adobe Camera Standard profile rendering "differences." So why does the discussion continue ad infinitum?

Adobe's objective with the Canon "Camera" named profiles (for all models) is to emulate the Standard picture style rendering of the "in-camera" JPEG file and DPP with Adobe RGB color space. From my testing Adobe has done a good job and the raw image file with Camera Standard applied matches the camera JPEG quite well with Adobe RGB camera setting. When DPP's working color space is set to Adobe RGB it also matches LR Camera Standard rendering very closely. This is true for all models I tested including 5D MKII and MKIII cameras.

So why is the 5D MKIV Camera Standard profile different? This is a decision the camera manufacturer (Canon) made with its newer model 5DS, 5DS R, 5D MKIV, and 1DX MKII camera models. Canon has changed the 'Standard' picture style tone curve for these cameras to include a negative area near the 0 Black Point level. They may have done this to improve the "perceived" shadow noise level....don't know?

Given the amount of work involved it's not likely Adobe will start providing two sets of "Camera" named profiles (DPP Adobe RGB and sRGB renderings) for Canon cameras. Especially since they haven't done so for any other camera makes. As already discussed here you can change the Tone Curve color rendering using the DNG Profile Editor (DPE) or dcpTools so that the Camera Standard profile behaves the same as with older camera models (i.e. 5D MKIII). For those with less technical skill you can also apply the corrections using the LR Tone Curve and Camera Calibration panels. The ColorChecker Passport is also good alternative for creating custom camera profiles both with their software and Adobe DPE.
Photo of benjamin vingrief

benjamin vingrief

  • 30 Posts
  • 2 Reply Likes
Todd, you pointed exaaactly the whole thing !
The thing is that for many years Adobe has taken the sRGB "look", but for only new Canon Camera, they took the AdobeRGB "look". And yeah you are right, they have done it Perfectly.
Problem is : it looks fake, colors are completely wrong. Wrong in the AdobeRGB view in DPP AND wrong in the CR view so.
If you see my pictures, it's totally wrong with girls with tanned skin.
When I say it's wrong, I don't say they "copy" in wrong way, because like I said they have copied perfectly the adobeRGB view from DPP.

The main problem : if you try to manipulate the colors in Camera Raw with all the settings, you are wasting your time with that. If you are a pro photographer and retoucher, and if you shoot a lot and retouch a lot, you can't edit as hell all these like this. And the worst is : whatever you try, you are not accurate.
Plus, it's very hard to have an "AdobeRGB look" and transform it to a "sRGB look".
As you understand, to much things to manipulate :
- contrast because of the differents curves between sRGB / adobeRGB (the sliders "contrast" can't manage this)
- colors rendering (Red are too saturated)

You can't do a "general" settings, because when you change the lights during the shoots, the skin colours rendering for example changes.


With the 5DMarkIII i have never been dependant, with my nikon cameras, never.

And for me doing two dcp profile like "low contrasts" and "high contrasts" is not a nightmare for them.
Or at least, I just hope Adobe could correct the last profile to put them in sRGB, but the problem is : they must admit they have taken the adobeRGB instead of sRGB...
And they will never do that ! 

For us, we have to manage with different tools, code lines, tries and tries. But they got the tools to do that in 5min haha

The thing is, for example, Capture One has taken that in account in their "camera settings", you can choose between different rendering, and they are aware of this.

Curve with shadow control, Curve with High Contrast (oh, AdobeRGB to be specific), Curve with Standard Contrast (oh, sRGB !), and... Linear.

Dxo does the same thing.

Well...

To be honest, Adobe will never correct that. And Canon doesn't care more with this problem.
These forums have ZERO impact, and the posts are not taken in serious.
No engineers will read that, no one in Adobe/Canon will transfer this info.
Photo of Andrew Rodney

Andrew Rodney

  • 660 Posts
  • 126 Reply Likes
>>The thing is that for many years Adobe has taken the sRGB "look", but for only new Canon Camera, they took the AdobeRGB "look".


There is no 'sRGB look'. The facts are, sRGB is an encoding color space based on a CRT display (theoretical with specific aims: TRC, white point and chromaticity values that define the color gamut). 

The camera doesn't produce sRGB. It encodes raw data, rendered by some process (human or non human) and places that into the sRGB container.

