jake/vOOice
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
997 Commits 1 Branch 0 Tags
Commit Graph

78 Commits

Author SHA1 Message Date
Jake Shore
f6a613e8ef Use 320ms streaming chunks and fix bundle ID 
- Switch from 160ms to 320ms chunk size for better streaming accuracy
- Fix corrupted bundle identifier (was "--com.jakeshore.VoiceInk-com.jakeshore.VoiceInk")

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
 2026-02-13 05:24:30 -05:00
Jake Shore
de1c1e51aa Add hybrid streaming transcription for improved accuracy 
- Implement real-time streaming preview using Parakeet EOU (160ms chunks)
- Add batch transcription on completion for accurate final result
- Prefer Whisper large-v3-turbo (2.7% WER) over Parakeet (6.05% WER) when available
- Remove audio preprocessing that hurts ASR accuracy (gain control, noise reduction)
- Add streaming audio callback support in Recorder and CoreAudioRecorder
- Raw audio passthrough - SDK handles resampling internally

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
 2026-01-16 07:35:53 -05:00
Beingpax
652859414c v1.69 Update 2026-01-12 19:50:37 +05:45
Beingpax
c530367a04 Replace audio recorder with CoreAudio AUHAL 
New low-level recorder targeting devices directly. Includes device switching during recording, enhanced logging (transport type, format, buffer), and log export feature.
 2026-01-10 20:45:37 +05:45
Beingpax
072b57f365 Update to 1.67 2026-01-07 11:30:17 +05:45
Beingpax
5d6a2208cf Update selectedtextkit package 2025-12-21 22:48:18 +05:45
Beingpax
557ac95478 1.64 update 2025-12-06 08:43:41 +05:45
Beingpax
ab9d6ad830 Add option to dismiss Affiliate promotion 2025-12-06 08:18:35 +05:45
Beingpax
c2666c3c24 Update to 1.63 2025-12-04 22:30:22 +05:45
Beingpax
88676362b4 update to v1.62 2025-11-29 08:49:08 +05:45
Beingpax
e5e194de5f v1.61 update 2025-11-08 12:00:45 +05:45
Beingpax
7c86ac6928 Added Swift atomics dependency. 2025-10-31 11:31:25 +05:45
Beingpax
ea0e5d840c Rewrote the default prompts to make them simpler and better. 2025-10-29 23:27:49 +05:45
Beingpax
615a0c4fad Isolate whisper.cpp dependency build outside the project. 2025-10-29 11:59:49 +05:45
Beingpax
357804c03d feat: Integrate SelectedTextKit for robust text retrieval with multi-strategy support 2025-10-28 18:30:25 +05:45
Beingpax
29d139285a Update it to version 1.59 2025-10-10 22:37:35 +05:45
Beingpax
b7627c9deb Remove screen capture toggle gating 2025-10-06 17:26:38 +05:45
Beingpax
7dcf982b60 Updated to version 1.57 2025-09-24 21:08:22 +05:45
Beingpax
abaa54eb02 Fix build no. 2025-09-16 10:43:50 +05:45
Beingpax
e6225050cc Prefer enhanced text with graceful fallback: paste enhanced falls back to original; copy last transcription now mirrors same behavior; updated Settings description to reflect single-shortcut-friendly design. 2025-09-16 10:30:58 +05:45
Beingpax
12c850f77b Open With and title cleanup 2025-09-05 11:22:14 +05:45
Beingpax
68e8a21746 Clipboard context with their own tags 2025-08-26 16:55:01 +05:45
Beingpax
f0ddbed5e8 Update to version 1.5 2025-08-25 14:09:17 +05:45
Beingpax
6a308b81bf Update app to support Parakeet B3 model 2025-08-25 13:00:35 +05:45
Beingpax
889162633c remove api header 2025-08-18 19:48:09 +05:45
Beingpax
a8db010900 Added support for adding fine-tuned local whisper models. 2025-08-13 14:04:34 +05:45
Beingpax
6935f58737 Recorder error loggin. 2025-08-12 21:05:53 +05:45
Beingpax
1b4f09731b 1.50 2025-08-12 10:30:44 +05:45
Beingpax
0b8eb71048 Enable native Apple Speech Analyzer in build 2025-08-11 09:07:29 +05:45
Danny Ricciotti
2c12003b84 Update project.pbxproj 2025-08-10 18:59:53 -04:00
Beingpax
b14bf8733e Added tip jar 2025-08-08 21:34:12 +05:45
Beingpax
4a14b7cd52 1.47 update 2025-08-08 09:02:22 +05:45
Beingpax
d363808a7b revert copy to clipboard 2025-08-07 10:09:02 +05:45
Beingpax
f7987cd4bd Redesign recorder views with reusable components 2025-08-06 13:10:21 +05:45
Beingpax
7c74f57b8f update version 2025-08-03 13:03:16 +05:45
Beingpax
3eebbc4e3b Better Parakeet error handling 2025-08-03 12:44:13 +05:45
Beingpax
29722d0a31 more logging in parakeettranscription service 2025-08-03 09:35:49 +05:45
Beingpax
cfd09ed36b Add correct model name for Qwen-3-235b-a22b-thinking-250 2025-08-02 12:24:28 +05:45
Beingpax
d09a9fba7f Experimental new models 2025-08-01 17:26:08 +05:45
Beingpax
50c7b9a354 Support for playing/pausing media during recording 2025-07-27 10:23:36 +05:45
Palash Oswal
e8d49b1797
Update identifiers to reflect latest release 2025-07-25 11:56:53 -04:00
Beingpax
f9adfd5269 cerebras AI provider integration 2025-07-23 21:08:14 +05:45
Beingpax
9968ac3dcc Add sysmte info in analysis view 2025-07-11 19:11:34 +05:45
Beingpax
4a83c5400e Fix for paste fail window messsing with Paste operation 2025-07-04 05:38:47 +05:45
Beingpax
8c44efad64 Consistent background card 2025-06-30 11:27:45 +05:45
Beingpax
a0a1257e3b Add enhancement prompt support when power mode disabled 2025-06-22 23:37:03 +05:45
Beingpax
f7e160ac99 Fix cloud model default clearing 2025-06-15 23:46:53 +05:45
Beingpax
7aef8f04a6 Fix audio input mode persistence 2025-06-15 18:49:41 +05:45
Beingpax
b1524970c1 fix: power mode model selectio & removed current model in favor of currenttranscriptionmodel 2025-06-10 23:23:43 +05:45
Beingpax
e2e6c3ac94 feat: Migrate Power Mode to support cloud models 2025-06-10 17:13:55 +05:45