016a171a35
Another dependency the app should not be asking the platform for:
- audio_decode — MP3 (layer 3, LSF tables, synthesis) and Ogg Vorbis
(codebooks, floor, residue, MDCT) decoders, with tag reading. Both are
checked against oracle fixtures rather than against our own expectations.
- audio_encode — an Ogg Vorbis encoder: MDCT, psychoacoustics, floor and
Huffman coding, setup tables, plus `oggenc` and `audiobench` binaries.
- audio_picture — waveform and spectrogram rendering, and compositing.
- audio_lyrics — word-level lyric alignment (DTW plus a DP snap) and the
baked schema behind karaoke timing.
- audio_sidechannels — the side-channel plumbing between them.
libs/voice grows a CUDA backend and an alignment path beside its CPU decoder,
with a `whisper_parity` binary to keep the two honest.
19 lines
779 B
TOML
19 lines
779 B
TOML
# Pictures of audio: the log-frequency spectrogram and the antialiased
|
|
# waveform strip, and the composite that carries both.
|
|
#
|
|
# ONE definition of the recipe. The importer bakes these into catalog
|
|
# thumbnails; the shared preview widgets render the same pictures live from
|
|
# decoded samples. Before this crate the ramp, the window and the band lived
|
|
# in the importer, so the only way a widget could show a spectrogram was to
|
|
# fetch one somebody had already baked.
|
|
#
|
|
# Zero dependencies, by design: an iterative radix-2 FFT and a coverage
|
|
# rasteriser, nothing else. That is what lets a widget crate and a headless
|
|
# importer share it without either dragging the other's world along.
|
|
|
|
[package]
|
|
name = "makepad-audio-picture"
|
|
version = "0.1.0"
|
|
edition = "2021"
|
|
|
|
[dependencies]
|