I noticed the 1080p upload on YouTube looked really soft/low detail compared to the original rendered file. I created a 4K timeline, upscaled and sharpened the video, rendered out as MXF (H.264 ALL-I, 422 300Mbps) and uploaded it again to YouTube. The 1440 "HD" looks like the original 1080p; can't see much improvement for the 4K version on YouTube over their 1440 transcode.
The 1080p version above looks the same as the original 1080p upload, here (YouTube 1080 preset in PP CC):
For the C300 II, shooting in RGB 444 12-bit provides the best possible color and "costs" 225Mbps disk space. YUV 422 10-bit 4K costs 410Mbps disk space. I suspect that "actual" 4K won't look much more detailed on YouTube vs. upscaled and post-sharpened RGB 444 based on this 1DX II 1080p vs. 4K test:
For wide shots on the C300 II (and 1DX II), 4K is certainly useful vs. 1080.
The C300 II color is using optimized custom settings (developing Filmic Skin for the C300 II). This version was not using any form of log (will have log versions too to maximize DR when needed).
The 1080p version above looks the same as the original 1080p upload, here (YouTube 1080 preset in PP CC):
For the C300 II, shooting in RGB 444 12-bit provides the best possible color and "costs" 225Mbps disk space. YUV 422 10-bit 4K costs 410Mbps disk space. I suspect that "actual" 4K won't look much more detailed on YouTube vs. upscaled and post-sharpened RGB 444 based on this 1DX II 1080p vs. 4K test:
For wide shots on the C300 II (and 1DX II), 4K is certainly useful vs. 1080.
The C300 II color is using optimized custom settings (developing Filmic Skin for the C300 II). This version was not using any form of log (will have log versions too to maximize DR when needed).