Max Complexity Setting for Libopus Encoder

This article provides a direct answer regarding the highest complexity setting integer available for the libopus encoder, explaining its impact on CPU utilization and audio quality, as well as how to configure it.

The highest complexity setting integer available to configure in the libopus encoder is 10.

Understanding Libopus Complexity Settings

The complexity parameter in the libopus encoder controls the trade-off between encoder speed (CPU usage) and audio quality. The setting accepts integer values ranging from 0 to 10:

• 0 (Lowest Complexity): Minimizes CPU usage, making it ideal for low-power or legacy embedded devices. This speed optimization comes at the cost of lower audio quality at a given bitrate.
• 10 (Highest Complexity): Enables all of the encoder’s advanced psychoacoustic models and search algorithms. This results in the best possible audio quality and compression efficiency but requires the most CPU processing power.

Default Configuration

By default, the libopus library typically defaults to a complexity of 10 for high-quality encoding, though some applications and mobile frameworks may lower this default to conserve battery life and CPU cycles.

How to Apply the Setting

Depending on the environment or command-line tool you are using, the complexity setting is configured using different parameters:

• Using the Opus API (C/C++): You can set the complexity using the OPUS_SET_COMPLEXITY request in the encoder controller function:

  opus_encoder_ctl(encoder, OPUS_SET_COMPLEXITY(10));

• Using FFmpeg: In FFmpeg, the libopus complexity is mapped to the -compression_level option:

  ffmpeg -i input.wav -c:a libopus -compression_level 10 output.opus