I've been using blender for about 3 years now but I only recently built a computer capable of doing more complex renders. My renders look excellent when rendered as a png but when I render as a video I get very bad blotches in the dark areas. I assume this is from compression. I'm using h.264 and I have it set to perceptually lossless. I still get ugly banding in my objects that are black. This has limited me to only doing lighter scenes.
Right now I'm working on recreating an iphone commercial and it looks killer except the banding. I wanted a nice dark scene. (The banding only happens on the objects that are dark in color. The world is completly black and there is no blotchyness in it).
Not only does this happen in blender but in devinci resolve also. If I have a dark area in the video I get horrible blotchy areas after rendering even if I make no changes at all to the video.
I've thought about rendering as an image sequence and then converting that into a video outside of blender but I think I'd probably get the same result as just rendering as a video in blender to begin with.
I see lots of videos on YouTube that are dark scenes that have no blotchyness at all. And I've heard that youtube compresses the crap out videos when you upload. So how are these people getting such smooth dark areas even after all the compression?