-
Color Bit Depth for Film: When to choose 8bit vs 16 vs 32 bit?
Forgive me for my newbie question but can someone please help me to understand why I would choose to work in either 8-bit, 16-bit or 32bit (floating point) when creating compositions in After Effects. Especially when most of the video I’m importing or exporting is 8, 10, 12 bit etc. Why are these the standard export options compared to compositing in AE? Would working in higher bit depths and then exporting back to 8bit h.264 for example not just reduce the quality to the same state as if I worked in 8bit to begin with while animating?
So far what I’ve gathered is that 8 bit video can create banding artifacts, 16 bit will help with this and 32 bit allows you even greater color precision because it can create varying values between integers for finer/smoother transitions.
I’m not really sure when use each or why they’re better for certain tasks/situations and would greatly appreciate any direction.
-
4 Replies
-
[Nathan Scofield] “Would working in higher bit depths and then exporting back to 8bit h.264 for example not just reduce the quality to the same state as if I worked in 8bit to begin with while animating?”
The compositing process essentially performs mathematical operations on color values. If you perform math using more accurate values your results will be more accurate.
The outputs are the results encoded. With lower bit depths/accuracy while compositing errors can end up summing and multiplying to produce greater errors and as a result the output will be of a lower quality. The more math is being done during compositing the more potential error will be introduced into the output at lower bit depths.
Darby Edelen
-
Thank you Darby, that was helpful and I follow what you’re saying. I’m still unsure as to when to choose one option over another though. Why not always go for 32-bit when working in AE? I assume to save file size and or rendering time but are there any standard operations or situations in which one would choose 8 or 16 or 32 to work in?
-
[Nathan Scofield] “Why not always go for 32-bit when working in AE? I assume to save file size and or rendering time but are there any standard operations or situations in which one would choose 8 or 16 or 32 to work in?”
8bpc and 16bpc work similarly. They clip at black and white. 8bpc gives you a range of integers from 0-255 shades to express the brightness of a channel; 16bpc gives you a range of 0-32768 shades to express the brightness of a channel. 16bpc is just a more precise version. (Do two opposing Levels adjustments to the same image in 8bpc and 16bpc, see the difference in the histogram.)
32bpc works completely differently. It’s floating point, and it’s unbounded. Black is 0 and white is 1 with huge precision in between, and the possibility to go outside that normal range (with eventually decreasing precision).
This allows colors that are darker than black and brighter than white. This has a huge impact on how things look — especially things like blurs and glows, but potentially any stack of effects which may clip at a lower bit depth. With floating point color, you can blow out highlights with one effect and recover them with a subsequent effect.
See here for a lot more:
https://helpx.adobe.com/after-effects/using/color-basics.html#color_depth_and_high_dynamic_range_colorWalter Soyka
Designer & Mad Scientist at Keen Live [link]
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
@keenlive | RenderBreak [blog] | Profile [LinkedIn]
Reply to this Discussion! Login or Sign Up