#captionTHIS today! I’m in.

If you do not understand American Sign Language, please click on the ‘CC’ button on this YouTube video to understand.

One thing that I’ve been frustrated with is not just the lack of captioned videos out there, but the hurdles faced to add closed captions to uploaded YouTube Videos. For instance, YouTube’s Help Page for adding and editing captions can be found here. In it they say:

To add captions or subtitles to one of your videos, you’ll need to have transcript or caption files with the captions/subtitles in them. Please see this entry for detail.

If you visit their page to help create transcript files, you’ll soon realize that it’s not an exact science. In fact, the file that works best that you should create is what is called a Caption File. YouTube suggests you check out this help resource. Maybe one or two of these resources will prove to be more efficient in providing captions than the methods I’ve used in the past. No matter, the point here is that it’s kind of a pain to provide captions (though I’d argue the pain is worth the expanded potential audience).

But for heaven’s sake, please do not trust YouTube’s attempt to provide captions based on it’s audio recognition algorithm. If you need an example, watch any of Rhett & Link’s ‘Caption Fail’ video experiments! In these videos, they create a video and allow YouTube to create captions on the fly. They are always way off. So this dueo uploads a sketch, takes the YouTube ‘auto-captioned’ text as a script for Take 2, then do it all again. Here, I’ll show you…(and trust me, for this video example, turn your CC off)

Bottom Line: Auto-translation via audio input is very imperfect. Translation via text is fairly advanced, but to add a viable English-based caption file is currently cumbersome. I long for a day when uploading an ASL-based video is easily translated into written English easily.

This entry was posted in culture, technology, video. Bookmark the permalink.