diff options
| author | Chris Robinson <[email protected]> | 2018-01-07 05:32:07 -0800 |
|---|---|---|
| committer | Chris Robinson <[email protected]> | 2018-01-07 05:32:07 -0800 |
| commit | 4cc1c646466737ba411aa23ce4a6116936ada8c2 (patch) | |
| tree | 796fc698eb910630ac5f398fe38f26aef2cd456c /alsoftrc.sample | |
| parent | 0e1fd34c89d8f09f68c2c243ceccd0dab4f7c6c0 (diff) | |
Replace the sinc4 resampler with cubic
Turns out the C version of the cubic resampler is just slightly faster than
even the SSE3 version of the FIR4 resampler. This is likely due to not using a
64KB random-access lookup table along with unaligned loads, both offseting the
gains from SSE.
Diffstat (limited to 'alsoftrc.sample')
| -rw-r--r-- | alsoftrc.sample | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/alsoftrc.sample b/alsoftrc.sample index 41c884ad..68d45220 100644 --- a/alsoftrc.sample +++ b/alsoftrc.sample @@ -149,7 +149,7 @@ # Selects the resampler used when mixing sources. Valid values are: # point - nearest sample, no interpolation # linear - extrapolates samples using a linear slope between samples -# sinc4 - extrapolates samples using a 4-point Sinc filter +# cubic - extrapolates samples using a Catmull-Rom spline # bsinc12 - extrapolates samples using a band-limited Sinc filter (varying # between 12 and 24 points, with anti-aliasing) # bsinc24 - extrapolates samples using a band-limited Sinc filter (varying |