I built the last code from cvs and when I open any raw file (Canon 7D or 40D) it shows me multiple error messages stating that the multipliers are too big and will be truncated. The image, then, shows completely wrong colors.
Everything is fine with previous CVS code.
I have just commited a fix for this bug to cvs.
I confirm it is fixed now… Thanks!!
I was thinking to try digging the "famous" bug of the luminosity/saturation tab in the next days. Can you please point out which are the functions used there?
Is it this bug? http://sourceforge.net/tracker/?func=detail&aid=3007100&group_id=127649&atid=709086
I have not looked that much at the color management code lately so I am not able to help you out with where to start. Could you make a raw file and the profile showing the problem available for download? I might have a look at it too.
Exactly that one.
With concert photos (which I am working on a lot lately) it is a very annoying bug as it is recurring very often (and I am forced to work on sat/contrast in GIMP).
I will upload a raw rile ASAP.
here you find the raw file: http://albertoferrante.name/ufraw/IMG_8285.cr2
In this case you will not see the problem unless using a camera profile such as: http://albertoferrante.name/ufraw/7D-nonlinear.icc (gamma 0.45, linearity 0.1). On other images (it really depend on the image) you can see the problem even by using the color matrix, so it is not a problem of the profile.
Thanks for the file. I can see the problem and hope to find time to look into it. If you want to take a look at the code I think you should start with the function developer_prepare() in ufraw_developer.c.
I will do it maybe at the end of next week…
I tried to look into this problem, but it looks more complex than I tough…. Functions dealing with saturation/contrast make use of external libraries. It would take a really long time to debug this without a knowledge of what happens in each function. It looks like a rounding problem or something like this, but I could not point out where the problem could be.
Yes. It is part of the color management code which depends on the lcms library. Unfortunately I have not found time to look into the problem.
Please test the current cvs code which use lcms 2.x. Is this problem still present?
I'm just using Gimp and UFRaw… so, I'm really a very new user, but I've the same problem with multipliers and images troncated. What is cvs code ans how can it fix this bug ?
Sign up for the SourceForge newsletter:
You seem to have CSS turned off.
Please don't fill out this field.