Out of Gamut Colors in Unprocessed RAW File
Does anyone here have any idea how one can get out of gamut colors in an unprocessed RAW file? I have my camera set to use the Adobe RGB color space because I prefer the wider gamut. When I open the photo in the Editor and set the Camera Profile to Linear RAW so there is absolutely no processing being done to it, then I use Enable Soft Proofing with the Gamut Warning option turned on and the Adobe RGB (1998) profile selected I'm seeing out of gamut bands of color. Switching to sRGB amplifies the problem with far more banding showing up. Exporting to a JPEG with sRGB shows the banding quite visibly.
I'm having a hard time understanding why there would be out of gamut colors in a RAW file to begin with. How can the sensor capture out of gamut colors since all it sees is the amount of R, G, or B light hitting an individual sensor pixel according to the Bayer mask filter overlaying that pixel. I thought the color space setting wasn't applicable to the RAW data, only the demosaiced RGB image and the embedded jpeg preview. Am I wrong about that? My camera manual isn't clear on this point.
Anyone have any ideas about this? I sent a request to tech support but thought more eyes on the problem might help.
Brian your raw file is only light...then you wb it, demosaic it and use a colorspace...you should be using Prophoto during your editing or rec2020 or some wide gamut space and then in the export settings use something like Adobergb for printing and sRGB for jpg exports ON1 colormanagement will handle the conversions...color managing will handle things...your camera settings are only for the in camera jpg...they have no bearing on a raw file no matter how you set it......the raw file lives in the camera's colorspace until it gets processed......so you are creating the gamut warnings because of your selections...0
Yes, I understand all of that and I do use ProPhoto for my editing colorspace. And that is why I do not understand how I'm getting out of gamut colors before I even process the photo. If the camera was told to use Adobe RGB and I'm viewing it in Adobe RGB where did the out of gamut colors come from? ProPhoto is the only colorspace that does not show any banding but I can't post something online in that colorspace. It all gets transformed to sRGB by whatever site I post to.
The problem is that as I process the photo the banding becomes amplified and very visible even before exporting it to a smaller colorspace. Because of how the out of gamut colors are lying in bands I haven't been able to find a way to remove it during my processing. The things I've been able to find to reduce the banding affect the overall color in those bands so there is still banding visible even when everything falls within the colorspace. Or, I have to completely destroy the blues entirely to get rid of it.0
Brian, as you mention in your first post the camera color space setting does not impact the raw file. Isn't soft proofing and the out of gamut warning supposed to indicate that the colors are out of gamut for a particular printer / paper combination so it's just indicating your printer cannot print the exact colors being shown on the screen?0
Brian, See page 107 of the user guide.
Soft proofing is turned on via the View menu, or
by clicking on the soft proofing icon to the left
of the Preview button at the bottom of the main
Photo RAW window. You soft-proof various
paper types by choosing an installed printer
profile from the Profiles submenu. The image
will change to approximate the color of the
paper you’re profiling. You can also turn on an
overlay that will display out-of-gamut colors.
Hudson Henry has a good vide on Youtube about softproofing for printing but I can't find it now. I'll keep looking for it.0
My mistake the video is not on Youtube it is an On1Plus video course by Hudson called "Shoot to Print". He talks about soft proofing in Lesson 7: Soft Proofing. In it is a good discussion about making adjustments to print.0
Thanks David. In this case, my 'printer/paper' combination is a monitor which is why I'm using the sRGB profile. The banding I'm seeing in the soft proofing matches exactly the banding I see when an sRGB JPEG is being displayed.
I'll check out Hudson's Shoot to Print course again. It's been a while since I watched it.
Interesting, I just exported a reworked version of the shot and I'm not seeing the banding in the jpeg this time even though you can see it in the screen captures above.0
You mention again about the camera setting....just to be clear with all the wrangling I think David expressed it correctly.....Those camera settings have no bearing to ON1 at all unless you are editing a camera generated JPG file...your raw is processed by on1 and in general assigned a prophoto working space...this is the space that that all the calculations are done in and as long as you stay in gamut here you should be fine...next is your export profile and that again should handle the colorspace conversion and map the colors...I don't know what ON1 uses perceptual or relative?? Things can go offside if your monitor profile is not correct...for ON1 I think this has to be at the system level and then if you introduce softproofing ....I think you might have a wonky file if this only happens with this one file or if it is happening to all of them it is your combination of color profiles that is messed up...0
Thanks, Todd. Again, yes, I understand all of that. I just profiled my monitor a couple of days ago. I do it regularly. I have a BenQ SW2700PT and I use an x-rite i1DISPLAY PRO PLUS colorimeter to calibrate it with BenQs Pallete Master Element software according to ArtIsRight's videos on YouTube. The SW2700PT can be hardware calibrated, not just have a color profile applied to what gets sent to it.
I haven't looked in this level of detail at other photos because I haven't had the problem with previous photos. I only started looking at the cause of the problem when I couldn't resolve the out of gamut issues to my satisfaction. That is when I discovered the banding in the RAW file.
When I look closely at the unprocessed .CR3 files I took for this photo (a 2 exposure HDR) I can see bands of yellow noise in the sky particularly in the under exposed shot. Trying to process that shot by itself brings it out as I lighten the sky. I'm trying to figure out where they came from which led to looking at the unprocessed files with the soft proofing option.
It's been an interesting experience as I've processed this image multiple times in different ways trying to minimize the problem. I've tried the HDR module, manually blending them together, and processing one or the other singularly. I wound up using a combination of the HDR module and manually blending the sky with the HDR foreground.
