This plugin is a well structured, stateless audio recorder that uses videoJS recording capabilities. It will allow you to create audio recordings and play them back without having to upload to your Bubble S3 area first.
Each recording is held within the "recording" state, and each recording within that state has a bunch of fields containing details of that particular recording.

You will be able to create your own list of recordings that are all held in browser memory and then select any one of those to upload to either Bubble S3 or Wasabi once happy with it.

All of the standard options that are associated to audio recordings within the browser, are supported. This include options like buffer sizes, sample rates, bitrates, channels, noise suppression, auto gain control, max recording length  etc.

You can display the standard videoJS recording UI if required, or you can hide this and control the recorder using the actions instead. Wavesurfer is built-in so you get nicely animated sound waves appear as you speak or playback recordings.

If you're interested in voice cloning at all, there is a client-side action which will allow you to upload a recording directly to Cartesia's voice cloning endpoint.

Live transcriptions are also supported using either the built-in speech recognition API within the browser, or you can choose to use OpenAI's real-time WebRTC transcription service.

Other helpful events/states are also included, and you can control when to prompt for microphone permissions and change audio input devices at any time.