Hacker News new | ask | show | jobs
by Archit3ch 29 days ago
> the following audio processing is better done after converting the samples to FP32

Or, in some cases, FP64.