FLAMO: An Open-Source Library for Frequency-Domain Differentiable Audio Processing

Gloria Dal Santo; Gian Marco De Bortoli; Karolina Anna Prawda; Sebastian Schlecht; Vesa Välimäki

doi:10.1109/ICASSP49660.2025.10888532

FLAMO: An Open-Source Library for Frequency-Domain Differentiable Audio Processing

Gloria Dal Santo^*, Gian Marco De Bortoli, Karolina Anna Prawda, Sebastian Schlecht, Vesa Välimäki

^*Corresponding author for this work

Electronic Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

We present FLAMO, a Frequency-sampling Library for Audio-Module Optimization designed to implement and optimize differentiable linear time-invariant audio systems. The library is open-source and built on the frequency-sampling filter design method, allowing for the creation of differentiable modules that can be used stand-alone or within the computation graph of neural networks, simplifying the development of differentiable audio systems. It includes predefined filtering modules and auxiliary classes for constructing, training, and logging the optimized systems, all accessible through an intuitive interface. Practical application of these modules is demonstrated through two case studies: the optimization of an artificial reverberator and an active acoustics system for improved response coloration.

Original language	English
Title of host publication	ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Publisher	IEEE
Number of pages	5
ISBN (Print)	979-8-3503-6874-1
DOIs	https://doi.org/10.1109/ICASSP49660.2025.10888532
Publication status	Published - 7 Mar 2025
Event	IEEE International Conference on Acoustics, Speech and Signal Processing - Hyderabad, India Duration: 6 Apr 2025 → 11 Apr 2025 https://2025.ieeeicassp.org/

Conference

Conference	IEEE International Conference on Acoustics, Speech and Signal Processing
Abbreviated title	ICASSP
Country/Territory	India
City	Hyderabad
Period	6/04/25 → 11/04/25
Internet address	https://2025.ieeeicassp.org/

Bibliographical note

Keywords

audio processing
delay networks
optimization
machine learning

Access to Document

10.1109/ICASSP49660.2025.10888532

Accepted manuscript
© IEEE 2025. This is an author-produced version of the published paper. Uploaded in accordance with the University’s Research Publications and Open Access policy.
Accepted author manuscript, 4.45 MBLicence: CC BY

Cite this

@inproceedings{a39852b7bbff49579c7b90ada9f66af1,

title = "FLAMO: An Open-Source Library for Frequency-Domain Differentiable Audio Processing",

abstract = "We present FLAMO, a Frequency-sampling Library for Audio-Module Optimization designed to implement and optimize differentiable linear time-invariant audio systems. The library is open-source and built on the frequency-sampling filter design method, allowing for the creation of differentiable modules that can be used stand-alone or within the computation graph of neural networks, simplifying the development of differentiable audio systems. It includes predefined filtering modules and auxiliary classes for constructing, training, and logging the optimized systems, all accessible through an intuitive interface. Practical application of these modules is demonstrated through two case studies: the optimization of an artificial reverberator and an active acoustics system for improved response coloration.",

keywords = "audio processing, delay networks, optimization, machine learning",

author = "{Dal Santo}, Gloria and {De Bortoli}, {Gian Marco} and Prawda, {Karolina Anna} and Sebastian Schlecht and Vesa V{\"a}lim{\"a}ki",

note = "{\textcopyright} IEEE 2025. This is an author-produced version of the published paper. Uploaded in accordance with the University{\textquoteright}s Research Publications and Open Access policy.; IEEE International Conference on Acoustics, Speech and Signal Processing , ICASSP ; Conference date: 06-04-2025 Through 11-04-2025",

year = "2025",

month = mar,

day = "7",

doi = "10.1109/ICASSP49660.2025.10888532",

language = "English",

isbn = "979-8-3503-6874-1",

booktitle = "ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)",

publisher = "IEEE",

address = "United States",

url = "https://2025.ieeeicassp.org/",

}

Dal Santo, G, De Bortoli, GM, Prawda, KA, Schlecht, S & Välimäki, V 2025, FLAMO: An Open-Source Library for Frequency-Domain Differentiable Audio Processing. in ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP). IEEE, IEEE International Conference on Acoustics, Speech and Signal Processing , Hyderabad, India, 6/04/25. https://doi.org/10.1109/ICASSP49660.2025.10888532

TY - GEN

T1 - FLAMO: An Open-Source Library for Frequency-Domain Differentiable Audio Processing

AU - Dal Santo, Gloria

AU - De Bortoli, Gian Marco

AU - Prawda, Karolina Anna

AU - Schlecht, Sebastian

AU - Välimäki, Vesa

N1 - © IEEE 2025. This is an author-produced version of the published paper. Uploaded in accordance with the University’s Research Publications and Open Access policy.

PY - 2025/3/7

Y1 - 2025/3/7

N2 - We present FLAMO, a Frequency-sampling Library for Audio-Module Optimization designed to implement and optimize differentiable linear time-invariant audio systems. The library is open-source and built on the frequency-sampling filter design method, allowing for the creation of differentiable modules that can be used stand-alone or within the computation graph of neural networks, simplifying the development of differentiable audio systems. It includes predefined filtering modules and auxiliary classes for constructing, training, and logging the optimized systems, all accessible through an intuitive interface. Practical application of these modules is demonstrated through two case studies: the optimization of an artificial reverberator and an active acoustics system for improved response coloration.

AB - We present FLAMO, a Frequency-sampling Library for Audio-Module Optimization designed to implement and optimize differentiable linear time-invariant audio systems. The library is open-source and built on the frequency-sampling filter design method, allowing for the creation of differentiable modules that can be used stand-alone or within the computation graph of neural networks, simplifying the development of differentiable audio systems. It includes predefined filtering modules and auxiliary classes for constructing, training, and logging the optimized systems, all accessible through an intuitive interface. Practical application of these modules is demonstrated through two case studies: the optimization of an artificial reverberator and an active acoustics system for improved response coloration.

KW - audio processing

KW - delay networks

KW - optimization

KW - machine learning

UR - https://arxiv.org/abs/2409.08723

U2 - 10.1109/ICASSP49660.2025.10888532

DO - 10.1109/ICASSP49660.2025.10888532

M3 - Conference contribution

SN - 979-8-3503-6874-1

BT - ICASSP 2025 - 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

PB - IEEE

T2 - IEEE International Conference on Acoustics, Speech and Signal Processing

Y2 - 6 April 2025 through 11 April 2025

ER -

FLAMO: An Open-Source Library for Frequency-Domain Differentiable Audio Processing

Abstract

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Cite this