Personal View site logo
Make sure to join PV on Telegram or Facebook! Perfect to keep up with community on your smartphone.
Please, support PV!
It allows to keep PV going, with more focus towards AI, but keeping be one of the few truly independent places.
GH2 and Final Cut Pro 7 color space question
  • mangomonstermangomonster

    Hey guys, I feel I have a pretty straight-forward question but I haven't been able to find a straight-forward answer.

    Can anyone walk me through their workflow for FCP 7?

    Just read some article today and wondered if my workflow is a problem and what I can do about it.

    I simply film on my GH2 running Flow Motion v2.02, then do a log and transfer in FCP7 into a ProRes422 file. I have thought many times the brightness/color seems a bit off but I'm not sure. So reading the above article made me wonder. FCP 7 doesn't seem to have the same input/output levels adjustment or at least I don't know where to find them.

    Is there a gamma or colorspace mismatch happening?

    You can just increase your gamma to 1.22 to fix this problem with the 5dmiii...

    Thanks for any help.

    ~Tim

  • 4 Replies sorted by
  • stoneystoney

    Gh2 records avchd files in broadcast luminance range. Some editors assume dslr's source footages in full luminance range. So they convert the source into broadcast luminance range again. This is a bug.

    The tip from EOSHD fixes the bug by negating the last luminance range transformation by boosting the shadow and suppressing the highlight. Similar concept has been covered here when using 5DtoRGB.

  • DrDaveDrDave

    There is currently some disagreement on the tonal range of the AVCHD files of the GH2. Without wishing to start a flame war over this, I would suggest you should make some test images and see if there are any shifts in the process. Different apps handle these things differently, and if you are viewing the results within the app itself, you can easily wind up in a Catch 22 situation as far as knowing whether the footage has been squeezed and then unsqueezed. I load my footage into Premiere and and grade it, if I set the blacks as high as eoshd recommends, I would have some serious blacks. But I shoot a lot of dark material, and I shoot a bit underexposed to avoid clipping the highlights. And eoshd certainly has a good eye for detail. The scope in premiere shows that data of some kind is present in the full range, but some say you can't trust the scopes and so on. I don't know about that, the scopes seem really accurate, but then again, how would I test them and why would I care if the results look OK and seem to coincide with reality? It's turtles all the way down. Youtube screws it all up anyway. Bottom line, if you are much happier using the eoshd workflow, and the results are compelling, then go for it. I also think if your NLE is resampling your vid without express directions to do so, that's always a problem. If it resamples it and then "puts everything back" when you view it or render it, it is going to be extremely difficult to tell exactly what is going on under the hood.

  • mangomonstermangomonster

    So how do I apply the tip in Final Cut Pro after I've brought the footage in through L&T? I'd like to compare 5DtoRGB vs L&T but the color corrector filter in FCP is different than the one in Premiere. I don't have input/output level control... at least not labeled the same and I'm not really that good with scopes. I just want to make sure I'm getting th

  • mangomonstermangomonster

    the best image and if the image ends up looking pretty much the same either route, then I'll be satisfied.

    Later:

    Shoot, I just used 5DtoRGB vs. L&T on the same clip and I can't see what this guy is talking about. The difference is so minor I can't tell. Whatever's going on isn't some kind of huge shift in dynamic range. I do not see this huge problem I've heard about with Final Cut Pro 7 interpreting AVCHD footage. Looking at the waveform, there's probably .01% change going on... I can't even find it when zooming way in at the pixels and comparing.

    So maybe there's isn't a bug after all with FCP7's L&T? I dunno

Add a Comment