>>4308031Z6ii probably line skips or at least is definitely not downsampled. Quick way to see is using DPReview's video comparison tool and looking at the Siemens star targets. Jagged lines/false color/moire are the giveaways.
Downsampling is a very electronically intensive process since you have to capture the whole sensor and then resize it down dozens of times per second. Upping the frame rate makes it even harder so skipping pixels was and often still is a common way to achieve that. Attached is a sample from the GH5 which downsamples and in its high frame rate mode so you have an idea of the difference. Also included is the A6400 which line skips but in high frame rate line skips even more.
Years ago, video resolution was just a spec sheet checkmark so they could plaster "FullHD" in marketing materials. The HD video from the older dSLRs (and the Sony's I mentioned in the earlier post) is a clownshow. It makes sense since companies were targeting photographers and if you wanted good quality HD, Canon wanted you buying the XC10 and not a Rebel. But a lot of buyers just going by the spec sheets would be disappointed. Panasonic usually avoided this in 1080p which is probably a reason why m43 had a strong following for years. Nowadays, downsampled 4K is common so use that if you can for 1080p.
Lots of factors involved in final quality. Bigger colorspace is better e.g. 4:4:4>4:2:2>4:2:0; higher bitrates are better; RAW > advanced codecs > primitive codecs; downsampling from 8K > 6K >5.3K > no downsampling > low to medium to high line skipping (for 4K output anyway. 2x resolution eliminates Bayer softening with more being introduced as you move down to no downsampling).
The crop factor falls under "no downsampling" and is worth thinking about since if you are losing sensor area a smaller format is worth looking at. E.g. the m43 LX10 cropped to smaller than the RX100 in video, EOS R cropped to near m43 size but with m43 you could get IBIS nothing else could touch.