[Scrum] - SUSI Interpretation Project - 20260713

[Eshaan Walia]

1. What did I do yesterday?

• (eventyay-interpretation) - Opened Issue(#38) - Remove Polling From Interpretation (WS Primary, SSE Fallback)
  open
• (susi_translator) - Updated PR (#117) - feat(tts): support selectable voice styles
  open
• (susi_translator) - Opened Issue(#116) - Add voice selection support to TTS translation stream
  open
• (eventyay-interpretation) - Updated Issue(#29) - Switch caption delivery to Websocket (primary) with SSE signaling/fallback, remove polling
  closed
• (eventyay-interpretation) - Updated Issue(#28) - fix: Server-side SSRF via `base_url`
  closed
• (eventyay-interpretation) - Updated Issue(#26) - Task 3. Subtitles for non-HLS video inputs
  closed
• (eventyay-interpretation) - Updated Issue(#23) - Phase2: Feat: TTS, translated subtitles, non-HLS support
  open

• (eventyay-interpretation) - Updated PR  - Text-to-Speech (TTS) audio output 
  open

• Unified the caption relay into a single SSE stream for both captions and TTS. Added support for forwarding late-arriving audio events through an event queue, included audio=true support, and made language selection mandatory when tts=1 is enabled.

• Added support for selectable TTS voices in the relay. The selected voice is forwarded upstream as a query parameter, and the list of available voices is advertised to clients in the initial SSE connection event.

• (eventyay) - Updated commits - Feat/ Interpretation Plugin

• Added Text-to-Speech (TTS) support for spoken translations. When TTS is enabled, the original stage audio is muted, translated speech from SUSI is played, audio queue management was implemented, tts_url support was added, and interpretationTtsActive state was introduced.

• Fixed stage audio restoration when TTS is toggled. The original mute/volume state is now preserved instead of forcing the stage volume to zero.
• Fixed TTS resume behavior. Speech now continues from the current live captions when re-enabled, add stale caption/audio streams are discarded to prevent outdated playback. 
• Added an independent TTS volume control, allowing users to adjust translated speech volume separately from the original stage audio.
• Added selectable TTS voices. Available voices are fetched from the server, shown in a dropdown, and changing the voice automatically restarts TTS with the selected voice.

TTS feature fully implemented in eventyay videos, as part of the interpretation plugin. 
here is an image for it: 


2. What do I plan to do today?

• Update interpretation to work with websockets as primary connection and sse as fallback

• Prepare pr 19 for review

3. What is blocking me from making progress?

No Blocker at the moment