Hi, On Thu, Jul 7, 2016 at 5:25 AM, Alexandra Hájková < alexandra.khirn...@gmail.com> wrote:
> else if (lc->cu.pred_mode == MODE_INTRA && c_idx == 0 && > log2_trafo_size == 2) > - s->hevcdsp.transform_4x4_luma_add(dst, coeffs, stride); > + s->hevcdsp.idct_4x4_luma(coeffs); > This is not an idct. > + s->hevcdsp.add_residual[log2_trafo_size - 2](dst, coeffs, stride); Won't this be slower since there's a memory store intermediate? (I know it's faster now because you don't have inverse transform simd, but you should fix that by writing inverse transform simd, not by splitting the transform and the add.) Ronald _______________________________________________ libav-devel mailing list libav-devel@libav.org https://lists.libav.org/mailman/listinfo/libav-devel