Faster Dictaphone Processing During Long Retranscriptions

🔄 Improvements

• Improved dictaphone retranscription flow so processed sections can continue moving through the pipeline while later parts are still being retranscribed
• Improved handling of audio files sent for transcription by automatically detecting the correct audio format
• Improved reliability when transcription services return incomplete or empty results by automatically retrying affected audio segments

✨ New Features

• Added automatic recovery for empty transcription results, helping long recordings continue processing without manual intervention
• Added support for more accurate audio format detection for uploaded WAV and MP3 recordings

🛠 Fixes

• Fixed: segmented dictaphone recordings could remain waiting until the very end of a long retranscription before section processing started Screenshot

And other improvements