Fully Conditioned and Low-latency Black-box Modeling of Analog Compression

This page includes of audio examples for the paper Simionato R., Fasciani S., Fully Conditioned and Low-latency Black-box Modeling of Analog Compression, in Proceedings of the 26the Digital Audio Effects Conference (DAFx23), Copenhagen, Denmark, 2023.

 

 

 

Associated code, dataset, and trained models are available on GitHub.

 

Examples are available for the following compressor models:

 

Each example includes the following:

  • Input Audio Signal
  • Target Output Audio Signal (true output recorded from the device)
  • Predicted Output Audio Signal (generated by the inference of the trained model)
  • Waveform of Target and Predicted Output Audio Signal with respective RMS values
  • Spectrograms of Target and Predicted Output Audio Signal

The title of the waveform plot details the compressor control parameters, valid for all devices except the LA-2A:

  • at = attack time (ms)
  • re = release time (s)
  • ra = compression ratio (ra:1)
  • th = threshold (dBu)

For the LA2A, the control parameters detailed in the title of the waveform plot are the following:

  • M = mode (1 compression, 0 limit)
  • p = peak reduction (dimensionless)

 

CL 1B Dataset – Encoder Decoder Light Model

Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player



LA-2A Dataset – Encoder Decoder Light Model

Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player



CL 1B Dataset – Encoder Decoder Model

Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player



LA-2A Dataset – Encoder Decoder Model

Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player



PSP MicroComp Dataset – Encoder Decoder Model

Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


U-he Presswerk Dataset – Encoder Decoder Model

Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player



Softube FET Dataset – Encoder Decoder Model

Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player



CL 1B Dataset – Temporal Convolutional Network Model

Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player


Input Audio Player

Target Audio Player

Prediction Audio Player



Bookmark the permalink.

Comments are closed.