Please follow and like us:
RSS
Facebook
Twitter
YouTube
Pinterest
LinkedIn
Instagram

Interviewing a source can be an exciting part of your research, but hours-long transcription jobs can bog down the writing process. Transcription software might be the answer you’re looking for.

I consider the audio recorder on my phone to be one of my most valuable technological assets when it comes to writing. I record myself brainstorming notes and ideas on the fly, capturing creative phrasings or colorful images in ways that feel more articulate coming out of my mouth than through my thumbs. I also use my phone to record interviews for all sorts of projects I’m working on.

As great as it is to have easy, convenient, and nearly unlimited recording potential literally in your back pocket, the drag comes when it’s time to turn those audio files into words on a page. Transcribing can be tedious, especially if you’re trigger-happy with the “record” button and have lots of files to sift through — plus the prospect of massive, hours-long transcription can be a barrier to even beginning the research and writing processes.

At times, I’ve hired colleagues and interns to transcribe recordings for me, which is always helpful, but can get expensive. Most often, I’ve put on a set of headphones, poured a good cup of coffee, and gotten it done myself.

There is much to be said for muscling through. A writing colleague of mine once said that re-listening to and transcribing interviews made him absorb and experience the material in a completely new way, setting him up perfectly for the process of writing. So there’s a time and place for that approach, especially if you have the space in your calendar to make it happen.

There’s also a time and place to skip it entirely.

While working on a recent project for a private client, I learned about Temi.com, an online transcription software service that uses AI audio analysis to transcribe files for you. Since my project required the careful review and parsing of more than three hours of interviews — and the deadline was tight — I was intrigued and tried it out. Here’s what happened.

After a simple registration process, I uploaded an audio file and Temi got to work. Within twenty minutes, I received an email saying the process was done — and there was my transcription, laid out within Temi’s easy-to-navigate web interface, broken up by individual speakers, and ready for me to dig into.

That first transcription was free. Subsequent files were billed at ten cents per minute. Transcribing my full three hours of dialog? Around $18.00. Not bad at all.

Now, three hours of heated conversation translates into a whole lot of words on the screen. Being presented with pages and pages of transcribed material can be intimidating, and locating the content you need within that expanse of text can be a challenge. Fortunately, my interviews were recent enough that it was easy to remember the twists and turns of conversation, and what topics led into what. Given that, my process boiled down to scanning the transcription quickly for the topics that were most relevant to my article, or simply using Temi’s uber-helpful search function to find keywords, quotes, and snippets.

The transcription’s accuracy was, in some places, perfect; other times, it was jumbled and not easy to decipher. Some phrases, like this one, even came out sounding like beatnik poetry:

You’ll see keys. Cool. Who stick caught up the monster wallet. Robbins.

Luckily, Temi has a function that lets you easily listen back at any point in your original recording. When you’re logged in to the service, simply click on the sentence you’re having trouble with and Temi will start audio playback at that exact place while visually tracking through the transcription, word for word, perfectly in sync. I used this function repeatedly during my writing process to double-check quotes. It saved me a huge amount of time versus having to manually track back through my recording, pinpoint exact time codes for certain quotes, and check for accuracy.

The results? For my 3,000 word article, the service saved me a minimum of four hours, and possibly more. That’s a big deal.

If you’re thinking of using Temi, or any transcription software or service, here are a few tips.

  • Try to make your recordings as clean and clear as possible. Understandably, the service does its best work when ambient noise is at a minimum and spoken words are easy to understand.
  • Email yourself the transcriptions for backup. Temi makes this easy.
  • Make sure your files are labeled clearly and accurately from the start — “Tammy Smith interview 4/5/16 for government story background” as opposed to “TWH78836697.mp3.” I accidentally uploaded and transcribed the same file twice because both were labeled in the latter convention — a mistake I do not plan to repeat.
  • Edit your files before uploading them. Since Temi bills by the minute, an hour recording containing ten minutes of dialog costs just as much to transcribe as an hour-long file that’s filled to the brim with conversation. Perhaps future software versions will detect and edit out silences from the get-go, but in the meantime, it’s worth a few minutes to trim your files before uploading.
  • If you need a transcription to be completely accurate, word for word and from start to finish, you can efficiently edit the transcription within Temi. The “find and replace” utility is a great help. If your interviewee has a heavy accent, for example, perhaps Temi will regularly mistake the word “speech” for “spinach.” If a misidentification happens more than once, you can correct it throughout an hours-long transcription with a few key clicks.

Is Temi a perfect tool? No, but it doesn’t have to be. Did the service help me complete a challenging writing project in record time, and would I use it again? Without question. To anyone with transcription needs who is also writing against the clock, I recommend giving it a try.

How do you handle audio recordings and transcriptions in your own writing? Tell us in the comments below.

 

Free BookBaby Catalog - Your path to 
publishing

 

Related Posts
The Accidental Novelist – How Stolen Moments Can Make A Book
Conducting Interviews For Research: Identifying And Contacting Expert Sources
How to speak your book into existence
Write By Accident, Refine By Design
Conducting Interviews: Directing And Capturing The Conversation

 

Michael Gallant

About Michael Gallant

Michael Gallant has written 10 posts in this blog.

Michael Gallant is a writer, musician, composer, producer, and entrepreneur. He lives in New York City. Follow Michael on Twitter at @Michael_Gallant.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.