Cadence v0.2.100

Autotune and smoothing for your own voice

Two new controls under Advanced on the Generate screen, for songs sung in a
voice you have trained.

• Autotune (off / subtle / strong) pulls the singing towards the nearest note.
• Smoothing (off / light / strong) protects breath and voiceless consonants and
  cleans up the noise the conversion leaves behind. That is usually what makes a
  cloned vocal sound raspy, so if yours does, start with Light.

Both are greyed out for instrumentals and the generic male and female voices, where
there is no conversion to shape.

Downloads

Platform	File	Status
Windows 10 or 11, 64-bit	.msi	Tested
macOS, Apple silicon	.dmg	Under testing
Linux, Debian and Ubuntu	.deb	Under testing

macOS and Linux are built and smoke-tested on every release, so the app and its
engine start and run. Making actual music on them is not verified yet: the models
want an NVIDIA GPU, so Linux needs CUDA, and Apple silicon support is unconfirmed.
Intel Macs are not covered by this build. Windows is the tested path.

Before your first song

The installer ships the app and its engine, not the multi-gigabyte models. Run the
one-time backend setup once: scripts/setup-backends.ps1 on Windows, or
scripts/setup-backends.sh on macOS and Linux. The
setup guide walks through it.

The installers are unsigned, so Windows SmartScreen will warn on first run: click
More info, then Run anyway.

Full history in the
changelog.