Georgi Gerganov

The current format of the audio payload when using variable length messages is like this: - assuming 48 kHz sample rate is used - 1 frame contains 1024 samples -...

The fixed-length mode is a bit different, although the logic is simpler. In this mode: - there are no sound marker frames at all. - `N = n + E(n)`...

Thanks for the info. I am still not sure how to reduce these 60-ish ms variance. The only idea that I have is to compute the audio power within the...

Currently we generate single-channel audio. Not sure what would be the benefit to make a two-channel sound. I guess we can split half the frequencies in one channel and the...

Today I added a C interface to the library and implemented Python bindings. I think this C interface can be used for iOS development. A simple example to demonstrate how...

In Waver I used https://github.com/pthom/hello_imgui for the boilerplate and only had to plugin my C++ code. It uses SDL for audio capture/playback and the C++ SDL functions are automatically doing...

Sounds good. Ideally I would like to have a simple test for each platform/language which simply encodes and decodes some data without using any audio engine. I've currently made such...

Absolutely agree - I need to work on that. I added some more comments to `ggwave.h`: https://github.com/ggerganov/ggwave/blob/master/include/ggwave/ggwave.h#L21-L195 I hope these can answer your questions. While writing this, I realized I...

Yes - we generate and process only raw PCM. Whenever I used the word `waveform`, I actually mean PCM :)

No worries, I really appreciate the effort! Thanks!