subscribe to the RSS Feed

How to Create a Transcript for Podcasts

Posted by Darrell Hyatt on June 14, 2008

Here is the transcript of my latest podcast. I created this podcast in response to interest in how I created the transcripts of my podcasts.

Welcome to the Enabling Abilities to Appear podcast. My name is Darrell Hyatt. In this episode I will be talking about a piece of software called Dragon Naturally Speaking. This particular software is a voice-recognition system for people have difficulty using the keyboard and want to speak to type. Although the software has many uses and many advantages over typing, I’m going to be talking about it today from the perspective of using to create the captions and transcripts for an audio podcast.

This is the software that I use not only for the dictating my documents for my business, but I also use it as I say to create the transcripts for my podcasts as well as a captions for my video projects.

First I will start by giving you some details about Dragon itself. The current version that is available is version 9. You can get the software in four different varieties — there’s the standard version, preferred, the medical version or a legal version. I am using the preferred version of the software which is $199. The standard version is $99.
Now, in terms of the computer, you need to run the software according to the Dragon web site which is www.nuance.com, the preferred version requires a Pentium 4 processor or equivalent of 1GHz and a minimum of 512MB of ram. I would strongly recommend that you at least double that memory requirement. It also requires that you have a sound card, and if you want to use the Internet browsing capability of Dragon you need Internet Explorer 5.0 or higher. As long as your system means those minimum requirements you will be fine when you install the software it automatically checks to make sure that you do meet the minimum requirements if your system does not it won’t install the software for you.

In terms of creating the transcripts and so on, the first step is to record your audio for your podcast, which is what I’m doing at this very moment. I’m using a software package called Audacity and this is a free audio recording program that you can download from the Internet. I’m using this one because I found that for my voice it tends to work best with the recognition of Dragon, which I will talk about shortly.

Once you’ve recorded your audio file, you want to save it as a .wav file and so the first step once you open audacity you want to make some changes to some of preferences settings, which are under the edit tab. The first of which you want change is the audio input output tab and you want to make sure that your playback and recording device is set to your headset. When you purchase Dragon you will get a headset as well as the software. Once you’ve installed Dragon, go into Audacity or whichever audio recording program you using and make sure that your playback and your record device are set to your headset.

The next one that I want change in particular for Audacity is the channel. There you want to make sure that your only using one channel and creating a mono recording. Tthis is for two reasons: 1) a mono recording will be smaller in size, and 2) Dragon does not like and does not transcribe well with a stereo recording.

The next tab in Audacity that we want change is the quality tab. Under there we have the sample rate which we want to have as 22050; again this is a requirement for Dragon to be able to create our transcripts. The next one, again under quality, is the sample rate and you want that to be set for 16-bit.

The final tab that we want to make some changes to in Audacity is the file format tab. Here we want to make sure that the file format is set to wav 16-bit pcm. And finally, we want to make sure that under mp3 we’ve got Lame (yes, Lame!) selected there.

Now that we have made all the changes we want to our Audacity settings, you can go ahead, as I said, and record your audio podcast and make sure that you save it. Once you’ve saved it the next step is to go into Dragon itself. Load your user settings and, then once it comes up, go into the sound tab. There you will see transcribe your recording and here is the key to creating a transcript.

Tell Dragon that you want to transcribe your recording and it will come up with a dialog box that asks you which type of recording is it — is a wav file or is it from a Pocket PC or a mobile recording device? Under wav file, you simply browse until you find where you saved your audio recording and then Dragon will bring up its DragonPad, which is the dictation window and it will start transcribing your audio recording of your podcast. If it’s not in the proper format of wav and 16-bit and so on, Dragon will not transcribe it. You may have to fiddle with it, but hopefully with the changes you’ve made in Audacity or whatever audio recording program you are using, it will begin to transcribe your text. It will just be a straight transcription word for word with no formatting

Now you have to go ahead and transcribe you recording after you’ve made it, before you make any editing changes to the audio file in whichever editing program you us. Generally, editing the audio file first will make the recording unusable by Dragon. Therefore you have to transcribe the recording right away before you do any editing and then edit the file after you’ve created the transcription. Then, of course, you must do any editing you want on the audio file before you post to your web site, your blog or wherever you’re going to put the finished recording.

Once you start working with a transcript, you will want to check it for accuracy because, even if you’ve used Dragon for quite awhile and trained it as I have, there still can be some errors in the transcription of Dragon. What I recommend is having an audio player opened as well as your transcript text so that you can listen to the audio file and read the text at the same time to make any changes that are necessary to the text so they match your audio file.

After that, it’s simply put the audio and text together on your web site as you normally would and now you have the ability to create transcripts for any audio project that you wish, simply and easily as well as have a program that you can use for dictation in other areas of your business for other projects.

I see Dragon as a worthwhile investment for many reasons. Dragon can help you in many different ways throughout your work day. If you have any questions comments or concerns or any show ideas you’d like to see presented here please e-mail me at Darrellh@enablingabilities.com. Please feel free to visit my web site www.enablingabilities.com to see all the resources that I have available I look forward to hearing from you.

  • Tools and Tips for Creating Transcripts | Blog Accessibility said,

    [...] Hyatt explains how to use voice recognition software Dragon Naturally Speaking to create transcripts (This method tends to work best when only one person is [...]

  • Bill said,

    I generally find it easier and more accurate to “echo” the audio file into Dragon. If it’s slow enough you can even put in paragraph breaks for starting a caption file.

  • Darrell Hyatt said,

    I am not sure what you mean by “echo”, does that mean you record the audio & then listen to it as you dictate it into dragon?

  • Eric said,

    Podcasts for school, do you recommend this software? The school uses RSSFeeds available for download as wmv files.

  • Carey @ TranscriptionPRO said,

    It may be better at times to outsource transcription to professional transcription providers; they are affordable and accurate – don’t rely on Dragon for your transcription requirements.

  • Darrell Hyatt said,

    Very true Carey, Dragon does require careful editing for creating transcripts

Add A Comment

home | top