virtualenv/FFmpeg
1
0
Fork 0
mirror of https://github.com/FFmpeg/FFmpeg.git synced 2025-08-04 22:03:09 +02:00
Code Issues Releases Activity

avfilter/vf_premultiply: use correct premultiplication formula

Browse Source 
The previous formula was introduced without justification in 6e713841e8,
and the only thing Paul had to say about it over IRC was that it was copied
from an unspecified source on the internet.

I decided to do some testing and came to the conclusion that this term not
only produces "illegal" files, but also lowers PSNR score, over the naive
implementation without this extra term.

Here are the results of a round-trip test, using allrgb/allyuv (respectively)
as the input, and fade=alpha=yes:n=256 to cycle through every possible alpha
value, comparing the round-trip output against the input:

Before patch:
  PSNR r:26.677431 g:26.677431 b:26.677431 a:inf average:27.926818 min:6.012093 max:55.400791
  PSNR y:26.677431 u:21.101981 v:21.101981 a:inf average:23.548981 min:9.013835 max:53.182303 (full)
  PSNR y:27.348055 u:21.101981 v:21.101981 a:inf average:23.625238 min:9.554991 max:45.652221 (limited)

After patch:
  PSNR r:27.321996 g:27.321996 b:27.321996 a:inf average:28.571384 min:6.012093 max:52.424553
  PSNR y:27.321996 u:23.187879 v:23.187879 a:inf average:25.431773 min:9.013835 max:50.199232 (full)
  PSNR y:27.868544 u:23.187879 v:23.187879 a:inf average:25.515660 min:9.554991 max:45.078298 (limited)

It's worth pointing out that previous version sometimes artificially inflates
PSNR by producing values that are too high (i.e. RGB > A), such as for the
input pair (R = 255, A = 2) which should give R = 2, but actually gives R = 3
under the old logic.

As a second evaluation without this shortcoming, here is a comparison against
the reference value computed with a floating point format:

Before patch:
  PSNR r:53.600599 g:53.957833 b:53.540948 a:inf average:54.945316 min:50.508901 max:inf (premul only)
  PSNR r:30.734183 g:30.734183 b:30.734183 a:inf average:31.983570 min:12.058264 max:inf (round-trip)

After patch:
  PSNR r:61.751104 g:65.239091 b:61.339191 a:inf average:63.710714 min:55.441130 max:inf (premul only)
  PSNR r:32.611851 g:32.611851 b:32.611851 a:inf average:33.861238 min:12.058264 max:inf (round-trip)
This commit is contained in:
Niklas Haas
2025-07-22 12:35:38 +02:00
parent e74dbb02eb
commit 603334a043
1 changed files with 6 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
libavfilter/vf_premultiply.c
View File

@ -103,7 +103,7 @@ static void premultiply8(const uint8_t *msrc, const uint8_t *asrc,
for (y = 0; y < h; y++) {
for (x = 0; x < w; x++) {
dst[x] = ((msrc[x] * (((asrc[x] >> 1) & 1) + asrc[x])) + 128) >> 8;
dst[x] = (msrc[x] * asrc[x] + 128) >> 8;
}
dst += dlinesize;
@ -123,7 +123,7 @@ static void premultiply8yuv(const uint8_t *msrc, const uint8_t *asrc,
for (y = 0; y < h; y++) {
for (x = 0; x < w; x++) {
dst[x] = ((((msrc[x] - 128) * (((asrc[x] >> 1) & 1) + asrc[x]))) >> 8) + 128;
dst[x] = (((msrc[x] - 128) * asrc[x]) >> 8) + 128;
}
dst += dlinesize;
@ -143,7 +143,7 @@ static void premultiply8offset(const uint8_t *msrc, const uint8_t *asrc,
for (y = 0; y < h; y++) {
for (x = 0; x < w; x++) {
dst[x] = ((((msrc[x] - offset) * (((asrc[x] >> 1) & 1) + asrc[x])) + 128) >> 8) + offset;
dst[x] = ((((msrc[x] - offset) * asrc[x]) + 128) >> 8) + offset;
}
dst += dlinesize;
@ -166,7 +166,7 @@ static void premultiply16(const uint8_t *mmsrc, const uint8_t *aasrc,
for (y = 0; y < h; y++) {
for (x = 0; x < w; x++) {
dst[x] = ((msrc[x] * (((asrc[x] >> 1) & 1) + asrc[x])) + half) >> shift;
dst[x] = (msrc[x] * asrc[x] + half) >> shift;
}
dst += dlinesize / 2;
@ -189,7 +189,7 @@ static void premultiply16yuv(const uint8_t *mmsrc, const uint8_t *aasrc,
for (y = 0; y < h; y++) {
for (x = 0; x < w; x++) {
dst[x] = ((((msrc[x] - half) * (int64_t)(((asrc[x] >> 1) & 1) + asrc[x]))) >> shift) + half;
dst[x] = (((msrc[x] - half) * (int64_t)asrc[x]) >> shift) + half;
}
dst += dlinesize / 2;
@ -212,7 +212,7 @@ static void premultiply16offset(const uint8_t *mmsrc, const uint8_t *aasrc,
for (y = 0; y < h; y++) {
for (x = 0; x < w; x++) {
dst[x] = ((((msrc[x] - offset) * (int64_t)(((asrc[x] >> 1) & 1) + asrc[x])) + half) >> shift) + offset;
dst[x] = ((((msrc[x] - offset) * (int64_t)asrc[x]) + half) >> shift) + offset;
}
dst += dlinesize / 2;