I mostly use Virtualdub2 to upscale my videos (although I am looking for other methods) via two resize filters, Nearest Neighbor and Lanzcos3. Recently I am looking at the x264 encoder to encode my videos, I am not sure which is better for pixel art as I mostly deal with pixel art from video game dumps. Can anyone tell me which version x264 is good for pixel art? Which is better at preserving details and which is not?
It depends on the bit depth of the source. If the source material is greater than 8 bits, then 10 bit will give you more color fidelity and less banding in subtle gradients (if there is any in your “pixel art.” Flat cartoons wouldn’t have any gradient detail to preserve, e.g.) If the source is only 8 bits per channel, 10 bit won’t give you any extra detail.
Since a bit is binary (base 2), 8 bit is 2^8, or 256 shades of each color in a color channel. So 256 R * 256 B * 256 G, or roughly 16 million colors. The human eye can perceive more than that, which is why 10 bit color depth is one of the requirements for HDR.
Most consumer displays are also only capable of displaying 8 bit color, so consider your audience, also.
Most average viewers won’t notice the difference between 8 and 10, even if they’re using a 10 bit display. Where that difference becomes mostly apparent is when you’re developing the images; higher bit depths give colorists, editors and VFX artists more numeric/artistic headroom, which is the same reason Photoshop artists avoid 8 bit originals.
So, if you don’t plan to do any further processing, 8 bit is fine.