Encode with YUV420P, to be more compatible

[TobiX]

This fixes #256.

This is still suboptimal (IMHO), since we encode twice with the same options... Wouldn't it be better to do an "almost-lossless" encode while doing the x11grab (with a fixed set of options, perhaps -crf 0 -preset ultrafast -qp 0) and letting the user decide only the encoding options of the second step (via the FFMPEG_CODEC_ARGS option) - WDYT? If that is something desired, I can prepare a patch set with those changes.

Difference in ffprobe is now:

-    Stream #0:0(und): Video: h264 (High 4:4:4 Predictive) (avc1 / 0x31637661), yuv420p, 1900x1880 [SAR 1:1 DAR 95:94], 4161 kb/s, 10 fps, 10 tbr, 10240 tbn, 20 tbc (default)
+    Stream #0:0(und): Video: h264 (High 4:4:4 Predictive) (avc1 / 0x31637661), yuv444p, 1900x1880 [SAR 1:1 DAR 95:94], 5111 kb/s, 10 fps, 10 tbr, 10240 tbn, 20 tbc (default)

Profile is still High 4:4:4 Predictive - I think this is forced by the -qp 0 option.

PS: I saw the FFMPEG_FINAL_CRF=0 - is that still used anywhere, I didn't saw it used in the repository?

[CLAassistant]

All committers have signed the CLA.

[elgalu]

Thanks for the PR @TobiX !

I think maybe FFMPEG_FINAL_CRF was used at some point but for sure not anymore. Feel free to remove it on any upcoming PR.

letting the user decide only the encoding options of the second step

Yes I like also this approach. I will merge and release this change but feel free to send a new PR with the additional improvements.