Makes output exact with smaller number of samples per frame than taps.
Showing the changes to libavfilter (1 of 1 file). Show all
Unified · Split
| @@ -145,7 +145,7 @@ static void convolve(AVFilterContext *ctx, AVFrame *in, · 144 unmodified lines | ||
| 145 | 145 | dst = scalarproduct(taps, taps + len, s->filter[filter_ch], dst); |
| 146 | 146 | |
| 147 | 147 | // process current input |
| 148 | − if (2*in->nb_samples >= NUMTAPS ){ | |
| 148 | + if (in->nb_samples >= NUMTAPS) { | |
| 149 | 149 | endin = src + in->nb_samples - NUMTAPS; |
| 150 | 150 | scalarproduct(src, endin, s->filter[filter_ch], dst); |
| 151 | 151 | |