Making High Quality Transcripts with Fanfare

Joel McKinnon, CPWA
Fanfare logo with name spelled in red block letters and party favor image

Podcasts have exploded in popularity in recent years with about half of the US population having listened to a podcast in the past month. Topics cover true crime, politics, sports, and entertainment including science fiction, the topic of my own Seldon Crisis, dedicated to Isaac Asimov’s Foundation series. Seems that soon everybody will be listening to podcasts.

Except that they won’t, because there are a lot of people out there who can’t listen due to hearing difficulties or deafness. Is it right to just cut them out of a major way that people are coming to consume content in our world? Of course not! The way to close this gap is with text transcripts. A recent study by the Pew Research Center found that only about 13% of the top 100 podcasts offer transcripts. For smaller, more limited audience shows like mine, that number is much closer to zero. As an independent, budget-minded podcaster, I can relate to the challenges of creating transcripts. Many of us do this in our spare time and it takes all the time we have to write scripts, record, edit, and produce each show. Who has time for more steps to add to our workflow? 

As an accessibility professional, I knew I had to find a solution to this challenge, and I want to share my methods so that others can meet this challenge as well. It all starts with finding a reasonably quick and accurate way to transcribe each episode. I found an excellent solution in a product made by Fanfare, and I’m going to show you how I recently transcribed all 44 episodes of my show and now offer free transcripts in multiple formats to anyone who prefers or needs to read rather than listening.

Go to Fanfare and create an account.

Fanfare offers two kinds of accounts, a DIY plan at a low monthly rate and a concierge plan that is significantly more expensive, but takes the time and work out of your hands. I’m going to lead you through transcribing for yourself using the DIY plan.

Go to the Fanfare website and click the Sign Up button. You’ll be presented with two options; select the DIY plan.

Select and link your show

Once you are logged in to your new account, click on My Show in the left sidebar.

In the Podcastindex field, begin typing the name of your podcast until you see it appear in the autocomplete list, select it and click the Link button.

When your show has been linked, the main content panel will come up with the settings for your show in the main panel like you see here.

Customize your show information as needed

Most of the important info is autopopulated, but you will want to add social media links for your podcast and to those player platforms where listeners can hear it. Scroll down to the Socials section and click the More toggle control as shown here.

Then add urls to any player platforms and social media that pertain to your show.

You can also modify the accent color that is used for highlighting the text in the transcript as the audio plays. You can do this at any time and changes will be in effect for all web based transcripts for your show.

Import an episode to transcribe

In the left sidebar, click Import Episode, the click the Import Episode button in the main panel.

I’m switching to a different podcast now, because I’ve already imported all the episodes for Seldon Crisis. You should see something like the following for your podcast. You can select one or more episodes to transcribe, but I suggest you start with just one. Here I’ve selected the first episode for the excellent Trojan War Podcast, “The Apple of Discord.” Click the Import button.

Select My Episodes in the left sidebar and you should see any episodes you have imported in the main column. Click on the title of the episode you want to transcribe.

The episode will automatically begin to be transcribed by the AI engine and the first draft will appear within a few minutes.

Assign speakers

Click the Edit Transcript button.

The first task is to assign speaker names for hosts and guests if present. The AI engine will automatically label different speakers as Speaker A, B, C, etc, but this isn’t always perfect, so it’s best to doublecheck that it got it right. In this case there is only one host, labeled as Speaker A. Click the Speakers tab in the right panel.

Replace Speaker A with the first name of the speaker, and click Save. You will see all of the labels in the transcript for Speaker A change to the name you’ve assigned. If there were multiple speakers you would add each of their first names here.

In the empty field with the placeholder text Speaker Name, enter the speaker’s full name and click Add

In the main panel, select the first instance of each speaker’s name, and change it to the full name using the dropdown selector.

Most podcasts have audio that is not spoken text such as theme music or sound effects. A deaf person reading the transcript may want to know about these, so you can create a speaker name for audio transcriptions by adding a speaker that consists of a blank space. This will show up in the dropdown list as a colored dot without a name. Place the cursor in the text where the sound occurs, press the Enter key, and type in the description in brackets. Then change the speaker name to the blank speaker name.

Use Find and Replace

One of the most common problems in the AI transcription is misspellings of proper nouns like names and places. In this example there are a lot of names of Greek gods and goddesses which are often misspelled in various ways. I’ll use one of these as an example.

Press the Find and Replace button.

Add the misspelled name to the first field and the proper spelling to the second field. You will see a list of all the occasions of the misspelled name in the text. You can then press the Replace button to correct all of these at once.

Look through the transcript for other kinds of errors. Sometimes you will notice punctuation errors, paragraphs chopped awkwardly, cross talk errors where two speakers are speaking simultaneously, and other issues. You can click the play button to listen to the audio as you transcribe anything you’re not sure of. When you think you’ve got it all just right, press the Back to Transcript button.

Export and Publish

You should see a new menu of options in the right panel. Select the Export tab.

Press the Download button for any file type you would like to export. I typically create Word and PDF exports and offer these to my listeners.

Press the Publish tab.

Scroll down and select the Visible radio button, and click Save.

Find your Active Transcript page

Select the My Show tab in the left sidebar again. You should see a link below the title of your show. Click this link to open the list of published episodes for your show.

As you create more transcripts they will be added to this list. Click on the title of the episode you just transcribed to be taken to the Active Transcript page. 

Play your Active Transcript

When the page is loaded, click the Play button as shown. The text of the transcript will be highlighted as the audio plays with a moving underline (you can change the color by modifying the Accent Color in your My Show settings). You can change the speed, move forward and backward in the audio, and turn on the Autoscroll feature for hands free enjoyment if desired.

Share link to your Active Transcript!

Just copy the url for the transcript page and share it in emails, texts, or on social media. In the next part of this tutorial, I’ll show you how to how to make your PDFs more accessible with Adobe Acrobat Pro and to host your PDF and Word format transcripts in the cloud using Amazon AWS S3.