Crates.io | augmented-dsp-filters |
lib.rs | augmented-dsp-filters |
version | 2.5.0 |
source | src |
created_at | 2021-10-30 05:48:58.348691 |
updated_at | 2024-08-26 12:05:16.978953 |
description | Audio filters (port of C++ DSPFilters) |
homepage | https://github.com/yamadapc/augmented-audio |
repository | https://github.com/yamadapc/augmented-audio |
max_upload_size | |
id | 474279 |
size | 1,132,142 |
Mechanical port of Vinnie Falco's https://github.com/vinniefalco/DSPFilters/.
Only RBJ filters are ported over. No introspection is supported & the implementation is quite a different (as Rust would prefer composition to multiple inheritance).
Very untested, be careful with your speakers.
Depends on audio-processor-traits
. Exports FilterProcessor
which may be used for general filtering needs.
See synth
on this repository for a nice working example.
use audio_processor_traits::{
AudioBuffer, AudioProcessor, AudioProcessorSettings
};
use augmented_dsp_filters::rbj::{FilterProcessor, FilterType};
pub struct YourProcessor {
filter: FilterProcessor<f32>, // <- f32 may be f64 if you wish
}
impl YourProcessor {
fn new() -> Self {
Self {
filter: FilterProcessor::new(FilterType::LowPass),
}
}
fn set_cutoff(&mut self, midi_value_between_0_and_127: f32) {
let cutoff_ratio = midi_value_between_0_and_127 / 127.0;
let cutoff_freq_hz = 22000.0 * cutoff_ratio;
self.filter.set_cutoff(cutoff_freq_hz);
}
fn set_q(&mut self, q: f32) {
self.filter.set_q(q);
}
}
impl AudioProcessor for YourProcessor {
type SampleType = f32;
fn prepare(&mut self, settings: AudioProcessorSettings) {
self.filter.prepare(settings);
}
fn process<BufferType: AudioBuffer<SampleType = Self::SampleType>>(
&mut self,
data: &mut BufferType,
) {
self.filter.process(data);
}
}
The filter mutate functions recalculate coefficients for the filter. This should run on the audio-thread only.
In order to integrate with MIDI, see synth
; this won't be a problem as AudioProcessor
(stand-alone) will receive
MIDI on the audio-thread.
For integrating with a GUI thread, the best would probably be to have the audio-thread read the parameters from an
atomic store & update the filter when they change (see audio-parameter-store
in this repository).
MIT licensed as the original.