Hello everyone! I lead project to provide transcripts for the podcast, which is great for both accessibility for people who can’t listen to podcasts or find it easier to read transcripts, and for searching things that were said in podcasts. Now with SUSD’s heroic pledge to release a podcast every week this year, the current transcription team (i.e., me and on rare occasions my wife) needs some extra hands so that the transcription doesn’t slow down too much and/or we don’t get burned out and abandon it altogether. That’s just a reality of the amount of time something like this can take, but the good news is that many hands make light work, and transcription can be fun and rewarding!
So with that in mind, for anyone who would like to contribute to this project, please send me a message so that I can add you as an editor in the Google Docs folder. Here are some FAQs:
What’s the time commitment? A: No commitment necessary! Contribute as much as you can, when you can.
Is there a set of transcription/formatting guidelines to follow? A: There are no official guidelines, so you can transcribe in whatever way is easiest for you. You don’t have to follow the style of the current transcriptions, as long as you’re including the names of the speakers. I will be going over any transcriptions to clean up and correct errors (a much quicker process than the transcription itself), so that will help smooth any minute style differences.
Are there any tools I can use to help transcribe better? A: Yes! There is a free software called Express Scribe that has playback controls, different speed settings, and pause-and-repeat functionality that I’ve found very useful.
If you have any other questions, post them below and I will answer them and add them to this list. Again, if you’re interested, message me and I can get you set up. It’s a good, low-pressure way to contribute to something minor but meaningful and support a team that provides so much excellent content for us.
I don’t have any question, but whoa am I amazed to learn that the “amazing community” that transcribe every podcast is actually just you and your wife. Impressive work!
I am a journalist so obviously I keep an eye on promising transcription technology and I strongly encourage you to check out the tools that are now available in Office 365 on the web. Without much fanfare Microsoft has implemented speaker independent voice recognition and a great interface for tidying up the results similar to what you get with Express scribe. In other words, it’s like using Express scribe but with a lot of the work done already. There’s a limit to length you can use but I presume it would be possible to divide the task among several users to get around that limitation.
If you’re prepared to pay, there’s also Otter, which @Shimmin is using.
Hey @Derbius ! I don’t think we’ve ever had anyone make a first post thats so helpful. Welcome!
Thanks for the suggestions! I think as a volunteer project it makes sense to go as cheap (i.e. free) and that has worked well for me so far, but if anyone else wants to try something with a few more features here are some ideas.
I find Otter very good, and am looking at using it for IRTD episodes.
I don’t get close to the data cap, so I could potentially get some autotranscriptions done for you, but I can’t help with the real work.
Now this is interesting. Is that using a speech-to-text function? That might be good to get a base to work off of, at least it’s worth a try.
It seemed to me not unlikely that some were all volunteers might have Office 365 licenses anyway so using it would not cost anything extra.
Parenthetically, I used to be a technology journalist in a past life and I was always deeply skeptical about anything that would claim voice independent speech recognition but I was stunned when this Microsoft product came out and even more surprised that more was not made of it…
Yes, it’s actually pretty good - especially if you add likely custom vocab ahead of time. Even better if you can upload a sample of your voice, but that isn’t much use unless you’re only recording yourself. It can even split up and label speakers with some success.