Welcome to TVPaint Community forums !

So I see that I have to change my Photoshop file from 16 bit to 8 bit in order to import them into TVPaint with layers. Two questions: Does changing them to 8 bit compromise the image quality in any appreciable way? Would it be better to save each layer as a separate file and then manually load each layer? Why doesn't TVPaint load 16 bit?

Because TVP only works with 8bit per colour channel.

And IMHO this is enough.
1. You can't see a difference between an image done in 16bit converted to 8bit with your naked eye anyway.
2. You don't have a monitor which is calibrated for 16bit, so it can't display enough different shades anyway.
3. The 16bit are mostly a reserve for image calculating - in the end it's 8bit (print), 10bit (some TV) or 12bit (DCP) at most anyway.
4. Animated characters are never coloured with a subtlety which needs 16bits for correct representation.
5. Only in nearly black and nearly white parts of an image the 16bit really matter, and this is the job of a compositing program anyway.
6. If you can't maintain the 16bit resolution throughout all of your workflow, there's no use to start with it.

(I could provide a more detailed explanation, buut will spare you that, as long as nobody asks for it.)

This is what I was interested to know, the real world application of 8 or 16 bit. Thanks.

In some cases 16bit matters: Probably not for flat color, but if you got a pretty subtle gradient for example, you're getting banding pretty fast.

(pretty extreme example, but you get the idea, what I'm talking about)

Ha - you may tweak and optimize your gradient as much as you will, but in the end it will look only as good as the screening format permits - digital TV uses heavy compression (different for different TV channels), so you will not even get nice clean banding but ugly blocky compression artefacts.

But of course you're right, flat gradients look better in greater bit depth. Though in real world one would add a certain amount of noise or grain instead.

Yea, the final format is usually 8bit (in DCP cases 12bit), but if I use subtle gradients I tend to use 16bit while working on it and put some slight grain or noise on top of the whole film before I convert it to 8bit. So there are some cases where 16bit makes sense, but they're pretty rare, that's right.

Even if it's just 1 scene asking for 16bit, it still doesn't make much sense if the following steps in the pipeline don't recognize it. Is my editing software up to the task? I just researched a bit and found that Final Cut X can import images up to 16bit, and it's possible to export 16bit from there. (See https://images.apple.com/final-cut-pro/ ... _Gamut.pdf for some workflow tips, especially how to set up FC first.)