Aug 3, 2023, 18:26 by [email protected]: > These are in-place transforms, required for DCT-I and DST-I. > > Templated as the mod2 variant requires minor modifications, and is > required specifically for DCT-I/DST-I. > > Quite optimized, as there's no need for any additional buffer storage. >
Specifically, for R2R, vs fftw 26280 decicycles in av_tx (r2c), 1048574 runs, 2 skips 69940 decicycles in fftwf_execute, 1048576 runs, 0 skips And for R2I 25856 decicycles in av_tx (r2c), 1048571 runs, 5 skips 65561 decicycles in fftwf_execute, 1048576 runs, 0 skips _______________________________________________ ffmpeg-devel mailing list [email protected] https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email [email protected] with subject "unsubscribe".
