I have a colleague who emailed recently needing a transcript from a podcast episode. How could she get one? It’s surprisingly easy.
If you do not already have the file (mp3/m4a/wav—or mp4 if it’s a video), you’ll need to get it.
Step 1: Download the recording’s file
I use a Firefox browser add-on called Video DownloadHelper. For Chrome, try CocoCut. Visit the website that hosts the recording. The browser add-on icon will change when it detects a file it can download. The Video DownloadHelper and CocoCut icons will go from black and white to color. If the add-on doesn’t turn color, try playing the recording on the website. That will help the add-on see that there is an audio or video file it can download. Once the browser icon changes color, click the icon. The audio or video file will be downloaded to your computer. (Unless it’s YouTube. YouTube specifically blocks browser downloaders from working.)
If the downloaded file is something other than mp3, m4a, wav, or mp4, visit cloudconvert.com to convert the file to one of those formats.
Step 2: Get the transcription from Office 365 Word
Go to office.com, and log into your work or “premium” ($6.99/month) Office 365 account. With either kind of account, you get 300 minutes of transcription per month. If you’re doing a lot of transcription—say, up to 6,000 minutes each month—take a look at Otter.ai for $8.33/month.
Open a new document. On the Home ribbon, click the down arrow next to Dictate and select Transcribe.
A panel will slide out on the right side of your screen. Click “Upload audio” to upload the audio or video file.
Once completed, you will see the transcription on the right side of your screen. The transcription remains attached to whatever document you had open when you asked Word to do the transcription.
If you know the names of the speakers, you can change them. Mouse over one of the time stamps, click the pencil icon. Type in the name of the speaker. Be sure to check the “Change all Speaker [x]” box.
If the sound quality of the audio is pretty good, Word should do fine with the transcription. It’s probably worth a proof listen, though. Click the play button in the transcription box. As the recording plays, the related transcription will be editable.
When you’re happy with the transcription, click the “Add to document” button at the bottom of the transcription pane. Choose whether you want just the text, the text and speakers, the text and timestamps, or the text, speakers, and timestamps.
Accent considerations
How does Word transcription do with accented English? I downloaded this audio file from the Speech Accent Archive and ran it through Word’s transcription. The transcription was spot on.