Without a pixel with some numeric value and a defined color space, the color space is an empty container. 
Photo of Andrew Rodney

Andrew Rodney

  • 660 Posts
  • 126 Reply Likes
>>we are not idiots. we knows what we are talking about.

The reason there's so much ignorance on the subject of color management, is that those who have it are so eager to regularly share it! - The Digital Dog
(Edited)
Photo of Todd Shaner

Todd Shaner, Champion

  • 1164 Posts
  • 380 Reply Likes
benjamin vingrief said: To be honest, Adobe will never correct that. And Canon doesn't care more with this problem.
These forums have ZERO impact, and the posts are not taken in serious.
No engineers will read that, no one in Adobe/Canon will transfer this info.
Well that's not actually the case! See the below post concerning a similar issue with the Canon 5DS Adobe Standard profile. They have created an Adobe Standard v2 camera profile that corrects the issue.

https://forums.adobe.com/message/7649141#7649141
(Edited)
Photo of Andrew Rodney

Andrew Rodney

  • 660 Posts
  • 126 Reply Likes
>>Andrew, so you are saying in DPP if you use sRGB working space you don’t see a change in an image when you switch it to Adobe rgb?

Depends on the image and it's gamut. 
Photo of Laura Kent

Laura Kent

  • 98 Posts
  • 15 Reply Likes
Perhaps in Photoshop, but it doesn't matter in DPP- the entire image changes.  Have you actually tried comparing one image with both working spaces in DPP?   Can you try it with the files I provided in the link above?  Change it from the tool pallete and not under Adjustment tab so you can see the results instantly.  There's a whole different curve, or tone, or something. It's not just like the colors are different.  
Photo of Laura Kent

Laura Kent

  • 98 Posts
  • 15 Reply Likes
Can you provide a screenshot where the renderings are the SAME?  
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
Andrew, so you are saying in DPP if you use sRGB working space you don’t see a change in an image when you switch it to Adobe rgb? I’m not asking for color theory, I want to know if you’ve actually DONE it. There is indeed a different look. If you haven’t done it, I don’t know why you’re arguing with us.

Todd, I disagree that Adobe has copied the Adobe rgb look for other cameras. I just took similar shots (same lighting, exposure, lens, and same bored kid in the same spot) with the 5d Mark ii and iv and added new screenshots comparing them to the DPP sRGB and Adobe rgb rendering. I included the color checker passport so you can see more easily which DPP rendering the contrast/blacks/saturation resemble with the Camera standard profile in Lightroom. (The screenshots are sRGB). I think you’ll conclude that-at least the 5d Mark ii, but it sounds like most past cameras- resemble the sRGB working space in DPP while the 5d Mark iv clearly resembles the contrasty Adobe rgb rendering.

If you look at them and agree, then are you sure that adobe’s Intention WAS to copy the Adobe RGB look? This is the part that is frustrating me. I want to know if they intentionally did it or not.

That’s why we are belaboring this. Because I think that we think Adobe DIDNT mean to copy that rendering and that it is a mistake. Someone with Andrew’s knowledge of sRGB and Adobe RRB who don’t think that there IS a difference between the two in DPP. They just sat down with the “better” working space and started matching. If they intended to match the Adobe RGB look, why did they start with this camera?

Why does it matter if I’ve figured my way out of this and made a profile? For everyone who can’t or haven’t made a custom profile or the poor souls who don’t know any better, and are trusting Adobe and working extra hard to make their pictures look decent. Adobe should change it. They can do it better than us.

If that isn’t happening, I might be able to settle for people acknowledging it’s wrong. And then I can stop trying to prove it. I understand how you say we can adapt to it, and put a bandaid over the problem, but nobody without the problem is admitting it’s WRONG. People with other cameras aren’t in the same-I have to make a custom camera profile in order to use Lightroom like everyone else gets to-boat, and I don’t think that makes sense. They can make their profile better, we have to make it useable.

Basically what we have is:
Adobe lowered the contrast in Adobe standard. I’m fine with that. It’s a good, boring base when all else fails. And then they increased the contrast in the camera standard profile (intentionally or not, is the question I want to know) so there’s no middle ground.

