WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: video transcripts for talking heads: need tobesynchronized?

for

From: John Foliot
Date: Apr 8, 2015 5:54PM

Morin, Gary (NIH/OD) [E] wrote:
>
> If the video is important, how much work is really involved in timing
> and synchronizing the captions with the speaker? Is there that much
> time and cost involved that it makes a serious business reason to nix
> synchronization?
>

The time and cost is in getting the actual transcript of the audio portion of
the media. If it is a pre-approved script that is being presented by the
smiling talking head, then the hard part is already done. If it is a general
conversation [sic] that is not pre-scripted, then transcribing the audio will
take time (which = money). With a bit of practice, I have seen roughly 6
minutes to 1 (or, a 5 minute video would take roughly 30 minutes to
transcribe). For low volume producers, this may prove to be the most
cost-effective.

"Time-stamping" the transcribed text can be achieved in a number of different
ways: it can be done manually using free tools like MAGpie
(http://ncam.wgbh.org/invent_build/web_multimedia/tools-guidelines/magpie), or
Microsoft's HTML5 Video Caption Maker
(http://ie.microsoft.com/testdrive/Graphics/CaptionMaker/) where, again with
practice, it will take roughly 1 - 2 minutes per each minute of real time
audio (or, up to 10 minutes for that 5 minute video), or it can be done
programmatically (as previously noted), by using services such as YouTube, or
tools & services such as what DocSoft offers
(http://www.docsoft.com/Products/), Ceilo24
(http://www.cielo24.com/solutions/media/), or 3PlayMedia
(http://www.3playmedia.com/services-features/tools/transcript-alignment/)

HTH

JF
------------------------------
John Foliot
Web Accessibility Specialist
W3C Contributor - Accessibility
Co-Founder, Open Web Camp