Is your feature request related to a problem? Please describe.
Longer voice messages can be time-consuming to listen to, especially when the message contains a lot of information or when you need to find a specific piece of information.
It would be very helpful to have the option to read the content of a voice message as text instead of having to listen to the entire recording.
This would also make it easier to quickly find or copy information from voice messages.
Nextcloud Talk already provides voice message transcription in the iOS app. However, the current iOS implementation has a known issue where the transcribed text can disappear after the transcription has completed.
See: nextcloud/talk-ios#1269 – Voice message transcript resets during transcription
Describe the solution you'd like
I would like to have a "Transcribe" option for voice messages in the Nextcloud Talk desktop app.
After selecting the option, Talk should transcribe the voice message and display the resulting text directly in the conversation, for example below the voice message.
The transcription should:
- be available for existing voice messages;
- display the transcription directly in the conversation;
- indicate when the transcription is being processed;
- persist after the transcription has completed;
- allow the resulting text to be copied;
- support the languages provided by the selected speech-to-text engine.
Ideally, the transcription should be independent of the operating system and its local speech recognition capabilities.
Describe alternatives you've considered
Currently, the only option is to listen to the complete voice message.
Another option would be to download the voice message and use an external transcription service or application. This is inconvenient and may have privacy implications, as the audio would leave the Nextcloud environment.
An integrated transcription feature in Talk would provide a much more convenient workflow.
Additional context
The Nextcloud Talk iOS app already provides transcription for voice messages.
There is also an existing issue in the iOS app concerning the persistence of the transcription result:
nextcloud/talk-desktop#1269 – Voice message transcript resets during transcription
It would be useful to have the same functionality available in the Talk desktop application.
A platform-independent or server-side transcription solution could make the feature available consistently across Windows, macOS and Linux without depending on the speech recognition capabilities of the individual operating system.
Is your feature request related to a problem? Please describe.
Longer voice messages can be time-consuming to listen to, especially when the message contains a lot of information or when you need to find a specific piece of information.
It would be very helpful to have the option to read the content of a voice message as text instead of having to listen to the entire recording.
This would also make it easier to quickly find or copy information from voice messages.
Nextcloud Talk already provides voice message transcription in the iOS app. However, the current iOS implementation has a known issue where the transcribed text can disappear after the transcription has completed.
See: nextcloud/talk-ios#1269 – Voice message transcript resets during transcription
Describe the solution you'd like
I would like to have a "Transcribe" option for voice messages in the Nextcloud Talk desktop app.
After selecting the option, Talk should transcribe the voice message and display the resulting text directly in the conversation, for example below the voice message.
The transcription should:
Ideally, the transcription should be independent of the operating system and its local speech recognition capabilities.
Describe alternatives you've considered
Currently, the only option is to listen to the complete voice message.
Another option would be to download the voice message and use an external transcription service or application. This is inconvenient and may have privacy implications, as the audio would leave the Nextcloud environment.
An integrated transcription feature in Talk would provide a much more convenient workflow.
Additional context
The Nextcloud Talk iOS app already provides transcription for voice messages.
There is also an existing issue in the iOS app concerning the persistence of the transcription result:
nextcloud/talk-desktop#1269 – Voice message transcript resets during transcription
It would be useful to have the same functionality available in the Talk desktop application.
A platform-independent or server-side transcription solution could make the feature available consistently across Windows, macOS and Linux without depending on the speech recognition capabilities of the individual operating system.