Advice needed on a decent speech-to-text software program.

It doesn’t have to be free, but it doesn’t need to be professional quality, either. I need something that I can mess around with and learn the ropes on. I also would rather pay a few bucks than deal with ads and stuff.

Suggestions?

[UPDATE: What comes with Windows is, on first look, horrible. I want to transcribe interviews, which means that I need a success rate that’s better than 40%. If that.]

11 thoughts on “Advice needed on a decent speech-to-text software program.”

    1. We use it with kids who are too involved to type or do a lot of writing. However I do know that it has to learn your voice and patterns before using it so I dont know how it would do on your subject’s part of the vocals if they werent set up on it first….

  1. Some of the authors I follow swear by dragon.
    .
    Which means it has improved a lot since I tried it in the late 90s. I just swore at it.

  2. Do the current versions of Dragon still require learning a person’s speech patterns before using? Because that could make transcribing an interview with a stranger a bit difficult.

    1. Keeping in mind that my experience with the program was over 15 years, and an unknown number of versions ago…

      The accuracy rate was about 80% out of the box. You could train it to recognize your voice with greater accuracy.
      My regional dialect features soft consonants and a vowel lilt that the program had a great deal of difficulty with. Even with exaggerated enunciation, I never got it better than 90%.

  3. Learning the speech patterns can’t be that hard for Dragon. We used it at the Army Hospital where I used to work and Physicians don’t have patience for a long learning curve. Many of them had very short attention spans and difficulty following instructions, but they learned Dragon.

  4. The Spousal Unit uses Dragon, but it will not work for multiple voices.
    .
    From Dragon’s FAQ page:
    .
    • Can Dragon be used to transcribe meetings or interviews?
    .
    No. Given its amazing recognition accuracy, many assume that Dragon speech recognition would be an ideal solution for meeting or interview transcription. But Dragon is not designed to transcribe multiple voices since Dragon software is “speaker dependent” speech recognition.
    .
    A suggestion is to use Dragon to “parrot” the other speakers by listening to the recording of the meeting on headphones and repeat what you hear into your computer mic…but that sounds tedious and time consuming.

    There is a cloud based app by Koemei but I think the monthly fee is pretty pricy.
    .
    Here’s a workaround you might try:
    .
    <a href="https://www.youtube.com/watch?v=5eZJa-iM2Gw"A down and dirty way to convert audio to text using YouTube transcription services.
    .
    Given the popularity of podcasts it’s hard to believe someone hasn’t published an app.

Comments are closed.