Skip to content

Conversation

mattvenn
Copy link

@dcoredump
Copy link

Hi @mattvenn

does your PR mean that the problem also exists in sat16(n, rshift) signed_saturate_rshift((n), 16, (rshift)) because you are removing this part of the code?

Thanx, Holger

@mattvenn
Copy link
Author

mattvenn commented Sep 1, 2020

(From Neils) As far as I can remember, all the saturation routines have the same asymmetrical rounding which causes a buildup of the noise. So, yes, that code was removed on purpose.

@PaulStoffregen PaulStoffregen merged commit 5259876 into PaulStoffregen:master Feb 5, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants