After scouring the documentation further I came across the NSOpenGLPFAColorFloat
attribute, which according to the documentation:
A Boolean attribute. If present, this attribute indicates that only renderers that are capable using buffers storing floating point pixels are considered. This should be accompanied by a NSOpenGLPFAColorSize of 64 (for half float pixel components) or 128 (for full float pixel components). Note, not all hardware supports floating point color buffers thus the returned pixel format could be NULL.
With that additional information it must mean bits per pixel.
I did some experimenting as well, setting NSOpenGLPFAColorSize
to each of 8, 16, 24 & 32 and then checking what I got back. In each case I was returned a pixel format with NSOpenGLPFAColorSize
set to 32 - meaning 32-bits per RGBA pixel. Just passing NSOpenGLPFAColorFloat
with nothing set for the Color Size is enough to get back a pixel format with 64-bits per pixel.