• Rostislav Pehlivanov's avatar
    aaccoder: tweak PNS implementation further · da64bd6a
    Rostislav Pehlivanov authored
    This commit changes a few things about the noise substitution
    logic:
     - Brings back the quantization factor (reduced to 3) during
       scalefactor index calculations.
     - Rejects any zeroed bands. They should be inaudiable and it's
       a waste transmitting the scalefactor indices for these.
     - Uses swb_offsets instead of incrementing a 'start' with every
       window group size.
     - Rejects all PNS during short windows.
    Overall improves quality. There was a plan to use the lfg system
    to create the random numbers instead of using whatever the decoder
    uses but for now this works fine. Entropy is far from important here.
    Signed-off-by: 's avatarRostislav Pehlivanov <atomnuker@gmail.com>
    da64bd6a
Name
Last commit
Last update
..
api Loading commit data...
checkasm Loading commit data...
fate Loading commit data...
filtergraphs Loading commit data...
ref Loading commit data...
Makefile Loading commit data...
audiogen.c Loading commit data...
base64.c Loading commit data...
copycooker.sh Loading commit data...
fate-run.sh Loading commit data...
fate-valgrind.supp Loading commit data...
fate.sh Loading commit data...
ffserver-regression.sh Loading commit data...
ffserver.conf Loading commit data...
ffserver.regression.ref Loading commit data...
lavf-regression.sh Loading commit data...
md5.sh Loading commit data...
reference.pnm Loading commit data...
regression-funcs.sh Loading commit data...
rotozoom.c Loading commit data...
test.ffmeta Loading commit data...
tiny_psnr.c Loading commit data...
tiny_ssim.c Loading commit data...
utils.c Loading commit data...
videogen.c Loading commit data...