I have done a short test to compare and understand the quality factor in h.264 export. I actually used 54 sec. of video footage in FHD h.264 as a basis with some audio and exported this in different quality settings starting from 10% up to 100%. I used the YT-preset with unchanged parameters beside the quality (in the advanced - codec settings).
Here are my results:
Quality___ encoding time____file size__ artefacts
10%______38 s______________9.3 k__very massive artefacts, blocking, color flickering, loss of detail
20%______44 s_____________13.8 k__lots of visible blocking, color flickering
30%______49 s_____________24.8k__less visible artefacts, but still obvious
40%______53 s ____________ 44.3k__slightly visible artefacts
50%______58 s_____________85.8k__hardly visible to no artefacts
60%______78 s____________196.3k__no visible artefacts (for me on laptop screen)
70%______83 s____________372.2k__no visible artefacts
80%______95 s____________598 k___no visible artefacts
90%_____112 s____________921 k___no visible artefacts
100%____116 s___________1585 k___no artefacts
Best quality to file size ratio is somewhere between 50…60% i would assume. For me, already at 40% there are hardly any visible obvious artefacts. Over 50% you can be quite sure you wont see any artefacts unless you have an eagle’s eyes
More than 70% quality is a disk space waste unless you absolutely need 100% quality (for a later re-encoding e.g.)
As a rule of thumb (except for the extreme ends) you can approx. say 10% more quality means a doubling in file size.
Note: all measurement were done on a Thinkpad P50 laptop with render preset=“fast”. You can set the render preset on the EXPORT - ADVANCED - OTHER tab. “fast” was the default here, you can set it also to “veryfast” (or “medium”) which decreases render times about 30% on the cost of output quality (at the same quality setting).