According to the x264 Encoding Guide, the crf scale depends on whether x264 is 8-bit or 10-bit. Supposedly x264 --help shows the Output bit depth. But on Windows ffmpeg -h full doesn't say anything about Output bit depth and the crf scale option is described as from -1 to FLT_MAX. How can I check if my ffmpeg is using 8-bit or 10-bit version of the libx264 library?
You can run ffmpeg -h encoder=libx264 and look for values of supported pixel formats.
For bit-depth=8 of libx264 it will be something like:
Supported pixel formats: yuv420p yuvj420p yuv422p yuvj422p yuv444p yuvj444p nv12 nv16
And for bit-depth=10 of libx264 it will be something like:
Supported pixel formats: yuv420p10le yuv422p10le yuv444p10le nv20le