Voice-to-text Apps

dr_mabeuse

seduce the mind
Joined
Oct 10, 2002
Posts
11,528
I'm wondering if anyone here has ever tried using a voice-to-text app for writing. They were pushing Dragon a lot over the holidays, and I understand the programs are getting pretty good, and recently I heard about some author who won a big award for a novel he'd written using a speech transcription app. He'd dictate in the morning and then correct the text in the afternoons, and even so he said he wrote the book in record time.

Has anyone ever tried this?
 
I'm wondering if anyone here has ever tried using a voice-to-text app for writing. They were pushing Dragon a lot over the holidays, and I understand the programs are getting pretty good, and recently I heard about some author who won a big award for a novel he'd written using a speech transcription app. He'd dictate in the morning and then correct the text in the afternoons, and even so he said he wrote the book in record time.

Has anyone ever tried this?

The 'IBM' software for this is very good, but all of the software takes a good while to 'train' to get to understand an individual's characteristics.
 
One of my publishers has Dragon for his writing and says it's been too frustrating to try to figure out.
 
I'm wondering if anyone here has ever tried using a voice-to-text app for writing. They were pushing Dragon a lot over the holidays, and I understand the programs are getting pretty good, and recently I heard about some author who won a big award for a novel he'd written using a speech transcription app. He'd dictate in the morning and then correct the text in the afternoons, and even so he said he wrote the book in record time.

Has anyone ever tried this?

Doc, Rumple has been trying to master one of these programs for about a year now. He has even had training at the VA Hospital in Waco. So far, not so good. Of course, his being blind now does not help one bit.

I did some research for him before the VA got around to giving him the software and found that there is free software very similar to what he is using now. I can't remember the name right off hand but this software is sponsored by a member of the English Royal family.

So if you want to give it a try, free is better than the hundreds they want for anything else. There is also a voice to program built into windows but it sucks royally.
 
An elderly man I worked with called the sales folk for Dragon. After talking to him, the sales rep told him that it wouldn't work for him, because his voice was too slurred.
 
If you just want to "play" to see what its like, down load Dragon Dictation.

It's a free app I downloaded to my iPhone for doing emails (I freakin HATE virtual keyboards).

It really works SURPRISINGLY well! An amazingly few error corrections (99% of which are caused by ME vs. the app "understanding") and *poof* off goes a multi page email!

I LOVE that program and I'd imagine it's headset/real mic oriented big brother works even better!
 
Dragon

I use Dragon for personal and professional writing (though I am not an author, I routinely write white papers and technical analysis papers). Dragon works surprisingly well for me, with a couple of caveats. I have a clear midwestern speaking voice and I came to Dragon with years of experience dictating for others to transcribe - part of "training the software" is training yourself. Dragon has a cool feature that allows you to simultaneously record as it types - this means that when it seems to make a mistake you can play the actual speech back - and often I discovered the software was accurate, I was the one who made the mistake. I'd recommend giving it a try.
 
When younger I played a lot of cricket as a batter. That resulted in having my top hand on the bat crushed several times by fast balls. Result - premature arthritis and a problem typing.

So, I started using Dragon a couple of years ago for longer pieces and it works pretty well - but you must train it. That is a bit tedious, but worth it in the long run.

It can make you realize how sloppy your regular speech is!

One thing you will notice very quickly, (at least I did) is that your spoken words will be different to your written words.
 
When younger I played a lot of cricket as a batter. That resulted in having my top hand on the bat crushed several times by fast balls. Result - premature arthritis and a problem typing.

So, I started using Dragon a couple of years ago for longer pieces and it works pretty well - but you must train it. That is a bit tedious, but worth it in the long run.

It can make you realize how sloppy your regular speech is!

One thing you will notice very quickly, (at least I did) is that your spoken words will be different to your written words.

arthritis and flat keyboards have me down to two finger typing. One thing I've noticed is that the slow typing has slowed down the voices in my head. The ones that do the stories.

I tried the software I got for Rumply and found the voices sped up as i tried to dictate a story. I had to train the software yes and i had to train myself to slow down. I would go back and reread and see that I had left out sections, even whole sentences at first. I would have sworn I spoke them but they were not there.
 
What concerns me is not whether the thing would work, but whether I could actually write fiction by speaking, because I don't think I could. I think telling a story is radically different from writing one.

I think I use an entirely different part of my brain when I write than when I speak. Seeing the words in print on a page influences what I say and how I say it, and I take much more time constructing a written sentence than I would ever do when speaking. I access a much larger vocabulary, and I pay special attention to the way the words read, and I spend great gobs of time just staring at the page and trying different combinations of things in my head to get just the effect I want.

I don't do any of that when I talk. Not only do I slur and talk in fragments and inject a lot of "uhhhhs" and "errrs" and "you knows", but I just don't craft things right.

