Looks like no one’s replied in a while. To start the conversation again, simply ask a new question.

Exporting FCP/Color changes gamma

I'm working on a 90-minute film shot hot with a 50D. It has scenes where there is natural light, photo lamps and fluorescent light. I'm an old school tape video editor gone digital, I'm used to using a TBC to do: gain, black, hue, and saturation. It's been a bear learning to use a 3 range color adjustment and it's killing me trying to color those multiple light temperature shots.


But my real problem is this:


I discovered a problem with exporting that comes from the compression in QuickTime. Even going from 422 HQ to 422 HQ the re-compression is adding a gamma curve to everything and making the blacks darker and causing faces to shift into red more.


I read a bunch of articles on this problem, but no one seemed to really have an answer. It seems everyone thought it was a monitor calibration issue or who knows what. The only thing to do was use my scopes and see if I can reproduce the problem there...


So I exported a gradient fill black to white in 422HQ and when I opened it in FCP the gradient is flat on the scopes. But if I re-export it and look at it again there is a definite gamma curve on the gradient when you look at it in a scope. If you stack both clips in your timeline and do a wipe between them the difference is very clear onscreen as well. This explains why I always thought images exported from after effects, or anything that uses QT, had more contrast. I just never bothered to figure out where it was happening, I just assumed it was the programs, but it's the re-compression that does it. It’s Quick Time.


I'm sort of perplexed at this point. I render to send to Color. Then color renders to send to FCP. After this I need to add Film grain and digital makeup artist; client’s preference on the digital makeup, I like it as is.


Adding those effects means I need more rendering. At this point I've got two renders, each drops the mids by maybe 5 ire and the blacks by even more. My end result is not quite right. So if I compensate by making things brighter and doing an opposite gamma curve before export, I'll have no idea what I'm looking at as I do the color work.


Is this how professional colorists work? Do they make a file that looks wrong, but will look right after the final codec is applied?


Try it yourself. Make a gradient in FCP. Look at it on the scope: flat slope with no curve.

Now export it with QT and recompress it: flat gamma no curve.

Re-export it again with the same codec: Gamma has a slope toward black.


Stack them in the timeline and do a half screen wipe and you can see it clearly.


Do the same thing with a clip that has faces in it. The export creates pinkish faces.


Am i doing something wrong? Is there a tool out there for handling this?

Posted on Aug 29, 2012 10:48 AM

Reply
2 replies

Sep 7, 2012 10:32 AM in response to What happened to my amiga?

Interestingly I spoke with a professional colorist with years of experience and he was not aware of this issue. But he said he was able to recreate it. He said the only format that would be safe to use for preserving gamma shift is a DPX file. ( I may have the acronym wrong, it's a sequence of frames used in film transfer.)

Sep 19, 2012 2:08 PM in response to What happened to my amiga?

You are not alone. This issue has been persisting since at least 2008 from my research. I've found that this bug affects any application that uses QuickTime libraries, such as Keynote, PowerPoint, and even iMovie for export.


Please file a bug report for this: https://developer.apple.com/bugreporter/


I have already done so, but only bugs that squeak the loudest get fixed, especially in maintenance releases of OS X.

Exporting FCP/Color changes gamma

Welcome to Apple Support Community
A forum where Apple customers help each other with their products. Get started with your Apple ID.