Here at Spoolcentral we have crowdsourced our fair share of video transcription, translation and subtitling and it's not as simple as it looks. Our JobSpooler platform automatically divides video and audio media into chunks for assignment distribution to our crowd of Spoolsmiths. So far, so good. However, we've been looking at ways of integrating all the tools needed for transcribing and subtitling all in one place (automated timestamping, line-length control, error correction, etc.).
We don't like all the errors that get introduced when you aggregate a bunch of excel or word files together. We'd like to simplify and streamline the process further. I bet you'd like that too.
We've played around with the whimsically named Subtitle Horse which has a lot of nice features (keyboard controls, timecode offset, easy text editing) and a compelling price (nil). It isn't quite industrial strength though. It would be nice if it read quictime files and not just Flash Video and it would be fab if it could read a timecode track, either embedded or fed in manually.
We were excited to hear that the Center for History and New Media at George Mason university in Virgina received an NEH grant to create an open-source transcription tool specifically designed with crowdsourcing in mind. We'll report back we we get a chance to check out what they have in mind.
