On Wed, Sep 28, 2011 at 09:52:25PM +0100, Mans Rullgard wrote: > Writing zeros to the high entries in the array need only be > done once as the cutoff position is constant throughout the > loop. > > Signed-off-by: Mans Rullgard <[email protected]> > --- > libavcodec/dca.c | 5 +++-- > 1 files changed, 3 insertions(+), 2 deletions(-) > > diff --git a/libavcodec/dca.c b/libavcodec/dca.c > index db794a7..735d7ba 100644 > --- a/libavcodec/dca.c > +++ b/libavcodec/dca.c > @@ -898,6 +898,9 @@ static void qmf_32_subbands(DCAContext * s, int chans, > else /* Perfect reconstruction */ > prCoeff = fir_32bands_perfect; > > + for (i = sb_act; i < 32; i++) > + s->raXin[i] = 0.0; > + > /* Reconstructed channel sample index */ > for (subindex = 0; subindex < 8; subindex++) { > /* Load in one sample from each subband and clear inactive subbands > */ > @@ -905,8 +908,6 @@ static void qmf_32_subbands(DCAContext * s, int chans, > uint32_t v = AV_RN32A(&samples_in[i][subindex]) ^ ((i-1)&2)<<30; > AV_WN32A(&s->raXin[i], v); > } > - for (; i < 32; i++) > - s->raXin[i] = 0.0; > > s->synth.synth_filter_float(&s->imdct, > s->subband_fir_hist[chans], > &s->hist_index[chans], > --
Since filter input is const, OK _______________________________________________ libav-devel mailing list [email protected] https://lists.libav.org/mailman/listinfo/libav-devel