Link to screenshots:
https://www.dropbox.com/sh/6hkm3wineeoiogm/AADbid0DSyux7r-jVfEWPmlDa?dl=0
(Edited)
Photo of Todd Shaner

Todd Shaner, Champion

  • 1265 Posts
  • 419 Reply Likes

Laura Kent said: Todd, I disagree that Adobe has copied the Adobe rgb look for other cameras. I just took similar shots (same lighting, exposure, lens, and same bored kid in the same spot) with the 5d Mark ii and iv and added new screenshots comparing them to the DPP sRGB and Adobe rgb rendering. I included the color checker passport so you can see more easily which DPP rendering the contrast/blacks/saturation resemble with the Camera standard profile in Lightroom. (The screenshots are sRGB). I think you’ll conclude that-at least the 5d Mark ii, but it sounds like most past cameras- resemble the sRGB working space in DPP while the 5d Mark iv clearly resembles the contrasty Adobe rgb rendering.
Laura, I think you missed this paragraph in my previoyus reply:

So why is the 5D MKIV Camera Standard profile different? This is a decision the camera manufacturer (Canon) made with its newer model 5DS, 5DS R, 5D MKIV, and 1DX MKII camera models. Canon has changed the 'Standard' picture style tone curve for these cameras to include a negative area near the 0 Black Point level. They may have done this to improve the "perceived" shadow noise level....don't know?

The earlier model Canon cameras such as 5D MKII and MKIII are not affected by this issue.
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
I didn't miss that.  It's beside the point.  You said:
<<Adobe's objective with the Canon "Camera" named profiles (for all models) is to emulate the Standard picture style rendering of the "in-camera" JPEG file and DPP with Adobe RGB color space

And I disagree.  I am trying to prove that Adobe has been copying the sRGB color space in DPP for past cameras- in my examples I showed the 5d Mark ii.  I think you can see, especially in the contrast & black levels, that I am correct.  UNTIL the 5d Mark iv.  The 5d Mark iv samples show without a doubt, and I know you have agreed already, that Adobe matched the camera standard profile to ADOBE RGB instead.  

If I'm right, and Adobe used to copy sRGB working space preview in DPP until 5d Mark iv, it seems logical that they SHOULD have copied it for 5d Mark iv also.  And had they done so, the contrast issue as you described above wouldn't exist.  It'd be irrelevant.  Because the contrast in the sRGB working space previews in DPP are fine.

https://www.dropbox.com/sh/6hkm3wineeoiogm/AADbid0DSyux7r-jVfEWPmlDa?dl=0
(Edited)
Photo of Todd Shaner

Todd Shaner, Champion

  • 1265 Posts
  • 419 Reply Likes
I went back and downloaded 5D MKIII and MKIV CR2 files of a 2D test chart that don't have significant differences in lighting (i.e. 2D). They are well controlled test shots that allow making critical measurements. Here are the files for your own comparison of DPP vesus LR/ACR rendering:

http://www.imaging-resource.com/PRODS/canon-5d-mkiii/E5D3hVFAI000100.CR2.HTM

http://www.imaging-resource.com/PRODS/canon-5d-iv/E5D4hVFAI000100.CR2.HTM

I am trying to prove that Adobe has been copying the sRGB color space in DPP for past cameras
I processed the two above CR2 files in DPP with Adobe RGB and sRGB working and TIFF output color space and then compared them inside LR with the CR2 files with Camera Standard. Here are the results:

E5D3hVFAI000100.CR2
DPP Adobe RGB TIFF
The LR Camera Standard rendering of the color patches very closely matches the DPP Adobe RGB TIFF.
The LR Camera Standard grayscale patches near the 0 Black Point level have higher values indicating a very slightly lower contrast Tone Curve than DPP Adobe Standard rendering. Pretty close match!

DPP sRGB TIFF
The DPP sRGB TIFF has some color patches that are very noticeably lower in saturation and/or Luminace compared to the LR Camera Standard rendering. This is not even close to a match!
The LR Camera Standard grayscale patches very closely matches the DPP Adobe RGB TIFF. The Tone Curves are near identical.

E5D4hVFAI000100.CR2
DPP Adobe RGB TIFF
The LR Camera Standard rendering of the color and grayscale patches very closely matches the DPP Adobe RGB TIFF.

DPP sRGB TIFF
The DPP Adobe RGB TIFF has some color patches that are very noticeably lower in saturation and/or Luminace compared to the LR Camera Standard rendering. This is not even close to a match!
The DPP sRGB TIFF grayscale patches  near the 0 Black Point level have higher values indicating a slightly lower contrast Tone Curve than DPP Adobe RGB rendering.

Draw your own conclusions as to what Adobe and Canon are doing.

