J
jay-A
Guest
Actually I cannot let it go because this is the definition of raw. But around here hard facts are not welcome, I know, do whatever pleases you, I’m OK with it.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Canon colors are back
- Thread starter boldcolors
- Start date
Lawn Lends
Senior Member
So raw interpreters just randomly decide on colors to display on your monitor when you open a RAW file?
J
jay-A
Guest
This is what color profiles are all about. Adobe color will normalise the output of all cameras in LR. Dxo can give one raw file the color characteristics of another body. And if you like the way Canon does color then DPP will give your raw files a similar look.
Last edited:
I see someone watched a Tony Northrup video and believed everything Tony said.
https://www.dpreview.com/reviews/im...1&x=0.27913882898051506&y=-0.7445660279561264
Funny, how you can look at four samples of RAW files out of four Canon cameras, all processed by DPReview in Lightroom, and still get four completely different results.
(I’ll probably never know exactly why, but I still say the Off Chip ADC has better colours than Canon’s new On Chip ADC. As far as I’m concerned the 5DS is “The Ultimate Canon”.)
Also remember there’s no such thing as a “Real RAW File”, everything is baked, RAW is just uncompressed so you can change values without losing quality.
If you want a totally pure and unaltered readout from the sensor you’re going to have to build your own ADC and hook it up yourself.
Last edited:
J
jay-A
Guest
These RAW files are represented using the 'camera matching' color profiles of ACR, which differ from camera to camera in order to mimick the jpeg output.
You might want to do the experiment at home using 'Adobe color' or 'Adobe standard' as common color profiles for all these cameras, and then report back on how different they now are.
edit: did it for you.
https://www.dpreview.com/forums/post/61893738
I don't think that one can honestly see a meaningful difference. But apparently here some manage to see one. Draw your own conclusions.
Last edited:
Steve Balcombe
Forum Pro
That's not correct, there's much more to RAW than being uncompressed.
No time to write a dissertation now, suggest you Google it.
boldcolors
Senior Member
That is their ambition however due to sensor CFA design and internal image processing (yes it is used for RAW too) the result is not the same between different cameras. It can be, but more often than not, there are differences.
DXO color match does not work.
The funny thing about color and CFA discussions is that many claim that anything can be done in post as a general rule. But when asked questions about how...then answers start to fade away.
Being able to do anything you want in post doesn't automatically mean that it can be done. It's like saying "I don't see the point of going to a fancy restaurant. I can buy the same ingredients myself and cook at home. It will taste exactly the same..."
Fujifilm is famous for their color science. An entire team from the film-era has spent a lot of time to tweak the special sensor CFA and togehter with very balanced profiles they are able to achieve that look.
So... tell me. Are they nuts? Did they do all this in vain because on DPR, all you have to do is listen to "hard facts" and push some sliders in LR. Or install DXO.
Nah...
Last edited:
J
jay-A
Guest
Adobe profiles the cameras with a common target when creating the Adobe color profile for each camera. This of course levels off the variations in the CFA design. Any remaining difference relates to sample variations in a given model.
Are you a dxo user? I am a user and for me it works. I tried once to level off the differences between a 5D4 and A7R3 set of raw files, and match the output to a 5D4 jpeg. It worked sufficiently nicely that I could no longer tell the difference in a blind test.
It's always interesting (and a strength per se) to have an in-camera rendering of color that is pleasing to many people. That can also be emulated, though it requires more time that most do not want to spend or do not have.
It does. It captures a specific 3D projection of the spectrum throwing away all the other infinitely many dimensions.
It is not about levels. Different CFAs have different metameric failures. The color profiles are trying to achieve some balance.
I am since version 4. They never worked and they have some smart people there.
Oh, this is easy. Just copy one of the files and paste it on the other. Or desaturate both. If you are trying to use their profile though, they do not even have different profiles for many Canon cameras and yet they have different colors. Also, the DXO colors of the 5D4, for example, do not match the DPP colors of the 5D4...
How far do you want to go with emulating pleasant colors? One can shoot with a B&W Leica and paint after that, right?
Leitz
Leading Member
boldcolors
Senior Member
Nope
J
jay-A
Guest
These are two RAWs from imaging resource. One 5D4, one A7RIII. Adobe color and default settings for both. I see almost the same colors, or if there are differences they're too minute to matter. if you shoot raw you don't care about what color your body produces. It's just data and whatever differences in the way the data is acquired is levelled off if the raw converter profiles are properly specified.
Last edited:
I do not see the same colors - the yellows are quite different. And this is just one scene. It tells you almost nothing. The biggest weakness of Sony is foliage greens and they are completely absent here.
Different data.
J
jay-A
Guest
Let's be honest.
Don't you think any of the minute differences that you claim you see cannot be corrected by adjusting to taste? Don't you think that the 1/10 of a stop exposure difference between the two shots, or slightly different 'as shot' white balance is more significant that the differences that you claim you see?
TN has shown that people tend to be more brand-loyal than color-discerning. And therefore this is a neverending discussion. Pardon me if I leave it here.
Last edited:
Of course, they can. A good painter does not need a camera. But we are not talking about that here.
No, for a very simple reason: several decades experience with photography.
I could not care less what TN says.
Leitz
Leading Member
If I upload ooc jpegs from different camera's (stripped exif) you can tell me the camera brand they were taken with?
--
Klaas
Last edited:
boldcolors
Senior Member
Of course not. Why would I?
boldcolors
Senior Member
This is a color-chart in a controlled lighting with strobes. That is the most neutral environment there is and with color patches and objects which is far away from "memory colors" in real life.
The true difference between camera sensors will be detected after a while when you work with your camera and the files and in all the hundereds of different scenarios you end up in as a professional and/or serious amatuer photographer.
Last edited:
My point is that the only practical difference between JPEG and RAW is that the RAW file hasn't been through the demosaic process and it contains full bit depth.
Similar threads
- User review
Keyboard shortcuts
- f
- Forum