tune

• Improvement: microwave is now based on OxiSynth
  • Compilation to more targets is possible
• Improvement: Add phase parameter to oscillators
• Bug fix: Mixed-up GUI status updates

• Improvement: Add BufferWriter::buffer_len
• Improvement: Add Spec and AutomatedValue impls for Option<T>

• Improvement: fluid-xenth is now based on OxiSynth
  • All code is written in Rust
  • Compilation to WASM is possible
• Change: Several API changes due to the adoption of OxiSynth, mainly:
  • Calling into functions requires &mut Xenth
  • Xenth::write consumes a callback function

• New: Add bench command for performance testing
• Change: LF source Wavelength is now WaveformPeriod
• Change: LF source Property.Velocity is now Velocity
• Change: LF source Property.Pressure is now Controller[kind=KeyPressure]
• Change: LF source Control[controller=<cc>] is now Controller[kind=<cc>]
• Bug fix: Full keyboard AOT live retuning not working due to non-retuned notes being counted and exceeding the limit of 128 notes
• Bug fix: Non-monotonic scales were sorted when they should not

• Bug fix: Full keyboard AOT live retuning not working due to non-retuned notes being counted and exceeding the limit of 128 notes

• Change: AotXenthControl::set_tuning returns Result<usize, SendCommandResult> instead of Result<usize, SetTuningError<SendCommandResult>>

• Deletion: FullKeyboardDetuning and OctaveBasedDetuning
  • Use ChannelDetuning<G> instead
• Deletion: SetTuningError
  • Use AotTuner::tuned to test whether setting a tuning was succesful
• Improvement: Reject Single Note Tuning Change messages only when number of actually retuned notes is larger than 128

• New: core components of microwave have been extracted to a separate crate.

• New: Fixed-scale mode with FluidLite now uses fluid-xenth's AOT tuning model for playback free of artifacts
• Improvement: --kb2 option now allows for specifying a colored second keyboard

• Change: XenthControl has been renamed to JitXenthControl
• New: Add AotXenthControl for artifact-free playback of fixed-scale tunings
• New: Add TunableFluid implementing TunableSynth for better composability

• Change: JitTuner has been renamed to JitTuningModel
• Change: AotTuner has been renamed to AotTuningModel
• New: Add TunableSynth abstraction implemented by TunableMidi
• New: Add JitTuner consuming an instance of TunableSynth for better composability
• New: Add AotTuner consuming an instance of TunableSynth for better composability

• New: Support the Lumatone and other MIDI controllers with multichannel note output
• New: Add ANSI keyboard support
• Improvement: Support MIDI controllers that send NoteOn with velocity 0 instead of NoteOff
• Improvement: Multi-cyclic keyboard layouts are compressed along the sharp axis
• Internal: Use clap instead of structopt

• New: Support the Lumatone and other MIDI controllers with multichannel note output
• Improvement: Support MIDI controllers that send NoteOn with velocity 0 instead of NoteOff
• Improvement: Multi-cyclic keyboard layouts are compressed along the sharp axis
• Improvement: Print step sizes and fret constants in est command
• Internal: Use clap instead of structopt

• Improvement: tuner::MidiTarget now accepts an arbitrary channel list
• Improvement: keyboard::Keyboard::coprime compresses along the sharp axis

• Improvement: Add experimental GL support
• Improvement (Bug fix?): Update wgpu from 0.11.0 to 0.11.1 which should fix some driver compatibility regressions

• Improvement: The sample rate is now configurable
• Bug fix: Prevent StreamConfigNotSupported aborts by taking the audio device's preferred sample rate into account

This is a huge, feature-rich release!

• New: Support JIT-based MIDI-out
  • This means any pitch can be played without having to pre-tune the MIDI device
  • It is possible to update/bend a currently sounding note polyphonically
• Improvement: All tuning message types and control parameters from tune-cli are available
  • realtime / device ID / tuning program / ...
• Improvement: Both AOT and JIT live retuning can be used for MIDI-out
  • When tuning mode is "Fixed": Use AOT model
  • When tuning mode is "Continuous": Use JIT model
• Improvement: Display the tuning message type
• Improvement: Upgrade to Nannou 0.18.0
  • This should result in less non-Rust build dependencies
• Change: The CLI of microwave and tune-cli have been harmonized
  • The available tuning methods are full, full-rt, octave-1, octave-1-rt, octave-2, octave-2-rt, fine-tuning and pitch-bend
• Change: The --tun-method parameter is now mandatory if --midi-out is provided
• Change: To toggle between the tuning modes "Fixed" and "Continuous" use Ctrl+T (instead of Ctrl+C)

• New: Add support for Scale/Octave Tuning (2-Byte) messages
  • Command-line argument: octave-2 or octave-2-rt
• Improvement: Add control over Real-Time SysEx flag
  • Example: full vs. full-rt and octave-1 vs octave-1-rt
• Change: The live subcommand has been updated
  • octave is now octave-1, channel is now fine-tuning
  • The order of the command-line arguments is slightly different

• New: Add JitMidiTuner and AotMidiTuner
  • Those components have been extracted from tune-cli in order to be reusable
• Change: Rename ChannelTuner to AotTuner
• Change: Remove the G parameter from JitTuner
• Deletion: ChannelMessage::transform and ChannelMessage::distribute have been removed. They were too MIDI-centric and became useless in favor of JitMidiTuner and AotMidiTuner.