.\" Automatically generated by Pandoc 2.14.0.3
.\"
.TH "SONIVOXRENDER" "1" "February 14, 2026" "sonivox 4.0.1.0" "Sonivox MIDI File Renderer"
.hy
.SH NAME
.PP
\f[B]sonivoxrender\f[R] \[em] Render MIDI song files into raw PCM audio
.SH SYNOPSIS
.PP
\f[B]sonivoxrender\f[R] [\f[B]-h|--help\f[R]] [\f[B]-v|--version\f[R]]
[\f[B]-d|--dls\f[R] \f[I]soundfont\f[R]] [\f[B]-r|--reverb\f[R]
\f[I]0..4\f[R]] [\f[B]-w|--wet\f[R] \f[I]0..32767\f[R]]
[\f[B]-n|--dry\f[R] \f[I]0..32767\f[R]] [\f[B]-c|--chorus\f[R]
\f[I]0..4\f[R]] [\f[B]-l|--level\f[R] \f[I]0..32767\f[R]]
[\f[B]-g|--gain\f[R] \f[I]0..196\f[R]] [\f[B]-V|--Verbosity\f[R]
\f[I]0..5\f[R]] [\f[B]-R|--reverb-post-mix\f[R]]
[\f[B]-C|--chorus-post-mix\f[R]] [\f[B]-s|--sndlib\f[R] \f[I]1..3\f[R]]
\f[I]song_file\f[R]
.SH DESCRIPTION
.PP
This program is a MIDI file renderer based on the sonivox synthesizer
library.
It reads .MID (Standard MIDI Files), .RMI (RIFF-RMID), and .XMF
(Extensible Music Files) file formats, and writes an audio stream to the
standard output as raw 16 bit stereo PCM samples.
.SS Options
.TP
-h, --help
Prints brief usage information.
.TP
-v, --version
Prints the version numbers.
.TP
-d, --dls \f[I]soundfont\f[R]
Optional DLS or SF2 soundfont file name.
If not provided, it uses an internal embedded soundfont.
.TP
-r, --reverb \f[I]reverb_preset\f[R]
Reverb preset between 0 and 4: 0=no, 1=large hall, 2=hall, 3=chamber,
4=room.
.TP
-w, --wet \f[I]reverb_wet\f[R]
Reverb wet level between 0 and 32767.
.TP
-n, --dry \f[I]reverb_dry\f[R]
Reverb dry level between 0 and 32767.
.TP
-c, --chorus \f[I]chorus_preset\f[R]
Chorus preset between 0 and 4: 0=no, 1..4=presets.
.TP
-l, --level \f[I]chorus_level\f[R]
Chorus level between 0 and 32767.
.TP
-g, --gain \f[I]master_gain\f[R]
Master gain between 0 and 196, default is 100 (+0dB).
The number is relative to 100, in 1dB increments, e.g.\ 120 = +20dB, 80
= -20dB.
.TP
-V, --Verbosity \f[I]verbosity\f[R]
Verbosity level between 0 and 5, where 0=no, 1..5=severity levels.
.TP
-R, --reverb-post-mix
Ignore CC91 reverb send level.
The reverb effect will apply to mixed output audio, which is the old
behavior.
.TP
-C, --chorus-post-mix
Ignore CC93 chorus send level.
See also \f[B]--reverb-post-mix\f[R].
.TP
-s, --sndlib \f[I]index\f[R]
EAS sound library to use:
.RS
.IP \[bu] 2
1: wt_200k_G (default) - WT-only bank, used in Android devices.
Also named \[lq]Common\[rq].
Support 22050 Hz and 44100 Hz sample rates, 8-bit and 16-bit samples.
.IP \[bu] 2
2: GMdblib-3 - FM-only bank.
Support all sample rates.
Sample bit depth does not matter here.
.IP \[bu] 2
3: hybrid_22khz_mcu - Hybrid bank.
Use WT synth for drums and FM for melodic instruments.
This bank is a combination of \f[C]GMdblib-3\f[R] and
\f[C]Sonic_20Khz_Drums\f[R].
Support 22050 Hz sample rate and 8-bit samples only.
.PP
\f[B]Note:\f[R] This option does not affect DLS/SF2.
They will always use the DLS synth engine.
If the selected sound library is not compatible with the build
configuration, the program will fail with an error message.
.RE
.SS Arguments
.TP
\f[I]song_file\f[R]
Input [\f[B]MID|RMI|XMF\f[R]] file name.
.SH EXAMPLES
.PP
The following examples assume the default option USE_44KHZ=ON, which
means an output sample rate = 44100 Hz.
.PP
Example 1: Render a MIDI file and save the rendered audio as a raw audio
file:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid > ants.pcm
\f[R]
.fi
.PP
Example 2: pipe the rendered audio thru the Linux ALSA \f[B]aplay\f[R]
utility:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid | aplay -c 2 -f S16_LE -r 44100
\f[R]
.fi
.PP
is equivalent to:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid | aplay -f cd
\f[R]
.fi
.PP
Example 3: pipe the rendered audio thru the \f[B]lame\f[R] utility
creating a MP3 file:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid | lame -r -s 44100 - ants.mp3
\f[R]
.fi
.PP
Example 4: pipe the rendered audio thru the \f[B]sox\f[R] utility
creating a WAV file:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid | sox -t s16 -c 2 -r 44100 - ants.wav
\f[R]
.fi
.PP
Example 5: pipe the rendered audio thru the PulseAudio\[cq]s
\f[B]pacat\f[R] utility:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid | pacat
\f[R]
.fi
.PP
Example 6: pipe the rendered audio thru the PipeWire\[cq]s
\f[B]pw-play\f[R] utility:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid | pw-play --rate 44100 -
\f[R]
.fi
.PP
Example 7: pipe the rendered audio thru the FFmpeg\[cq]s
\f[B]ffplay\f[R] utility:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid | ffplay -i - -f s16le -ar 44.1k -ac 2 -nodisp -autoexit -loglevel quiet
\f[R]
.fi
.PP
Example 8: pipe the rendered audio thru the \f[B]mpv\f[R] media player:
.IP
.nf
\f[C]
$ sonivoxrender ants.mid | mpv --demuxer=rawaudio -demuxer-rawaudio-format=s16le --demuxer-rawaudio-rate=44100 --demuxer-rawaudio-channels=2 --no-video -
\f[R]
.fi
.SH BUGS
.PP
See Tickets at GitHub <https://github.com/EmbeddedSynth/sonivox/issues/>
.SH LICENSE AND COPYRIGHT
.PP
Licensed under the Apache License, Version 2.0
.PP
Copyright (c) 2022-2026 Pedro L\['o]pez-Cabanillas and contributors
.SH AUTHORS
Pedro L\['o]pez-Cabanillas <plcl@users.sf.net>.