Different colors between Capture One and Adobe Camera Raw
I'm experiencing different colors between Capture One and Adobe Camera Raw (Photoshop CS6). I have my camera tethered and the image that shows up in Capture One looks very different from when I open it in Adobe Camera Raw. Please see image (Capture One on top and Camera Raw below). Thanks!
http://s30.postimg.org/rn5fxeutr/Screen_Shot_2015_07_04_at_11_32_22_PM.png
http://s30.postimg.org/rn5fxeutr/Screen_Shot_2015_07_04_at_11_32_22_PM.png
0
-
The only thing we can say from this is that CO and ACR reads the camera set WB slightly different
Can you quantify the difference, and are there reasons for it? I thought white balance was white balance, that the same colour temp would be the same for all applications?0 -
[quote="Christian Gruner" wrote:
We need the raw-files...
- Preferably a shot of an X-rite Colorcchecker, correctly lit (no reflections, uniformly lit), filling the whole frame.
to have raw data skewed by mechanical/optical vignetting and cos^4 rule ? targets shall be shot to fill just the center of the frame, well far away from the borders0 -
[quote="meanwhile" wrote:
I thought white balance was white balance, that the same colour temp would be the same for all applications?
Yes, me too.
But it seems that Adobe software desn't take into account your white balance.
In French
http://couleureticc.com/wp/2015/04/27/l ... /#comments
To check it, it is simple: fix your white balance in your shot at 5000k for example, and develop it in ACR and Capture One.
The white balance will be different in ACR and 5000K in C1.
I don't know why but it is a fact.0 -
[quote="ChrisM" wrote:
Bottomline is however i.m.o., that the profiles for cameras are overall very accurate.
bottom line is - take a target like a colorchecker (passport or classic), measure with spectrophotometer, take a shot, do your raw conversion, use a tool like babelcolor patchtool to find a difference between conversion and measurements - do you think you will be able to get your conversion with C1 OEM profile within max dE2K = 1.x for 😎 ... only when you do that come back to claim that "profiles for cameras are overall very accurate"
here is one exercise a month ago, where I was approximating camera's SSF/CMF using the approach from http://www.cis.rit.edu/jwgu/research/camspec/ & Matlab , results using profile built based on that approximation and software from A. Torger (which allows to build profiles for both ACR/LR and C1 giving the same color transform) - upper left corner the target (passport) measured with spectrophotometer, bottom right corner - raw conversion :
http://s13.postimg.org/ih0qhg4fr/de125_1.jpg
http://s9.postimg.org/nr68fpihb/de125_2.jpg
http://s29.postimg.org/h91kh95qv/de125_3.jpg0 -
[quote="Peter50" wrote:
Studio shots on white seamless are actually too neutrally boring if you white balance on the seamless.
so they do make that paper with absolutely no optical brighteners ?0 -
"so they do make that paper with absolutely no optical brighteners ?"
That's a very good question. I haven't measured them, but I guess I could and see what the "b" values are. There are definitely two or three different "white" seamlessness out there and the one(s) designed for color photography are supposed to have a minimal amount of brighteners while the ones designed for black and white have a lot and are much bluer to the eye - and I'm sure to the spectrophotometer too.
But nevertheless, even shooting on medium gray seamless, and gray balancing on that produces essentially the same result. Another test, and this is one I've encountered on numerous (and I'm talking hundreds) occasions, is that products that have neutral values in them, shot on white seamless and balanced to those seamlessness retain their neutrality in the parts of the product that are, in fact, neutral. Hell, shooting with AWB on in those circumstances gives an almost perfectly neutral image as the camera is auto balancing on the background. Using the eyedropper in C1 might give you the slightest change, but it's not enough to argue about, and by the time you're done with whatever image in Ps, it won't matter in the least.0 -
[quote="meanwhile" wrote:
The only thing we can say from this is that CO and ACR reads the camera set WB slightly different
Can you quantify the difference, and are there reasons for it? I thought white balance was white balance, that the same colour temp would be the same for all applications?
I can't comment on how ACR does their wb, however, I can tell you how we do it.
When implementing a camera in CO we know all the kelvin values and also the value of the cameras presets. We use these to match numerical values to their respective look. That means that if you set your fully suported camera to 5200k, CO displays 5200, and also matches the OOC jpeg for whitebalance (given you are using the standard/neutral jpeg look).
We think this is the most consistent way.0 -
I had the orange/brown cast with my Nikon D700/D800 but now I have moved to Sony its fine. 0 -
Hi Sizzlingbadger
Out of interest (I may switch from Canon to Sony soon) which version of C1 are you using to process Sony files, and which Sony camera?
