32bpc vs 16 bpc for H.264 DSLR footage

[Walter Soyka]

[jim bachalo] “I understand the advantage in working in a higher bit depth than 8bpc but is there any advantage in working in 32bpc with H.264 footage for normal comping and color correction work, or am I just wasting RAM and CPU cycles?”

32bpc uses floating-point values instead of range-limited integers to represent the RGBA values in an image, which means that values brighter than white or darker than black won’t be clipped. It may not make a huge difference relative to 16bpc for normal color correction, but it will make a huge difference for synthetic effects like glows, blurs, and lights.

See Color depth and high dynamic range color [link] for a visual example of the same blur at 32bpc and 16bpc.

Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events

[Jim Bachalo]

Thanks Walter!

Local is the new global

[Walter Soyka]

I should add that floating poing calculations have a huge effect on blend modes and adjustment layers, too. If you’re stacking a lot of layers to get a specific effect, using 32bpc (and thus avoiding clipping) will let you push on one layer and pull on another.

Walter Soyka
Principal & Designer at Keen Live
Motion Graphics, Widescreen Events, Presentation Design, and Consulting
RenderBreak Blog – What I’m thinking when my workstation’s thinking
Creative Cow Forum Host: Live & Stage Events