I'm happy with the shot I showed in my previous post. Now I need to finish it up with a border then exporting the jpeg and posting to Facebook again. That's going to be the pudding that provides the proof. Hopefully it won't be as bad as the first version I posted the other day where it is horribly obvious.
Thank you both for taking the time work this through with me. I've been doing tech support for a very long time so when I have to ask for help it's usually not something simple to fix. 😄 It isn't unusual for me to have to educate the tech support folks I work with about the problem before we can start making progress on a solution. 😉 Not talking about ON1's folks in this case, I haven't heard back from them yet.0
I have no doubt that you will get it figured out. I am more used to Darktable where access to all the profiles is quite clear and gives you full control.... Although DT uses icc profiles like ON1 the ones created by Xrite-really don't work well for it. Always very dark and its doesn't matter about exposure of the test shots. Simple matrix profiles are better as it does scene-referred edits and is not working on bounded data from 0-100 or 0-255 and lut based profiles can present problems for this....Its a pitty ON1 does not have a forum like pixls.us for the opensource editing software. There you can upload your raw files and/or side car files and let other edit or respond to issues or questions...they call it playraw....its a very good learning tool and responses are quick as it is user driven not depending on support....0
Brian, here's an interesting article about color banding and fixes.
One of the fixes discussed is to add blur to the sky.
How to Resolve Color Banding in Your Images - Savage Universal0
David, Lol, that's one of the things I did differently for my last workup. :) Thanks again, I'll take a look at it.
Todd, thanks for that. I'm not using an x-rite color profile in this case. Just the out of the box Adobe RGB & sRGB profiles that are common to all editing software. I have used them on occasion, I also have one of the ColorChecker Passports but I'm not sure how well one of those would have worked in the pre-dawn light even if I had the time to take the calibration shot. I literally got out of the car and set up the camera and got off 2 frames before it was too light to see all the planets. I got up there a bit later than I had wanted so it was a hurried shoot.
If you are interested, Venus and Jupiter are in conjunction at 3AM ET on 30 April. They will have 0.2 degrees separation, less than the width of a full moon. No chance of a passport shot for that one but I will be shooting it just because. Shouldn't be any banding problems caused by the rising sun either!0
I think that it is likely that many unprocessed RAW images will show out of gamut warnings when soft-proofing against anything other than ProPhoto. That is, I believe, On1's internal working colour space which is larger than Adobe RGB and sRGB. I have just looked at several of my photos at random both processed and unprocessed and all show out of gamut warnings if I soft proof against anything other than ProPhoto.
I assume that MacOS deals with out of gamut colours in its own way depending on monitor profile, so not really visible.
So I don't think that your original question about why you get the warnings is relevant, files are demosaiced into ProPhoto regardless of any monitor settings etc. So there will be colours that are out of gamut in smaller colour spaces.
As Todd said it is probably something to do with this file, including how On1 does RAW conversion, denoising etc that produces banding. Possibly just the nature of the how the sky colours form a gradient.0
Ya I saw something checked like simulate paper and ink...I think really I would have to get more familiar with the softproofing in ON1 . I don't print so I don't use it really but people often use soft proofing against a color space to check gamut and I have been told in the past this is not really the correct use. It should be used in conjunction with printer and paper profiles and using those the display will be modified to try and simulate the output so you can adjust your edit accordingly ...to use srgb or adobe in the soft proofing context as profiles to try to check gamut is not really using the right way at least as I understand it and how it works in other software I have used.....If gamut doesn't clip while processing in the working space then when you select your output profile be that srgb or other and if the application is color managed that conversion should be handled correctly mapping colors in one of three chosen manners ie perceptual relative or absolute.... It could be in Brians case that those images shot in the conditions that he had to work with just created an image with some inherent issues. In a recent forum using another raw convertor it was noted that the Sony A7ii could have some weird colors and WB issues when shot in low light incandescent and even custom profiles for those conditions were less than effective so maybe just the extreme low light for that capture may have introduced some artifacts in this case that you normally wouldn't see....in any case there is always something to learn ....0
Photo RAW allows us to choose between Perceptual and Relative when changing the colorspace.
I had considered that my problem could have been caused by the lighting conditions I was working under. I also considered that it could be from how the program is handling the demosaicing and/or applying the colorspace.
As I recall from my Photoshop days, when it opens an image file (not raw because there isn't any colorspace in them) which does not have an icc profile embedded we are given the option to assign a colorspace to it which doesn't alter the colors in the photo, or applying one to it which does alter them. Of course assigning the wrong colorspace (i.e., we chose sRGB when it was created in a wider space) would result in all kinds of wrong colors when the photo was printed.
I'm still waiting for a response from tech support. I'll post there answer when I get one.0
Here is the response from tech support about this.Our dev team has indicated the following r regarding the issue of banding:
"When viewed at Fit you will see a banding effect when switching profiles. If you switch to 100 shouldn't see any banding effect when changing profiles.
This is most likely due to the previewer which uses a proxy to display the image when in fit view."0
There is this video https://www.youtube.com/watch?v=q5yK5hFQydo which describes a similar issue with Photoshop showing banding in previews that is not there in exported image because Fit previews are 8-bits regardless of bit depth of image being edited.
I did the test on On1 and didn't get the same results so I assumed that On1 didn't do the same, but perhaps it does in some cases.
I assume that it comes down to balance between quality of previews and performance.0
Please sign in to leave a comment.