Thank you.
Peter0 -
Hi Christian
Thank you for this explanation. So if I understand correctly Phase relies to a large degree on numerical quantification of colour values and Kelvin readings; and presumably on RGB output readings.
But how reliable are these readings? Do you mean "precise" or do you mean "accurate"? Or perhaps even "relative"? Judging by all the C1 colour problem posts relating to an apparently systemic overly saturated yellow-orange-brown tint it seems to me that your data, while doubtless being precise and repeatable are actually not accurate for mixed lighting conditions. Perhaps Phase One thinking is still influenced by their experience with controlled lighting in studio conditions?
If it is any consolation, other RAW converters can create colour casts. In my hands Silkypix outputs tend toward yellow-green and DxO is too red and too harsh. I don't use ACR much. C1 version 7.1.3 outputs are just right for me (Canon 7D and 5D2), but it won't process colour files for more recent DSLRs (why not?).
These colour casts can arise from relatively small differences in RGB readings (I have measured this). The human eye-brain system is remarkably sensitive and it is easily influenced by juxtaposed areas of different colours. And individuals' perceptions vary - perhaps as seen in the C1 posts?! Quite a challenge!
I conclude that all RAW converters will have their own colour signatures and many of us will see these differently. Similarly, we adjust the so-called colour temperature to suit personal preferences. But the colour tones that make up an image all have their own colour temperature. It is almost nonsensical to ascribe a single number (Kelvin) to describe a whole image.
Perhaps the best thing that Phase can do is to back off from any claims about superior colour rendition and review the terminology used to label specific C1 colour outputs. In the meantime, perhaps a new C1 beta-testing panel should be set up comprising individuals who have had their colour sight checked?
As Grant often writes, HTH.
Peter0 -
[quote="Peter" wrote:
Perhaps the best thing that Phase can do is to back off from any claims about superior colour rendition and review the terminology used to label specific C1 colour outputs. In the meantime, perhaps a new C1 beta-testing panel should be set up comprising individuals who have had their colour sight checked?
That reminds me of the old story of the Master Printer who had worked his way up through the company an knew all of the printing presses settings intimately, eventually being promoted to the position of Quality Controller. His was the final say about whether the print quality and colour accuracy was acceptable for release.
After some years in the job he came to retirement and a dinner was held in his honour. Making his speech of thanks he surprise everyone by sharing a secret he had kept for his entire working life as a printer.
He was colour blind.
His decisions to accept or reject print results was based on asking others, whom he trusted to be consistent in their views and about whom he had worked out well enough how accurate were their opinions, what they thought of the results. None of those questioned knew about the others.
Based on the answers he would release the job or require it to be re-run.
The thing is that to do that with printed media at one location where only lighting conditions, ink pigments and media colours are in question is probably not too difficult.
Trying to set up a similar sort of controlled environment in the world of electronics with people perhaps scattered around the world ... not impossible I suppose but somewhat expensive and time consuming for those in the team.
Furthermore, how would you go about getting a certificate of colour interpretation competence?
Grant0 -
The point is that somewhere along the line the results changed.
For example the default profile for the Canon 40D is fine. The initial profile they produced for the more recent 7D MkII wasn't.
It's clear to me that Phase One did know how to produce realistic camera profiles, but for some reason they started doing it differently.0 -
Hi Grant
A typical to-the-point but sideways response from you, as usual. And very welcome it is too.
The answer to your query on colour competence is a colour blindness test. You can find this on the internet.
Before I got too involved with the C1 colour saga I took the test. I passed.
If Phase staff and their beta testers have not done the test..... ? Surely this should be a prerequisite?
As I am fond of writing: just a thought!
Peter0 -
Hi RichardT
Thanks for your note. It is possible that Phase has moved staff around with the unintended consequence of a different perception of colour, leading to a step-change in the C1 outputs.
Another possibility is that Canon and Nikon have tweaked their camera colour reference files without telling Phase. Technically feasible I am sure.
Like you, I just use C1 to the best of my ability.
Peter.0 -
[quote="Peter" wrote:
Hi Grant
A typical to-the-point but sideways response from you, as usual. And very welcome it is too.
The answer to your query on colour competence is a colour blindness test. You can find this on the internet.
Before I got too involved with the C1 colour saga I took the test. I passed.
If Phase staff and their beta testers have not done the test..... ? Surely this should be a prerequisite?
