jethrien: (Default)
I'm trying to transcribe an interview I filmed with a curator. I'm going nuts, and it's taking forever. I guess the real problem is that I type a little slower than he talks, so I keep having to pause the video every couple seconds and catch the typing up. So. Boring. And it really makes me hyper-aware of how long people can talk for, and how many filler syllables get thrown in. (And he's unusually good at this, too - I've transcribed much worse.)

Oh god, I'm at 33:40, and the bar isn't even 2/3 of the way through yet...

Date: 2008-12-17 10:01 pm (UTC)From: [identity profile] shnayder.livejournal.com
If you do this kind of thing often, look into getting some speech recognition software to do a first pass that you just have to correct. But yeah, transcription is really annoying....

Date: 2008-12-17 10:04 pm (UTC)From: [identity profile] edgehopper.livejournal.com
Stenographers use special keyboards and tricks (they go at one combined keystroke per syllable, rather than per letter), and before the advent of typewriters, they used tricky shorthand methods. Almost no one types fast enough to transcribe on a normal keyboard, QWERTY, Dvorak, or otherwise.

Date: 2008-12-17 10:07 pm (UTC)From: [identity profile] jethrien.livejournal.com
Not often to justify that - two or three times a year, basically.

Date: 2008-12-17 10:07 pm (UTC)From: [identity profile] jethrien.livejournal.com
I did not know that. Hey, that's random information, and makes me feel better about my typing speed.

Date: 2008-12-17 10:11 pm (UTC)From: [identity profile] edgehopper.livejournal.com
During the deposition I attended back in October, we had a program called LiveNote that displayed the transcript as the stenographer typed. The current state of the art is pretty cool--as the stenographer types, the program displays the typed syllables phonetically, and then software automatically tries to guess the correct word from context at the word breaks. Afterwards, the stenographer reviews the transcript to fix errors. Check "Stenography" on Wikipedia for more cool info.

Date: 2008-12-17 10:20 pm (UTC)From: [identity profile] chuckro.livejournal.com
I actually knew this--In one of my classes senior year, there was a guy who was severely hearing-impaired, so he actually had a stenographer come to the lectures and transcribe them so he could read them off a laptop while the professor spoke. I ended up sitting behind him fairly often, because I found the subtitles really helpful, myself.

Date: 2008-12-18 12:13 am (UTC)From: [identity profile] towk.livejournal.com
I'm doing a lot of nitty-gritty video and sound editing for my company, so I know a lot of hacker tools for working with those sorts of files.

It's actually possible to extract the audio and slow it down by whatever ratio you want, using some free software such as

http://www.surina.net/soundtouch/

The technical details are tricky. If this is something you'll be doing often, I could help. :)

Date: 2008-12-18 12:41 am (UTC)From: [identity profile] hyouneko.livejournal.com
A friend of mine in Seattle is taking classes to become an Stenographer and I've actually seen her practice keyboard. It's a pretty interesting little device.

And I agree about being able to type what people have to say. I'm generally a very fast typer (as much time as I spend at a computer, it's not surprising) and I have problems keeping up with speech.

Date: 2008-12-18 12:50 pm (UTC)From: [identity profile] jethrien.livejournal.com
Again, I'm only doing this two or three times a year - I suspect it would actually be faster to just finish rather than try to muck around with a technical solution.

Date: 2008-12-19 11:40 pm (UTC)From: [identity profile] fyrna.livejournal.com
Transcription is really hard. I take minutes at a lot of W3C meetings, and given that my top typing speeds are upwards of 90wpm, the result looks a lot like a transcript.

But it's not. I transcribed a speech off a recording once and it's much, much more painful. When I'm minuting I'm effectively paraphrasing whatever the person just said. Filler syllables and phrases usually get dropped, discursive sentences get summarized to be more concise. When you're transcribing, you want word-for-word, and that sucks a lot more time spent transcribing per minute of speech.

Date: 2008-12-20 04:12 am (UTC)From: [identity profile] jethrien.livejournal.com
I do quite well at note-taking. But yes, it's amazing how much of most people's speech is either filler or circular. It's not so hard to get the gist of what they're saying down. But with this, I'm trying to get word-for-word, with frequent timestamps so we can go back and edit.
Page generated Jan. 18th, 2026 11:36 am
Powered by Dreamwidth Studios