[FFmpeg-devel] [PATCH 0/5] mlpenc: fix lossless check failures for 16-bit encoding

[Jai Luthra]

This patchset fixes lossless check failures for 16-bit mlp/truehd encoding. 
Please check with more samples and let me know if you find any lossless 
failures.

Bug 6216 https://trac.ffmpeg.org/ticket/6216 still cannot be closed as last 
few frames aren't encoded, thus output is not bit-exact.
I will continue on the WIP patch Paul posted earlier and try to fix this.

32-bit encoding still has many lossless failures, so keeping it disabled for 
now. Will work on it later.

Jai Luthra (5):
  mlpenc: fix lossless check error in number_sbits
  mlpenc: fix huff offset calculation
  mlpenc: prevent negative lsb_bits lshift
  mlpenc: improve lpc filtering
  mlpenc: clean up

 libavcodec/mlpenc.c | 74 +++++++++++++++++++++------------------------
 1 file changed, 35 insertions(+), 39 deletions(-)

-- 
2.25.0