Uncompressed Quicktimes are the most requested format requested from me. But it’s a lie! Most uncompressed Quicktimes are compressed. Wikipedia is a good place to start reading. Search for stuff like RGB, YUV, 4:4:4, 4:2:2, digibeta, chroma subsampling, ntsc, pal.
You back? Ok. So a typical “uncompressed” quicktime is usually compressed color-wise. But we don’t care, since everything is chroma subsampled from here on out anyways. Finishing in “true” 10 bit uncompressed 4:4:4 colorspace will be visually identical to a 8bit uncompressed 4:2:2.
10 bit vs 8 bit
I usually stick with 10 bit, since that’s what we get off digibeta. The only time I don’t do 10 bit is when I’m outputting from After Effects, because it normally only does 8 bit, so you waste 2 bits outputting back to 10 bit. Bit bit bit bit bit.
Most “real world” transport streams are at most 4:2:2. In fact, they usually get subsampled even further. Look it up on Wikipedia. So it doesn’t matter if it’s 4:4:4 or 4:2:2, it’s going to get mangled downstream anyways. As for other colorspaces like 4:1:1 or 4:2:0, you can have problems. This is a good read. In particular… this link within that page.
- 2Yuv – don’t use this one, Final Cut Pro doesn’t know what to do with this.
- 2yuv – use this one instead. This is 8 bit 4:2:2
- v210 – 10 bit 4:2:2
Also, when making uncompressed Quicktimes from outside Quicktime or Apple, make sure your dimensions are a multiple of 8. Otherwise, your video may end up distorted.
If you really want to make an trully 4:4:4 uncompressed video, I’d try using ‘none’ as the compressor or ‘Animation’ set to ‘100’ quality. Animation uses RLE (run length encoding) compression or something so there’s a performance hit but the best in terms of file size. Lossless is inbetween, with a mild performance hit but also has space savings. Apple Lossless is still relatively new and has much less support (e.g. It won’t work in Windows, just like DVCPro HD) so I don’t touch the stuff.