comparison libmpcodecs/vf_ass.c @ 32261:0dd2b27940b5

vf_ass: avoid divisions in alpha blending. On a test case, it speeds up the blending by about 20%.

32260:56dadca4f36d

                           int dst_x, int dst_y, unsigned color)
{
    unsigned char y = rgba2y(color);
    unsigned char u = rgba2u(color);
    unsigned char v = rgba2v(color);
    unsigned char opacity = 255 - _a(color);
    unsigned char *src, *dsty, *dstu, *dstv;
    int i, j;
    mp_image_t *dmpi = vf->dmpi;

    src = bitmap;
    dsty = dmpi->planes[0]     + dst_x + dst_y * dmpi->stride[0];
    dstu = vf->priv->planes[1] + dst_x + dst_y * vf->priv->outw;
    dstv = vf->priv->planes[2] + dst_x + dst_y * vf->priv->outw;
    for (i = 0; i < bitmap_h; ++i) {
        for (j = 0; j < bitmap_w; ++j) {
            unsigned k = src[j];
            if (!k)
                continue;
            k = k * opacity / 255;
            dsty[j] = (k * y + (255 - k) * dsty[j]) / 255;
            dstu[j] = (k * u + (255 - k) * dstu[j]) / 255;
            dstv[j] = (k * v + (255 - k) * dstv[j]) / 255;
        }
        src  += stride;
        dsty += dmpi->stride[0];
        dstu += vf->priv->outw;
        dstv += vf->priv->outw;

32261:0dd2b27940b5

                           int dst_x, int dst_y, unsigned color)
{
    unsigned char y = rgba2y(color);
    unsigned char u = rgba2u(color);
    unsigned char v = rgba2v(color);
    unsigned opacity = 255 - _a(color);
    unsigned char *src, *dsty, *dstu, *dstv;
    int i, j;
    mp_image_t *dmpi = vf->dmpi;

    opacity = (0x10203 * opacity + 0x80) >> 8; /* 0x10203 = (1<<32)/(255*255) */
    /* 0 <= opacity <= 0x10101 */
    src = bitmap;
    dsty = dmpi->planes[0]     + dst_x + dst_y * dmpi->stride[0];
    dstu = vf->priv->planes[1] + dst_x + dst_y * vf->priv->outw;
    dstv = vf->priv->planes[2] + dst_x + dst_y * vf->priv->outw;
    for (i = 0; i < bitmap_h; ++i) {
        for (j = 0; j < bitmap_w; ++j) {
            unsigned k = src[j];
            if (!k)
                continue;
            k *= opacity; /* 0 <= k <= 0xFFFFFF */
            dsty[j] = (k * y + (0xFFFFFF - k) * dsty[j] + 0x800000) >> 24;
            dstu[j] = (k * u + (0xFFFFFF - k) * dstu[j] + 0x800000) >> 24;
            dstv[j] = (k * v + (0xFFFFFF - k) * dstv[j] + 0x800000) >> 24;
        }
        src  += stride;
        dsty += dmpi->stride[0];
        dstu += vf->priv->outw;
        dstv += vf->priv->outw;