mirror of https://git.ffmpeg.org/ffmpeg.git
aacdec_usac: do not round noise amplitude values
Use floating point division instead of integer division.
This commit is contained in:
parent
a18d0659f4
commit
9b41cc0430
|
@ -930,7 +930,7 @@ static void apply_noise_fill(AACDecContext *ac, SingleChannelElement *sce,
|
|||
float *coef;
|
||||
IndividualChannelStream *ics = &sce->ics;
|
||||
|
||||
float noise_val = pow(2, (ue->noise.level - 14)/3);
|
||||
float noise_val = powf(2, ((float)ue->noise.level - 14.0f)/3.0f);
|
||||
int noise_offset = ue->noise.offset - 16;
|
||||
int band_off;
|
||||
|
||||
|
|
Loading…
Reference in New Issue