Importing wrong colour space

[HalloweenJack2]

Hi guys,

I've just imported 2 movie clips from another company. One for ref and one HD to use.

They were nice enough to stick Bars on the front for reference on both clips.

Looking at the ref clip the bars are correct.

Looking at the clip to use the blacks are raised and whites too low.

I imagine this is importing as YUV instead of RGB (or vice versa, I never remember) Or that they exported one of them with the wrong settings. I always get a bit confused with this stuff as it always works fine for me beween different systems. I see it all the time from externally produced files but I can never trace the history.

Anyway, I thought the Include YUV headroom option sorted this but it doesn't seem to do anything.

Using a Gamma LUT and putting White Ref at 236 and Black Ref at 14 looks about right.

I'm working in 8bit obviously 🙂

So can it be tied down what the issue is?

Can Flame compensate without doing it by eye?

I thought there might be a LUT that would do it.

Thanks 

[ManChicken]

What format/codecs are these movies?

There are a couple of Color Transforms in the 'levels' directory for converting full-to-legal and legal-to-full, the latter which you could apply to your clip.

The YUV Headroom switch would actually make your problem worse if it was working;  That it's not doing anything at all, I suspect your file is an RGB codec of some sort (ProRes 444, uncompressed RGB, etc)

[HalloweenJack2]

Hi,

The Ref is QT H264 and the full res is FFMPEG.

Legal to Full works perfectly. Thanks!

So what's happening here? Is there a set of rules to note and adhere to?

Or do I need to set aside a few weeks for a Colour Space course 🙂

Cheers

[ManChicken]

ffmpeg isn't a codec, it's an encoding program.. but that in and of its self indicates it was possibly exported from something, and then compressed after-the-fact with ffmpeg.

It's hard to speculate without really knowing where it originated from (what software/NLE) and what codec it actually is.  The file isn't necessarily wrong;  The file you have is HD range (16-235 in 8-bit), but is proably in a codec that is normally full-range RGB.. so Flame is treating it as full-range and just 'passing through' the values.

[HalloweenJack2]

So I've found out the file in question came in as part of an Avid bin. The edit assistant put it into an Avid then exported it as a QT with animation. 

In the process you can choose whether you export as 601 or RGB he choose 601 thinking it was a broadcast standard and in an Avid.

He exported again today as RGB and Flame exported this fine.

So I guess the footage was full range not 601.

Is there a way of telling which it is?

The LUT you suggested seems seemless, should it be excatly the same when you apply the LUT to the limited range file compared to a file that was exported full range?

Thanks for your help on this.