As I am fond of writing: just a thought!
Peter
Peter,
Is there a new version of the typical colour blindness test I recall having seen over several decades?
See, the challenge here, it seems to me, is not so much about single colour blindness and more about the precision of colour interpretation across the spectrum.
Recently I have seen articles that seem to suggest that people involved with understanding vision - both eyes, the signals they create and how the signals are interpreted by the brain - are thinking that there may be a significant variation in how individuals seem the spectrum even within the majority groups who are not identifiably colour vision challenged.
This, to me, seems entirely logical - far more logical than assuming that everyone's vision is identical unless they are one of a relatively small group with red or green impairment (for example).
Then there is the whole subject of colour persistence, especially notable when using screens, and how that might affect the way we subconsciously "read" colours.
The Mk. I eyeball is a great asset for humans but from a manufacturing perspective its standards are few and build quality can be haphazard. The supporting interpretive engine is also not known for consistency, unit by unit. Both may vary over time.
One might look at things statistically and have a suitably large invited group of people who could be offered a set of colour interpretations and indicate which they prefer, taking the mean or the mode answer to be a presumed "public preference". That might provide justification for the chosen settings.
However if 90% were happy enough and 10% were not that would still hardly stand as a solution to the problem with which all would feel aligned. 99.9% versus .1% might offer a stronger case.
I suspect that "the answer", assuming there is a single answer, will not be found by changing some paramaters after finding a set with which everyone is happy. They won't be.
Providing a base line and allowing people to adjust it for their own preferences might be better.
C1 can do that now but maybe the creation and the saving are not yet as readily "accessible" to a range of users as it might be possible to make them. Perhaps working on a "re-make and save" tool would be a better option? Something that would allow people to interact even more easily then the tools allow currently
But then I will confess that for the cameras I have I am not suffering with terminal default colour issues. I do, however, like to re-adjust some of the results, sometimes quite significantly. But I can do that already so for me the defaults, if not quite what I want, are not the end of the world - merely a starting point for changes.
Probably just as well given that one's eyes are not entirely consistent at the best of times - or so I find.
Grant0 -
So the original poster on this thread posts once and then disappears, never returning to find and answer, comment or clarify some of the questions being raised. My vote says that he shot on AWB and never did and probably didn't know how to do a custom white balance on the raw file. There is a difference how ACR/Lr and C1 read AWB information from the camera, but once you custom WB in each program, the results, while somewhat different, are not so far apart as to raise suspicion. 0 -
Thank you for this explanation. So if I understand correctly Phase relies to a large degree on numerical quantification of colour values and Kelvin readings; and presumably on RGB output readings.
No, only the Kelvin values, so they are identical with the camera in question.If it is any consolation, other RAW converters can create colour casts. In my hands Silkypix outputs tend toward yellow-green and DxO is too red and too harsh. I don't use ACR much. C1 version 7.1.3 outputs are just right for me (Canon 7D and 5D2), but it won't process colour files for more recent DSLRs (why not?).
As I mentioned, we need this in support-case backed by pictures.
I conclude that all RAW converters will have their own colour signatures and many of us will see these differently. Similarly, we adjust the so-called colour temperature to suit personal preferences. But the colour tones that make up an image all have their own colour temperature. It is almost nonsensical to ascribe a single number (Kelvin) to describe a whole image.
The point of setting the WB on a greycard is to take out WB from the eqation. If the same tone of grey/white has a blue tint on one image, and a red tint on the other, their color-profiles simply cannot be compared.0 -
[quote="Peter" wrote:
C1 version 7.1.3 outputs are just right for me (Canon 7D and 5D2), but it won't process colour files for more recent DSLRs (why not?).
The simple answer here is that, in common with most other applications that I am aware of, developments move on and it is rare for anyone to go back and add validated support for new products to an older application that is no longer available for sale.
Conversely some older tools - notably for hardware support across the years - may also be dropped as being obsolete and having no further value in terms of providing support. This, for example, Canon have been known to omit older models from new tethering software releases. Apple drop entire applications. Nikon, as I recall, have done similar things.
It's the way things are.
Usually the effort required, which might involve some considerable re-coding effort, is simply not worth doing when a product has moved on in other areas. Even saying that assumes that there are no other technical limitations that would make the task impossible.
Grant0 -
This article is probably a bit off topic but adds an interesting aspect to this discussion. I don't think it matters whether LR and COP have different colors. The one important point to me is, who serve the BEST colours. And "best" means generally closer to reality. From my experience in product photography I can say that COP is way better than all other raw programs out there. 0 -
I saw the article.
