Repository navigation
audiobiquad: buffer_size, so a host's small blocks are filtered once - #324
Merged
Merged
Conversation
Biquad and AllPass take buffer_size in bytes, as the CircuitPython effects do, from one frame up to the default 256. Each pull then hands out that many frames and takes exactly one block from a source of the same size, so a host running small blocks gets the filter with no latency and coefficients that can move once a block. The filter is per sample, so the audio is the default render cut smaller; renders at the default size are byte-identical to before.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Biquad and AllPass now take
buffer_sizein bytes, the same as audioladder.Ladder: whole frames, from one frame up to the default 256.Before this, a host feeding 128-frame blocks got 256 frames from each pull. The node filled them by taking the same looped block twice, so it filtered every block twice and the filter state ran ahead. At the host's own size, each pull takes exactly one source block.
Renders at the default size are byte-identical to before. The filter is per sample, so a smaller block size gives the same render, just cut smaller.
Tested: