Is it possible to reuse the same statistics file generated by the first pass of a 2-pass encode when doing multiple second pass encode of different bitrates/crf values? I want to create several encodings of my video, and I would like to avoid redoing the first pass analysis phase each time if this is allowed. I tried it and it seemed to work, but I'm not sure whether something is going subtly wrong under the hood, or whether it might break entirely in a future revision.
Here's a practical example of what I want to do
First pass analysis
ffmpeg -i input.mp4 -c:v vp9 -b:v 1M -crf 20 -deadline good -cpu-used 0 -auto-alt-ref 1 -lag-in-frames 16 -pass 1 -passlogfile reusedlog -f webm -y /dev/null
Second pass encodes that all share the same logfile
ffmpeg -i input.mp4 -c:v vp9 -b:v 2M -crf 20 -deadline good -cpu-used 1 -pass 2 -passlogfile reusedlog highquality.webm
ffmpeg -i input.mp4 -c:v vp9 -b:v 1M -crf 30 -deadline good -cpu-used 1
-pass 2 -passlogfile reusedlog lowquality.webm
Is this safe? Is the first pass analysis independent of my requested encoder settings, and thus safe to reuse for several different encoder settings in the second phase?
Thanks for any insight you can offer.