Repository navigation
Replies: 1 comment
|
@Drew3296 VU meters: yes, and mostly already here. The level-meter primitive is in ui/primitives, playback already has the sample stream, and what is missing is computing RMS/peak per track and publishing it to the timeline header and the monitor. A day or two of work, and no new dependency. @jub0t Just a note. VST: a much bigger proposition, and probably not the right shape for this project. VST3 is a C++ SDK with its own licence, needs per-platform window embedding, plugin sandboxing (a plugin that crashes should not take Concat down), and a real-time-safe audio thread. That is I think months of work for several people, not weeks. Also licencing can open a discussion. If there are future plans for a similar integration, maybe CLAP is the modern open-standard alternative with an MIT licence and a much simpler hosting story; it is what Bitwig, u-he and Reaper build against now. LV2 is the Linux-native option, though it does not travel to macOS and Windows as cleanly. If the goal is just "more audio effects without a team writing them", the cheaper/simpler path is what Concat already does: WGSL shaders for picture and a small catalogue of built-in audio filters (which could also be expanded). Neither VST nor CLAP can change that overnight. |
Uh oh!
There was an error while loading. Please reload this page.
All reactions