-
Comparing Color Accuracy of FCPX, FCP7, PProCS6, Technicolor Assist
Interesting article about the color accuracy in various NLEs as seen through their scopes
Why video color changes on different monitors and programs
https://blog.technicolorcinestyle.com/blog/bid/305517/Why-video-color-changes-on-different-monitors-and-programsApple’s TechNote which also includes link to download test files.
https://developer.apple.com/library/mac/#technotes/tn2257/_index.html -
10 Replies
-
The link doesn’t works.
Rafael -
I’m sure it’s correct. Their whole site has been on the fritz this weekend.
-
Both links working for me. Perhaps intermittent issues.
-
I’m all for accuracy, but there seems to be some inconsistencies, at least with fcp7.
As we should now know, fcp7 is a pretty good YUV renderer. If you notice their scopes in fcp7, they have the grey bars above 100 and below 0. This means they are working with “always render in RGB” checked. If they switched to a YUV rendering they would probably get more accurate (or expected) results provided that the test images they are using haven’t already been truncated to Legal RGB.
So, while I appreciate the effort to sell me some accurate software scopes, I think we need to look at what is really happening with the source material.
FCP7 and X allows to conform to regular 709 gamma, and also FCP7 has exclusive features for gamma correction of RGB material (most default to a “source” gamma of 1.8). There’s even a preference to conform to 2.22 in FCP7 prefs.
-
I’m with you Jeremy.
There is something that I don’t like here.
First I don’t like the Gray Scale.
Up to Photoshop and up to the Mac Digital Color meter, values are wrong.
– 0 IRE should be 16.
– 50 % should be 128
– White should be 235Then his scope and FC scope (don’t know PP) are using different unities.
As you point, the FC’s VideoScope is in RGB (so no IRE.)
The signal is being crunched down.
Is impossible that the bars levels fall at the targets lines.
rafael -
[Rafael Amador] “- 0 IRE should be 16.
– 50 % should be 128
– White should be 235”And I think this is where the information might not be quite right from their end. I wish we could have their source gradient as obviously, the still grab they post on the website isn’t going to be correct.
I think they are assuming 0-255 instead of the more common 16-235 RGB values for component video.
I’m not saying software scopes are the most accurate means of measuring a signal, but there does seem to be something weird with their process, and in most cases, assuming 16-235 is the right thing to do.
-
You are right.
The article is about VIDEO.
The FC.7 VideoScope is OK for RGB graphics, but not for video.Also, the signal is showing FCPX Parade, is not the same he is showing on his own scope.
FCPX Parade Is showing a perfectly Broadcast signal with 100% Luma.
Where is the SuperWhites bars (Luma=109%)?Parades, are not graded to show absolute RGB values, but Broadcast legal RGB values.
100% on a PARADE doesn’t means 100% R, G or B (255).
100% on a PARADE marks the legal White so R, G or B = 235.
Rafael -
Anyone use Color Assist this on a routine basis? I’m thinking about getting it (or FilmConvert), but the site’s been down for me for the last four days.
-
Seems like Color Assist was pretty short lived:
https://www.technicolorcinestyle.com/cinestyle-color-assist/
Reply to this Discussion! Login or Sign Up