Encoding settings to avoid lose of quality in desktop recording

[origliante]

I record the desktop with a lossless codec (FSV1) and would like to achieve a similar quality when encoding with x264. Below you can see the images of the original capture and the one from the encoding

Original

[Attachment 68651 - Click to enlarge]

Encoded

[Attachment 68652 - Click to enlarge]

As You can see, the icons are not as crisp and vivid like the original capture. These are my Handbrake settings

[Attachment 68654 - Click to enlarge]

[jagabo]

Video is normally encoded as YV12 (YUV 4:2:0). That means the chroma channels are half the width and height of the luma channel. For example, a 1920x1080 video will have a 1920x1080 luma channel (greyscale) and two 960x540 chroma channels (colors that are added or subtracted from the luma to produce the final on-screen colors). The reduced chroma resolution will always blur sharp colored edges like that.

You can avoid that by encoding without chroma subsampling, using YV24 (YUV 4:4:4) or RGB (4:4:4). But whether that is suitable depends on your application. If you want videos that play on most devices you cannot use those.

[origliante]

Video is normally encoded as YV12 (YUV 4:2:0). That means the chroma channels are half the width and height of the luma channel. For example, a 1920x1080 video will have a 1920x1080 luma channel (greyscale) and two 960x540 chroma channels (colors that are added or subtracted from the luma to produce the final on-screen colors). The reduced chroma resolution will always blur sharp colored edges like that.

You can avoid that by encoding without chroma subsampling, using YV24 (YUV 4:4:4) or RGB (4:4:4). But whether that is suitable depends on your application. If you want videos that play on most devices you cannot use those.

Thanks for the explanation