Hi Edwin,

Thank you for your interest.

[multiconvolve~] combines time domain convolution for the early portion of an IR with more efficient FFT-based partitioned convolution for the latter parts of the IR[1]. Moreover SIMD instructions are used in the FFT-convolution scheme.

Best,
Marco

[1]: http://eprints.hud.ac.uk/id/eprint/14897/1/HISSTools_IR_Toolbox.pdf.pdf

Il giorno dom 11 apr 2021 alle ore 19:02 Edwin van der Heide <pd@evdh.net> ha scritto:
Dear Marco,

Very nice, I’l do some testing soon!

Does multiconvolve~ use a combination of different FFT sizes in order to achieve both low latency output and a cpu-efficient form of partitioned convolution?

Best!

Edwin

On 11 Apr 2021, at 17:25, Marco Matteo Markidis <mm.markidis@gmail.com> wrote:

Dear all,

I would like to announce the release of *piro*[1].
Piro is a Pd-port of an impulse-response measurement object and a zero latency convolver from the HISSTools Impulse Response Toolbox (HIRT). Moreover in the package it is included a general-purpose object that contains several utilities for managing IRs; this is still in progress.
Source code and binary for major architectures are provided.
For any questions about the Pd version of these objects please do not hesitate to contact me.
Moreover in the readme there is a link for the original project, where you can find an article that describes the original package and its utilities.

Finally I would like to thank Lucas Cordiviola and Giuseppe Silvi for their patience with this project.

Best,
Marco

_______________________________________________
Pd-announce mailing list
Pd-announce@lists.iem.at
https://lists.puredata.info/listinfo/pd-announce
_______________________________________________
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list