How to create a transcript of part of a video
June 22, 2023 12:16 PM Subscribe
I'm trying to create a transcript of a 10 minute portion of a video. Is there a better way than what I'm already doing?
There's a section of a video that I really want to be able to read and study, so I started creating a transcript. I'm not technically savvy at all, so what I did was open up NOTES on my iPad and enabled dictation (which I'd never used before). Then I played the section of the video on my computer, with my iPad right next to the computer.
It worked, but with lots of mistakes. Also, it seemed that the dictation stopped at regular intervals (based on time or number of words maybe), so it missed parts of some sentences. I estimate it will take at least another hour to finish the transcript, including replaying the video and doing manual cleanup.
Is there a better way, with these requirements: (1) I only want to capture 10 minutes of a one hour video. (2) It has to be simple because if it takes me more than an hour to set up and learn, then I'll just do the cleanup manually. (3) It has to be free. I'm willing to pay for things, but this is likely a one time thing, can't imagine needing to do this again.
There's a section of a video that I really want to be able to read and study, so I started creating a transcript. I'm not technically savvy at all, so what I did was open up NOTES on my iPad and enabled dictation (which I'd never used before). Then I played the section of the video on my computer, with my iPad right next to the computer.
It worked, but with lots of mistakes. Also, it seemed that the dictation stopped at regular intervals (based on time or number of words maybe), so it missed parts of some sentences. I estimate it will take at least another hour to finish the transcript, including replaying the video and doing manual cleanup.
Is there a better way, with these requirements: (1) I only want to capture 10 minutes of a one hour video. (2) It has to be simple because if it takes me more than an hour to set up and learn, then I'll just do the cleanup manually. (3) It has to be free. I'm willing to pay for things, but this is likely a one time thing, can't imagine needing to do this again.
You can use otter.ai for free if you just have one thing. It's pretty easy to use.
posted by pangolin party at 12:58 PM on June 22, 2023 [1 favorite]
posted by pangolin party at 12:58 PM on June 22, 2023 [1 favorite]
We use a tool called Temi (temi.com) for this. We pay for it because we use it all the time, but your first transcription is free (up to 45 minutes). We have great results with them.
posted by anastasiav at 6:40 PM on June 22, 2023
posted by anastasiav at 6:40 PM on June 22, 2023
Microsoft has a tool called video indexer that does a whole bunch of things, one of which is to create captions. You need a Microsoft or Google (Gmail or hosted domain) account.
But if you upload the file there, you'll see captions and if you click the "timeline" on the right that is the text synced to the video.
Under the "Download" button in the top left, the captions can be downloaded in TXT or CSV (excel) format. The other formats (SRT/VTT etc) are timed formats used with video players and are not what you want. You can even get captions in other languages.
posted by bitdamaged at 7:43 PM on June 22, 2023
But if you upload the file there, you'll see captions and if you click the "timeline" on the right that is the text synced to the video.
Under the "Download" button in the top left, the captions can be downloaded in TXT or CSV (excel) format. The other formats (SRT/VTT etc) are timed formats used with video players and are not what you want. You can even get captions in other languages.
posted by bitdamaged at 7:43 PM on June 22, 2023
I've had good luck with Macwhisper, it's free (improves quality for $20), and nothing leaves your machine.
posted by duende at 11:55 PM on June 22, 2023
posted by duende at 11:55 PM on June 22, 2023
I've had really good luck with using the free tier of Descript to do exactly what you're describing.
posted by Doktor at 8:18 AM on June 23, 2023
posted by Doktor at 8:18 AM on June 23, 2023
I also use Otter for regular meetings and like that - you can upload video or audio. One issue with transcription on all of these platforms, and indeed human transcription, is that it won’t be perfect, so you will have to correct it.
Some of the platforms make that really easy, so for example with Otter you can play the audio from a section and then use edit mode. Others not so much, meaning you might need to hit play on the video and fix your transcript in Word or whatever.
Rule of thumb from research transcription is to allocate 1 hour of time for 15 minutes of audio, and that’s using a foot pedal. Using one of these services speeds that up some, but you’ll still have to put in time if you are wanting anything close to 100% accuracy.
posted by ec2y at 2:12 PM on June 23, 2023
Some of the platforms make that really easy, so for example with Otter you can play the audio from a section and then use edit mode. Others not so much, meaning you might need to hit play on the video and fix your transcript in Word or whatever.
Rule of thumb from research transcription is to allocate 1 hour of time for 15 minutes of audio, and that’s using a foot pedal. Using one of these services speeds that up some, but you’ll still have to put in time if you are wanting anything close to 100% accuracy.
posted by ec2y at 2:12 PM on June 23, 2023
I have used a version of Whisper with really good results, so I encourage you to try duende's suggestion of Macwhisper (or any other version of Whisper that will work for you).
posted by kristi at 2:22 PM on June 23, 2023
posted by kristi at 2:22 PM on June 23, 2023
This thread is closed to new comments.
posted by wemayfreeze at 12:45 PM on June 22, 2023 [1 favorite]