mirror of https://github.com/odrling/Aegisub
Allocate memory for fftw with fftw_alloc_real/fftw_alloc_complex rather than fftw_malloc
Originally committed to SVN as r5974.
This commit is contained in:
parent
b4e3208405
commit
5d8a802e8a
|
@ -145,8 +145,8 @@ void AudioSpectrumRenderer::RecreateCache()
|
|||
cache.reset(new AudioSpectrumCache(block_count, this));
|
||||
|
||||
#ifdef WITH_FFTW
|
||||
dft_input = (double*)fftw_malloc(sizeof(double) * (2<<derivation_size));
|
||||
dft_output = (fftw_complex*)fftw_malloc(sizeof(fftw_complex) * (2<<derivation_size));
|
||||
dft_input = fftw_alloc_real(2<<derivation_size);
|
||||
dft_output = fftw_alloc_complex(2<<derivation_size);
|
||||
dft_plan = fftw_plan_dft_r2c_1d(
|
||||
2<<derivation_size,
|
||||
dft_input,
|
||||
|
|
Loading…
Reference in New Issue