Whisper AI: Automatic Audio Transcription
Transcribing audio and video files (e.g., from expert interviews or focus groups) is often the most time-consuming step in qualitative research. With Whisper AI, we offer you a tool in our computer labs that saves you a tremendous amount of time.
Whisper is a state-of-the-art artificial intelligence (AI) system that automatically converts spoken language into text with impressive precision.
Data Protection and Security:
Unlike cloud-based services on the internet, Whisper runs 100% locally on the hardware of the computer lab PCs. Your sensitive research data never leaves the university and is processed directly, in compliance with data protection regulations, and securely on your personal university network drive (Drive S:).
Availability:
The tool is available to you on all computers in the computer labs of the Center for Network Media.
Step-by-Step Guide: How to Use Whisper
Operation is fully automated and requires no prior technical knowledge.
Step 1: Start & Record
Launch the “Whisper Transcription” program by clicking the microphone icon on the desktop of the computer lab PC.
A black information window will open, and your personal network drive (S:Whisper_Transkripte). Now simply copy your audio or video files (MP3, MP4, WAV, M4A) into this exact folder.
Step 2: Start Transcription
Switch back to the black program window. Type the word “YES” to confirm and press the Enter key.
The AI will now automatically process all media files in the folder. Color codes (green, yellow, red) show you during the live preview how confident the AI is in its word recognition.
Continue to Step 3 →
Step 3: Retrieve Results
As soon as the process is complete, a green success message will appear. You’ll now find the finished transcripts on your network drive next to your original audio file.
The system automatically generates multiple formats for your further work:
- .txt (plain text document)
- .json (For structured data analysis)
- .srt (subtitle file for videos)
- _MAXQDA.txt (Directly optimized for MAXQDA import)
Pro tip for working from home:
If you set up the SOFS drive on your personal computer, you’ll have convenient access to your finished transcripts right from home. Equally convenient: You can upload your audio and video files to the drive in advance from home and then transcribe them directly in our pools.
The direct network path is:sofs1.uni-koeln.deihr_benutzernameWhisper_Transkripte
What happens after transcription?
Once you have the raw text, the actual scientific analysis (known as coding) begins. In addition to MAXQDA, which is available on selected pool PCs, we recommend the powerful open-source software QualCoder for working from home on your own PC.
Video Tutorial: Import and Analysis in QualCoder
This basic tutorial (Open Educational Resource) introduces you to the key fundamentals of the free QualCoder software and provides a practical demonstration of how to import your transcripts—freshly created with Whisper—and analyze them methodologically:
If you have any methodological or technical questions, please feel free to contact the Center for Network Media:
✉ computerpools-hf@uni-koeln.de