Mercurial > mplayer.hg
changeset 31641:6c5a559d340c
Faster paletted to OSD conversion.
author | reimar |
---|---|
date | Sun, 11 Jul 2010 13:54:23 +0000 |
parents | 0d09c303bfe3 |
children | af4d89e131b4 |
files | spudec.c |
diffstat | 1 files changed, 12 insertions(+), 7 deletions(-) [+] |
line wrap: on
line diff
--- a/spudec.c Sun Jul 11 13:48:16 2010 +0000 +++ b/spudec.c Sun Jul 11 13:54:23 2010 +0000 @@ -1285,6 +1285,8 @@ int x, int y, int w, int h, double pts, double endpts) { + int i; + uint16_t g8a8_pal[256]; packet_t *packet; const uint32_t *pal = palette; spudec_handle_t *spu = this; @@ -1304,17 +1306,20 @@ packet->packet = malloc(packet->data_len); img = packet->packet; aimg = packet->packet + stride * h; - // TODO: this would be a lot faster by converting the - // palette first. - for (y = 0; y < h; y++) { - for (x = 0; x < w; x++) { - uint32_t pixel = pal[pal_img[x]]; + for (i = 0; i < 256; i++) { + uint32_t pixel = pal[i]; int alpha = pixel >> 24; int gray = (((pixel & 0x000000ff) >> 0) + ((pixel & 0x0000ff00) >> 7) + ((pixel & 0x00ff0000) >> 16)) >> 2; - *aimg++ = -alpha; - *img++ = FFMIN(gray, alpha); + gray = FFMIN(gray, alpha); + g8a8_pal[i] = (-alpha << 8) | gray; + } + for (y = 0; y < h; y++) { + for (x = 0; x < w; x++) { + uint16_t pixel = g8a8_pal[pal_img[x]]; + *img++ = pixel; + *aimg++ = pixel >> 8; } for (; x < stride; x++) *aimg++ = *img++ = 0;