Executive Summary
The Canon 5D MKIV Camera Standard profile Base Tone Curve differs from the 5D MKIII, but color rendering is the same. The 5D MKIV Camera Standard profile edited using dcpTool to replace just the Tone Curve values with that from the 5D MKIII Camera Standard profile provides near identical rendering with both cameras. You can download the profile here:

https://www.dropbox.com/s/jlkvizsbyn1y8xy/Canon%20EOS%205D%20Mark%20IV%20Camera%20Standard%20With%20...

No sense beating a dead horse since Adobe already sent it to the glue factory!
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
Interesting! I’m out of town for a week but am curious to check when I get home.

Did you draw the same conclusions with my two files (same light, exposure, lens, pose, but mark ii and iv cameras?). And to make sure, you’re not using the wide gamut sRGB in DPP, right? Because that looks the same as Adobe rgb. Just checking because with my test, I don’t get the results you got, but am glad you tested. Maybe I see the difference more drastically going from the mark ii to the Mark iv, skipping the mark iii. It doesn’t explain what others are noticing who have the 5d mark iii, but I guess I need to get over it.

One more question about profiles if I haven’t annoyed you too much
Photo of Andrew Rodney

Andrew Rodney

  • 701 Posts
  • 140 Reply Likes
Laura, do you own a wide gamut display?
Photo of benjamin vingrief

benjamin vingrief

  • 30 Posts
  • 2 Reply Likes
Hello Todd,

The problem with testing with sharts is that the light with flash for example is flashing on a "flat scene". If you do the same thing with a scene with a human, with a portrait, you can see difference clearly. That's what i was talking about in a Canon center.
I sent them my camera and raw files to show them the problem, and they can't test an a "non-flat" scene. They do the same thing : charts. But, it is not a relevant test.

And that's exactly why all the camera tests online are limited in their analyzes, they all do the same mistakes : flat scene

With your DCP Profile with the 5DMarkIII curve, I already done that before and i sent many links about that.

The problem is :

Because Adobe have copied the "adobeRGB" rendering from DPP, the saturations are complety to saturated.
So yeah, with this curve the lights are more nice. But it's not optimal (curve should be calculated for and only for the current camera, and the exposure too).

So it's a better thing, but not the one Adobe could make with their tools.

BTW thank you for your time
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
Andrew,
Honestly, I’m not sure. I have a 27” 5K retina iMac which is P3 color. (I don’t think it always was, there was an update that changed it at some point). I thought P3 color gamut was considered wide gamut, but all I could find that clearly stated were wide gamut monitors were late 2015 models, and mine is late 2014. Why do you ask?
Photo of Todd Shaner

Todd Shaner, Champion

  • 1265 Posts
  • 419 Reply Likes
Laura Kent said: Did you draw the same conclusions with my two files (same light, exposure, lens, pose, but mark ii and iv cameras?). And to make sure, you’re not using the wide gamut sRGB in DPP, right? Because that looks the same as Adobe rgb. Just checking because with my test, I don’t get the results you got, but am glad you tested.
You have a couple of issues:

1) The file naming you used 2018-01-14_Canon EOS 5D Mark II_107_EF70-200mm f-2.8L IS II USM_1⁄160 sec at ƒ - 10_ cam sRGB.CR2 caused EXIF tool to balk and show NO FILE METADATA. I suspect the shutter speed fraction 1/160 caused the issue. Do yourself a favor and don't use illegal characters in filenames..and keep them short! Use  only alphanumeric, underscore _ and dash - characters.

2) After renaming the CR2 files I was able to see the camera MakerNotes. Your 2nd issue is that you have 'Auto Lighting Optimizer' set to Strong in both CR2 files (MKII and MKIV). This applies tonal changes to the in-camera JPEG file, but not to the raw file data. The raw file MakerNotes indicates the setting used and DPP applies it when editing the file. Since LR cannot apply the Auto Lighting Optimizer function it's rendering will be different than DPP's regardless of camera profile issues. To make an apples-to-apples comparison of the in-camera JPEG to LR's Camera Standard rendering you need to change all of your Canon camera's Custom Function Settings for 'Auto Lighting Optimizer' to DISABLE. Ditto for 'Highlight Tone Priority' and make sure the Picture Style controls (Contrast, Saturation, Color Tone) are at their default settings. This is what Adobe uses to determine Camera Standard rendering. Obviously any change you make to these settings will cause LR's rendering to be different than the camera JPEG.