Also I read some of the comments which seemed to be a little offensive.
I have been using CapOnePro for 10 years.It has been my raw-convertor of choice.
For Fine Art Repro I have to disable CapOne's secret sauce and implement In Cameras ICC profiler.
BUT For landscape and portrait I like what CapOne is doing. For the most part.0 -
[quote="Grant Kernan" wrote:
I saw the article.
Also I read some of the comments which seemed to be a little offensive.
In terms of accurately reflecting what a dismally wrong-headed (and for the most part, fundamentally just wrong) waste of time the article is, I'd argue that they've not offensive enough.0 -
[quote="Peter50" wrote:
My vote says that he shot on AWB and never did and probably didn't know how to do a custom white balance on the raw file.
Jeez. How about trying to be a little bit less dismissively condescending next time, new boy? I'm sure you've got it in you.
This exact issue has been argued about on here for a bloody age now, and it's nothing to do with ignorance of how white balance works.
Keep your arrogant, uninformed assumptions to yourself until you actually know what we're talking about, eh?0 -
[quote="Keith Reeder" wrote:
Jeez. How about trying to be a little bit less dismissively condescending next time, new boy? I'm sure you've got it in you.
This exact issue has been argued about on here for a bloody age now, and it's nothing to do with ignorance of how white balance works.
Keep your arrogant, uninformed assumptions to yourself until you actually know what we're talking about, eh?
You are funny. You complain. Some of us give you information, causes and remediary but you refute and you take we haughtily while saying that it is the others who are.
I'm sure the 4 pages of discussion, there are at least four relevant interventions that could make you grow.
If you wish to be accurate and well balanced, only one way: one custom profile AND a good color management.
Try.0 -
Goodness !
I was preparing to continue a dialogue with Grant.
I think I'll wait until everything calms down.
Peter0 -
Guys, 1 more post in bad or offensive language, and this thread will be locked. If you are in doubt how to act here on the forum, please refer to our forum-rules.
Keep a sober, proper, professional tone.0 -
OK. Here's where I really ran into a problem with C1 rendering.
Both photos are straight up RAW conversions with no corrections.
Shot with D800E, A single Profoto B1 with a beauty dish.
The first is Nikon NX2
The second is C1
The young couple looks like the Nikon version.
Her skin is very light.Her hair is "red" Her boyfriend's hair is dirty blond. He works outside and has a bit of a tan.
I was never able to get this looking right in C1. However Nikon's version got it with no fiddling.
http://paintboxphotography.com/img/s12/v171/p1333623897-5.jpg
http://paintboxphotography.com/img/s2/v73/p1333623900-5.jpg0 -
Thank you for some more concrete examples finally.
The example from NX is something you will never see copied in Capture One as it has green-tinted look to it.
However, do feel free to send us the raw for further analysis. It seems to me that WB's are not completely similar, without having looked at the numbers though.[quote="Jimmy D Uptain" wrote:
OK. Here's where I really ran into a problem with C1 rendering.
Both photos are straight up RAW conversions with no corrections.
Shot with D800E, A single Profoto B1 with a beauty dish.
The first is Nikon NX2
The second is C1
The young couple looks like the Nikon version.
Her skin is very light.Her hair is "red" Her boyfriend's hair is dirty blond. He works outside and has a bit of a tan.
I was never able to get this looking right in C1. However Nikon's version got it with no fiddling.
http://paintboxphotography.com/img/s12/v171/p1333623897-5.jpg
http://paintboxphotography.com/img/s2/v73/p1333623900-5.jpg0 -
I don't see a green tint on my fully calibrated system. I see a magenta/red tint on the CO image which exactly what I get with my D800 images and I'm not alone...
viewtopic.php?f=52&t=16148&p=76148&hilit=d800+colour&sid=ab3fd6661a98c96ba2b77e592c773fea#p76146
viewtopic.php?f=47&t=15451&st=0&sk=t&sd=a&hilit=d800+colour
viewtopic.php?f=47&t=15327&p=71563&hilit=d800+colour&sid=ab3fd6661a98c96ba2b77e592c773fea#p715630 -
[quote="Peter" wrote:
Hi Sizzlingbadger
Out of interest (I may switch from Canon to Sony soon) which version of C1 are you using to process Sony files, and which Sony camera?
Thank you.
Peter
Hi Peter,
I use the full version of CO and a Sony A7II.0
投稿コメントは受け付けていません。
コメント
68件のコメント