Is this page helpful?

About the Audio Super-Resolution Effect#

The Audio Super-Resolution effect upsamples the audio. For low-frequency audio, this feature predicts the higher frequency spectrum of input audio, which improves audio quality.

Note

In this guide, the term Super-Resolution is used interchangeably with Superres and Superresolution (referred to as superres in the API).

This effect has the following characteristics:

  • Supported input/output audio format is 32-bit float audio.

Note

The main purpose of this effect is to enhance the sampling rate of input audio. The level of enhancement seen in the output audio depends on the type of audio.

Audio that is captured on Windows with the audio enhancement settings disabled produces better superres outputs than when this setting is enabled.

Acquire the Windows feature files once from the SDK features directory:

powershell -ExecutionPolicy Bypass -File .\download_features.ps1 -Effects superres-8k_to_16k,superres-16k_to_48k -Version 3.0.0 -Output_Directory .

Run from an existing package without network access. The x64 helper directory is shown; on NVIDIA N1X, use samples\build-arm64\Release\scripts:

Set-Location $env:AFX_SDK_ROOT\samples\build\Release\scripts
.\run_effects_demo.bat -e superres -isr 8k -osr 16k
.\run_effects_demo.bat -e superres -isr 16k -osr 48k

Note

For more information, see Use the Helper Script to Run the Sample Application.

Acquire the Linux models once from the SDK models directory:

./download_models.sh --gpu <gpu> --effects superres-8k_to_16k,superres-16k_to_48k

Run locally from the built Linux effects_demo sample directory:

./run_effect.sh -g t4 -s 16 -e superres
./run_effect.sh -g t4 -s 48 -e superres

Note

For more information, see Use the Helper Script to Run the Sample Application.

  • Supported upsampling of 8-kHz input audio to 16-kHz output (2x) and 16-kHz input audio to 48-kHz output (3x).

  • In the Linux SDK, this effect has the following maximum throughput (the number of batches supported in real time):

    Architecture

    Maximum Throughput for the 8K-to-16K Effect

    Maximum Throughput for the 16K-to-48K Effect

    T4

    410

    180

    A100

    2110

    920

    A10

    910

    380

    L40

    3450

    1350

    H100

    2430

    1030

    B100

    5720

    2330

    RTX PRO 6000

    3940

    1650