Player not loading? Watch on YouTube
The video is in German; this English overview is based on its full German captions.
The video introduces local audio transcription for interviews, meeting notes and recorded talks. The speaker raises two concerns about uploading recordings: sharing other people's voice profiles and sending personal information to external services. He presents local processing as a way to avoid those transfers; his discussion does not establish legal compliance for every recording or workflow.
The noScribe walkthrough covers selecting an audio file and output location, choosing a starting point, setting or detecting the language, and balancing speed against accuracy. The speaker describes it as a free, open source tool for Windows, Linux and macOS that uses Whisper. His test identifies speakers, detects German and saves the transcript as a text file. He then suggests removing names and other personal information before further AI processing, or using local AI software such as LM Studio to prepare meeting notes.
For Mac users, the speaker introduces MacWhisper as a paid, macOS-only option built around Whisper. He discusses recording Zoom meetings locally, identifying speakers and adding their names afterward, alongside app audio capture and dictation.
The final section demonstrates iOS Voice Memos as an on-device recording and transcription option. The speaker says it works offline and reports using it for talks longer than an hour. He advises checking whether transcription apps require an internet connection. He does not demonstrate an Android equivalent.
