I am working in AE comping greenscreen plates shot with a sony f950 directly into FCP using the a Kona with the Kona 10bit 4:4:4 codec. When I switch AE to 16bit the image gets lifted. I wrote to support about this but never heard back. Is there a newer software codec to fix this? I typically like to do most of my work 8bit then switch to 16bit for rendering, but this codec is preventing me from doing so since CC’s look much different in 16bit.
Are you working in regular HDcam or HDCam SR?
If you working in regular HDCam, then working in 444 RGB is way overkill. HDCam is not a 444 format. It’s still 422 Uncompressed 10-bit, like D-1, though the frame size and number of scan lines is different. Outside of this knowledge, I can’t say that what you are trying to do is necessarily wrong. It’s just that your monitor is probably not a 444 RGB monitor, and your footage is being ‘converted’ to 444 in the capture process. So maybe it’s the monitor???
Just guessing here.
Try doing it with the 10-bit UC 1080 23.98 Easy Setup [or whatever the native acquisition frame is].
double wide post
4:4:4 is not overkill… The whole show is greenscreen with virtual environments. This was shot with the Sony F950 camera which is the SR camera. We had an SR deck as a backup, but to avoid the SR compression we went dual link out of the camera directly into FCP with Kona. It worked flawlessly except for this luminance bug.
For the record, your specs are really inaccurate. I suggest finding the specs chart from Miranda.
HD-CAM is far from uncompressed… 7:1 compression, 8bit, 3:1:1 colorspace. It’s dead technology.
D1 is (was) not a 10bit format… it is 8bit uncompressed 4:2:2. Some folks still like to use it for effects shots. I prefer the larger color space in Digibeta (10bit 4:2:2 2.3:1 variable compression which is lossless a lot of the time).
All computer monitors are 4:4:4RGB. I see this shift on computer monitors and out 4:2:2 HD broadcast monitors.
[Matt Silverman]“4:4:4 is not overkill… The whole show is greenscreen with virtual environments.”
Agreed Matt —
this is the optimum usage of a 4:4:4 RGB master, it is what the format was created for.
Back to your original very first post-
it is my understanding that the 16bit version of the codec that AE uses is not optimized for the Mac based version of QT but for a Windows file based ( DPX or Targa) workflow.
Have you tried the 10bit Kona RGB or RGB Log versions of the Codec to bypass the native QT version of 10bit?
HD & Film Consultation
Post and Production Workflows
You are correct on all counts. I was thinking F900, not F950. Also, since Sony does not give up its native codec to Apple, the logical capture mode for HDCam [assuming shot with F900] would’ve been 10-bit [or 8-bit] Uncompressed 422, which is a transcode of sorts. My point, of course, that for regular HDCam, anything more than 422 would be overkill.
Not the case for F950 HDCam SR.
I stand corrected. Thanks.
double wide post