Captioning and transcription

Transcription, or converting a video or audio track into text format, is a modern way of producing, for example, captions or text equivalents from video/audio sources. Artificial intelligence will give us a good base to work with, which can then be edited manually afterwards if necessary. This serves us both when it comes to transcription needs in research and the accessibility requirements for videos published online. You can also use transcription with webinar recordings and podcasts, for example.

Videos published online must be accessible, i.e. captioned. In addition to people with impaired hearing, other people also benefit from video captions. The video can be viewed in a location where it is not possible to increase the volume, the video could have poor sound quality or background noise could make it hard to hear the video. 

Various tools for captioning, subtitling, transcription and dictation are listed here.

• Word Dictate (see dictation instructions on Microsoft’s website)
  See separate instructions for transcription (on the Microsoft website)
  Versatile use for dictation and transcription needs. Also allows diarisation, i.e. automatically separating the speech of different speakers.
  Please note! Transcription in Word is limited to 300 minutes per month.
• AvidNote AI assistant, read more in Flamma’s news article (login required)
  A handy transcription tool that allows diarisation, i.e. automatically separating the speech of different speakers.
  Please note! AvidNote allows time-limited transcription.
• Unitube automated captioning
  Captioning (videos only), (transcription available from caption file). 
• Live captioning in Microsoft PowerPoint live (Microsoft instructions)
  Microsoft’s tool for captioning and dictation.
• Automatic captioning with Microsoft Stream
  For captioning and dictation.
• Captioning with the ScreenPal software
  Captioning tool.
• Subtitle Edit subtitling tool 
  Subtitling tool.

Please note! When using AI assistants, the University’s general cloud service principles and AI guidelines must be followed. Licensed materials, personal data or sensitive data may not be entered into the programmes. 

1) Plan everything well in advance (main structure/script for the video).
2) Handle the captioning using your preferred method. Various captioning program options are listed above. 
Please note that you can either burn the captions permanently into the video OR add a caption file in Unitube Uploader. If you are using a caption file, please do not burn the text into the video.
3) Publish in Unitube, for example.

• cc (closed captioning) = video caption track
• Speech to text = speech recognition service that recognises spoken text and generates text based on it.
• Sync point = synchronisation point used to mark where the text starts. The software identifies such points automatically, but you can also add your own sync points.
• Scripted = the video has been scripted

Subtitles and captions

• Zoom, streaming and instructional videos (Instructions for Teaching) 

Accessibility

• Digital accessibility of teaching (Instructions for Teaching)
• Accessibility instructions on the Celia website
• University of Helsinki’s accessibility instructions
• Accessibility in the University’s digital services (in Flamma)
• Accessibility (HY’s content creator’s guide – Flamma workgroup)

Artificial intelligence

• Use of generative AI in research
• Generative AI at the University of Helsinki instructions on the Helpdesk page