> ## Documentation Index
> Fetch the complete documentation index at: https://rockboxzig.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Audio settings

> Volume, EQ, DSP, ReplayGain, crossfade and the rest of the rbcodec pipeline.

The Rockbox DSP pipeline runs between the codec output and the active PCM
sink. Settings live in `global_settings` (in the C firmware) and are
mirrored in `settings.toml`. Most can also be changed at runtime via
GraphQL or gRPC and they persist on the next save cycle.

<CardGroup cols={2}>
  <Card title="Equalizer" icon="chart-column" href="/audio-settings/equalizer">
    10-band parametric EQ — gain, centre frequency, Q per band.
  </Card>

  <Card title="DSP" icon="waveform-lines" href="/audio-settings/dsp">
    Crossfeed, surround, PBE, AFR, compressor, dithering.
  </Card>

  <Card title="ReplayGain" icon="volume" href="/audio-settings/replaygain">
    Track / album normalisation with optional clipping protection.
  </Card>

  <Card title="Crossfade" icon="shuffle" href="/audio-settings/crossfade">
    Overlap track ends with the next track's start.
  </Card>
</CardGroup>

## Where settings live

| Layer                | Lives in                              | Lifetime             |
| -------------------- | ------------------------------------- | -------------------- |
| Compiled-in defaults | `apps/settings_list.c`                | Build-time           |
| `settings.toml`      | `~/.config/rockbox.org/settings.toml` | Read once at startup |
| Runtime (API)        | In-memory `global_settings`           | Persisted on save    |

Hardware settings flow through `firmware/sound.c → sound_set_*()`. DSP
settings flow through `lib/rbcodec/dsp/*` and are applied in the PCM pipeline
before samples reach the sink.

## Volume

`global_status.volume` — decibels relative to the device's clipping point.
**0 dB** is the maximum undistorted level. Negative values reduce output;
positive values may distort. On the CPAL target, volume is implemented as a
software mixer.

```toml theme={"theme":{"light":"catppuccin-latte","dark":"min-dark"}}
volume_limit = 0   # ceiling in dB (default = device max)
```

## Channels & stereo

| Setting                    | Storage          | Range / values                                       |
| -------------------------- | ---------------- | ---------------------------------------------------- |
| Balance                    | `balance`        | −100..+100                                           |
| Channel config             | `channel_config` | Stereo / Mono / Custom / Mono L / R / Karaoke / Swap |
| Stereo width (when Custom) | `stereo_width`   | 0..255 %                                             |

## Pitch & time-stretch

Persisted in `global_status` so they survive across restarts. Time-stretch
uses a TDHS algorithm — best for speech, may sound rough on dense music.

| Setting             | Storage                               | Range       |
| ------------------- | ------------------------------------- | ----------- |
| Pitch               | `global_status.resume_pitch`          | \~50..200 % |
| Speed               | `global_status.resume_speed`          | \~35..250 % |
| Time-stretch enable | `global_settings.timestretch_enabled` | bool        |

## Output sample rate

```toml theme={"theme":{"light":"catppuccin-latte","dark":"min-dark"}}
play_frequency = 0   # 0=auto, 44100, 48000, 88200, 96000
```

## UI feedback

```toml theme={"theme":{"light":"catppuccin-latte","dark":"min-dark"}}
beep              = 0   # off / weak / moderate / strong
keyclick          = 0
keyclick_repeats  = false
```