Prevent distortion filter from introducing NaNs in the audio buffer #46199
Conversation
This file contains 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
Calinou
added
bug
cherrypick:3.x
reduz
approved these changes
Feb 19, 2021
|
Thanks! |
|
Cherry-picked for 3.2.4. |
akien-mga
removed
the
cherrypick:3.x
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Outdated suggestions cannot be applied.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.
Fixes #38372
Currently the base argument passed to
powf()is negative sometimes. I don't understand the math here well enough to catch what that means conceptually or if it's indicative of a larger problem, but I do understand that for non-integer exponents that's going to causepowf()to NaN, which propagates a discontinuity in the output beyond just this filter. I tried this change and it works well. Distorted audio sounds nice and distorted, and undistorted audio also sounds clear since there aren't NaNs propagating through.Notably, I tried to just clamp the value of
abefore thepowf()step to the range [0,inf), but the distortion didn't sound good and the parameters applied to the filter didn't seem like they were changing things much. I dug a bit deeper and noticed that theundenormalisestep if I understand it correctly seems to have an output range of (-inf,-1] and [1,inf). So I figured that sign was important and decided to try passing the sign bit through and only using the magnitude of the number forpowf. This sounds great, but will definitely need a closer look by someone who knows what this filter is actually doing (reduz?).Test using the minimal repro project for #38372.