Onward!
Photo of Todd Shaner

Todd Shaner, Champion

  • 1265 Posts
  • 419 Reply Likes
Apple P3 Retina displays are wide-gamut, but shifted slightly compared to Adobe RGB:

http://www.astramael.com/

I'm using a display that has 99% Adobe RGB gamut. The images we've been testing have very little "visible" gamut outside of sRGB, so it shouldn't be an issue.
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
I need to look into the auto lighting optimizer because I have it off and it shows that in my DPP info. Same with highlight tone priority. I find it makes me think the image is brighter than it actually is in Lightroom (since it doesn’t read that metadata). Not sure why it would be different for you...
(Edited)
Photo of Laura Kent

Laura Kent

  • 99 Posts
  • 17 Reply Likes
I spoke too soon. I am sure about the settings for 5d Mark iv, but can’t promise for 5d Mark ii (if it’s even an option). Maybe that will do it, I’ll check at home.
Photo of Todd Shaner

Todd Shaner, Champion

  • 1265 Posts
  • 419 Reply Likes
Yep, 5D MKII is set to Strong and MKIII to Disable. You can manually disable ANY CR2 file camera setting inside DPP, but the in-camera JPEG file has those settings "baked in."

From inside DPP hit CTRL/CMD + I keys.

Photo of Ievgen Morskyi

Ievgen Morskyi

  • 5 Posts
  • 1 Reply Like
Thread bump. Adobe, please provide a fix in the form of V2 profiles as you did for 5Ds(R).
It's strange why so few ME TOOs... does every 5dm4 owner create custom profiles?
(Edited)
Photo of Steve Sprengel

Steve Sprengel, Champion

  • 2668 Posts
  • 343 Reply Likes
I am not new to Lightroom but there may be a new capability I am unaware of: 

Your screenshot doesn't appear to be the import panel, so how would you see the camera-embedded preview of a raw file in LR beyond the Import panel?  LR immediately replaces the preview with the Adobe rendering so my guess is that the middle image is the natural-looking Adobe-rendering using the default camera profile. the same look as the lefthand "Before" image in each of the three side-by-sides in my previous screenshot.

Now if that is somehow the camera-embedded preview, still, the CR2's thumbnail/preview may be in the AdobeRGB colorspace which when viewed in the assumed colorspace of sRGB will be less saturated.  This could be so the camera histogram and clipping warnings are displayed according to the AdobeRGB JPG rendering.
(Edited)
Photo of Todd Shaner

Todd Shaner, Champion

  • 1266 Posts
  • 419 Reply Likes
I've converted this to a reply since it's not relevent to Amanda Byler's threadabove, which unfortunately pushed it here.

"Your screenshot doesn't appear to be the import panel, so how would you see the camera-embedded preview of a raw file in LR beyond the Import panel?  LR immediately replaces the preview with the Adobe rendering so my guess is that the middle image is the natural-looking Adobe-rendering using the default camera profile."

Steve I'm using the new Embedded & Sidecar Import setting and workflow introduced in LR Classic 7.0. More details here:

http://blogs.adobe.com/sunil/2017/10/18/embedded-preview-workflow-in-lightroom-classic/

Canon and Nikon cameras embed a full-size preview and when with this import setting, which LR uses for the Standard and 1:1 Preview. When you switch to the Develop module and apply any change to the settings LR will build a new Standard preview and 1:1 preview with these settings. At the above link:

"Latest Nikon and Canon cameras have full size embedded previews. Means, the camera-generated embedded previews can be used for culling. Sony, Olympus, and Fuji cameras don’t have full sized embedded previews. In this case, Lightroom checks if the embedded preview size is 50% of that of the raw image. If not, the standard preview is created instead of reading the embedded preview. Create a jpg sidecar image along with the raw. If the image has the sidecar file, it’s being read for creating the preview in the Library module."

"Now if that is somehow the camera-embedded preview, still, the CR2's thumbnail/preview may be in the AdobeRGB colorspace which when viewed in the assumed colorspace of sRGB will be less saturated.  This could be so the camera histogram and clipping warnings are displayed according to the AdobeRGB JPG rendering."

It appears you are correct. This image file has Adobe RGB color space selected in-camera and LR is displaying the image in sRGB color space. Again, this is because LR doesn't read the proprietary MakerNotes, which is where the color space setting is defined. Good catch!