Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

frickin gamma

Justin McAleece

Well-known member
Joined
Jan 21, 2007
Messages
807
Reaction score
3
Points
18
Age
47
Location
Fresno, CA
Website
www.blaremedia.net
i just re-taught myself a lesson that has been pissing me off for a while now. in FCP I made a sequence that was originally uncompressed 10 bit 4:2:2 but then I switched it to every other compressor I could think of and then I changed the user preferences to alter the gamma that FCP would supposedly apply on import then I changed the RGB -> YUV settings back and forth a bunch of times.

anyway, I tried everything and no matter what my clips in the timeline were still darker then they were in the viewer if I just imported and double clicked. The problem ended up being that I had set the dimensions of the sequence to a sub-standard resolution. We shot greenscreen footage of people standing up and we didn't want to switch the RED 90 degrees because it was on a teleprompter and so long story short I had a lot of extra resolution on either side of the screen that I had no use for. So I started with a 2k (2048*1152) and just cropped off about 500 pixels from either side. this = 1048*1152.

For some completely retarded reason this causes FCP to shift the gamma to make everything darker. when I just drug the clip to a new sequence and let it figure out the settings the gamma levels were fixed. So I said, what was the difference? So I went back and changed the resolution of the one that I just made that had proper gamma. as soon as I changed the res to something substandard the gamma shift returned. have I mentioned that QT and FCP suck in so many ways. (i still use it 12 hours a day so who is the chump?)
 
i just re-taught myself a lesson that has been pissing me off for a while now. in FCP I made a sequence that was originally uncompressed 10 bit 4:2:2 but then I switched it to every other compressor I could think of and then I changed the user preferences to alter the gamma that FCP would supposedly apply on import then I changed the RGB -> YUV settings back and forth a bunch of times.

anyway, I tried everything and no matter what my clips in the timeline were still darker then they were in the viewer if I just imported and double clicked. The problem ended up being that I had set the dimensions of the sequence to a sub-standard resolution. We shot greenscreen footage of people standing up and we didn't want to switch the RED 90 degrees because it was on a teleprompter and so long story short I had a lot of extra resolution on either side of the screen that I had no use for. So I started with a 2k (2048*1152) and just cropped off about 500 pixels from either side. this = 1048*1152.

For some completely retarded reason this causes FCP to shift the gamma to make everything darker. when I just drug the clip to a new sequence and let it figure out the settings the gamma levels were fixed. So I said, what was the difference? So I went back and changed the resolution of the one that I just made that had proper gamma. as soon as I changed the res to something substandard the gamma shift returned. have I mentioned that QT and FCP suck in so many ways. (i still use it 12 hours a day so who is the chump?)
FCP changes the pixel format it requests based on: Pixel depth information in the QT wrapper, Image size, Display size, Whether the display is cropped, window or Cinema Desktop, playing or scrubbing. These are the ones I know about, but if someone said "phase of the moon" I would believe it.

Some of the pixel formats are YUV so do not have gamma issues, but have 601/709 color space issues, others are modern high bit depth so do not have gamma issues, and some are 8-bit RGB, which depending on what it is doing FCP assumes are 1.8 - no matter what you tell it, and sometime 2.6:waaa:
 
have I mentioned that QT and FCP suck in so many ways. (i still use it 12 hours a day so who is the chump?)

You're not a chump. At the end of the day FCP is a great tool.
(Just like Avid. Both have their pros and cons. Neither are perfect, neither is better than the other. It's more what is better for you.)

FCP/QT are known for gamma problems...
 
Try the FCP.mga LUT in File > import > display LUT in Color... try to grade without it and render with it OR grade with it and render without it...

One of the two should work... The FCP.mga LUT is for this problem...
 
What is your monitor gamma?

the monitor gamma shouldn't matter because the picture in FCP is changing depending on the size of the sequence which shouldn't have anything to do with it. nothing else on the screen is changing - just the canvas. thanks though.
 
Why does any gamma in the middle of the chain matter?

I myself correct in Color, then send to FCP, and it looks the same on NTSC output in Color as it does when it is finally output.

In other words, why does anything matter except for the last step?
 
up-down-up-down...

up-down-up-down...

Why does any gamma in the middle of the chain matter?

In general, if you make gamma adjustments up then down then up then down on the image data in serial form the "real image data" bit depth goes down, and histogram gaps develop. On each adjustment and save the values are rounded to the binary "word" size bits.

If you start with large "word" sizes like 48bpp 16bit you can alter the gamma, contrast, brightness, etc. more times than starting with 10bits or 8bits before the gaps get wide enough to see.
 
In general, if you make gamma adjustments up then down then up then down on the image data in serial form the "real image data" bit depth goes down, and histogram gaps develop. On each adjustment and save the values are rounded to the binary "word" size bits.

I think you misunderstand my question. A lot of the questions, the way I understand it, is why does gamma look different in this program vs. that program.

I think the answer, in effect, you're providing is that if you're doing baked renders out of one piece of the chain to then enter a new step, then yes, those errors would stack up and that would suck.

Since a lot of the complaints stem from Quicktime and the FCP/Color flow, my question is, why does one care what it looks like in FCP if you're finishing in Color?
 
Hello Nate,

I guess that is because some people have to go back to FCP to finish with effects and stuff. People who goes to other online system to finish, probably won't matter how FCP treats gamma.
 
Back
Top