Maybe I'm wrong, and maybe when I was facing a mic a blinking cursor waiting for my first pronouncements I'd start speaking in perfect prose, but I really doubt it.
 
What concerns me is not whether the thing would work, but whether I could actually write fiction by speaking, because I don't think I could. I think telling a story is radically different from writing one.

I think I use an entirely different part of my brain when I write than when I speak. Seeing the words in print on a page influences what I say and how I say it, and I take much more time constructing a written sentence than I would ever do when speaking. I access a much larger vocabulary, and I pay special attention to the way the words read, and I spend great gobs of time just staring at the page and trying different combinations of things in my head to get just the effect I want.

I don't do any of that when I talk. Not only do I slur and talk in fragments and inject a lot of "uhhhhs" and "errrs" and "you knows", but I just don't craft things right.

Maybe I'm wrong, and maybe when I was facing a mic a blinking cursor waiting for my first pronouncements I'd start speaking in perfect prose, but I really doubt it.

this is my major hang up with voice-to-text. hearing a story and reading one are entirely different things, as is telling or writing one. from an artistic standpoint, and im fairly certain from a neurological standpoint you just dont think the same way talking to the screen as you do clicking little buttons to affect the screen, or writing with an inkpot and quill for that matter. unless you have no hands or are blind, your writing is going to be changed by using dragon or whatever, whether you realize it or not.
 
What concerns me is not whether the thing would work, but whether I could actually write fiction by speaking, because I don't think I could. I think telling a story is radically different from writing one.

I think I use an entirely different part of my brain when I write than when I speak. Seeing the words in print on a page influences what I say and how I say it, and I take much more time constructing a written sentence than I would ever do when speaking. I access a much larger vocabulary, and I pay special attention to the way the words read, and I spend great gobs of time just staring at the page and trying different combinations of things in my head to get just the effect I want.

I don't do any of that when I talk. Not only do I slur and talk in fragments and inject a lot of "uhhhhs" and "errrs" and "you knows", but I just don't craft things right.

Maybe I'm wrong, and maybe when I was facing a mic a blinking cursor waiting for my first pronouncements I'd start speaking in perfect prose, but I really doubt it.

At first you are going to get a lot of short choppy sentences and more narrative than you know what to do with. Personal experience.

Editing it makes for some interesting realizations. My mind is more liner than I thought and those choppy sentences can be connected fairly easy.

I just find it more work than it is worth.

If I couldn't type or see then it might be worth the time and money.
 
What concerns me is not whether the thing would work, but whether I could actually write fiction by speaking, because I don't think I could. I think telling a story is radically different from writing one.

I think I use an entirely different part of my brain when I write than when I speak. Seeing the words in print on a page influences what I say and how I say it, and I take much more time constructing a written sentence than I would ever do when speaking. I access a much larger vocabulary, and I pay special attention to the way the words read, and I spend great gobs of time just staring at the page and trying different combinations of things in my head to get just the effect I want.

I don't do any of that when I talk. Not only do I slur and talk in fragments and inject a lot of "uhhhhs" and "errrs" and "you knows", but I just don't craft things right.

Maybe I'm wrong, and maybe when I was facing a mic a blinking cursor waiting for my first pronouncements I'd start speaking in perfect prose, but I really doubt it.

You are right, particularly in the learning phase, and especially when writing dialogue. When speaking a part, one tends to dramatise it with voice, speaking softer or louder, persuasively or dangerously, for examples, but without the need to explain, because one can hear the differences. But given time I got used to it and it makes little difference now.

PS I forgot to add .You cannot use Dragon where there is any discernible background noise or chance of many interruptions.
 
Last edited:
I couldn't write by dictation. Years ago I tried it with work memos. Just couldn't do it.
 
You are right, particularly in the learning phase, and especially when writing dialogue. When speaking a part, one tends to dramatise it with voice, speaking softer or louder, persuasively or dangerously, for examples, but without the need to explain, because one can hear the differences. But given time I got used to it and it makes little difference now.

Yeah, that's what I think. I think that, assuming I had the patience to stick with it for some months, I'd eventually learn to speak something like prose, or at least something that was less like speech, but I don't think it would ever be anything like the prose I write.

Here's my goofy and useless aesthetic thought for the day: That prose is like perfected speech. Writing gives us the chance to sculpt and idealize speech into something beautiful. Just like classical painting gives an artist the ability to capture and idealize something visual. So would that mean that voice-to-text vs. writing would be analogous to photography vs. painting?
 
Question: Wouldn't VtT make for MORE realistic dialog? I often find a lot of people's dialog to be TOO perfect to be realistic and it makes me slip out of the story a bit.

Thoughts?
 
What do you mean by dialogue that is too perfect? Do you mean too formal?

If the dialogue doesn't agree with how a character or characters are developed up to that point in the story, then that is a fail but it has more to do with the lack of skill of the writer than whether it is a voice to text transcription or not.
 
Back
Top