mirror of
https://github.com/fergalmoran/ladybird.git
synced 2026-04-23 21:14:31 +00:00
As long as possible, entire decoded frame sample vectors are moved into the output vector, leading to up to 20% speedups by avoiding memmoves on take_first.