Could you finish your book or thesis without actually writing anything? Probably not, unless you can afford to hire a scribe. It’s possible, though, to reduce the amount of time you spend typing and slumped at your desk. I’ve just run Thesis Boot Camp at the University of Sussex, where we discussed the problem of writing when we’re tired. After a long day at work, it can feel like a gargantuan effort to get into the chair and start clattering at the keyboard. And trying to extract three-dimensional ideas from our heads and into linear words is tough, even at the best of times. Wouldn’t it be nice if we could just talk about our research, then it would pop up on the screen? Well, it could.
In this post, I’ll discuss a few speech recognition tools for making your writing process a little easier.
1. Dragon Dictation
Who wouldn’t want a friendly dragon that takes dictation? Dragon is probably the most popular speech recognition software, and it’s also the best. It works with applications such as MS Word by transcribing your spoken word into text. While there are many other tools that perform the same job, Dragon is the most accurate. It learns the words and phrases you use the most to minimise corrections, and is also very clever at dealing with accents.
I’ve just bought Dragon and am planning to spend some time learning it over the Christmas break. Like any sophisticated software, it takes a while to get the most out of it. You have to practice speaking clearly, and also familiarise Dragon with your vocabulary and preferences. As you can see from this tweet, my first attempt wasn’t entirely successful:
“Currently training Dragon Dictate to transcribe my speech. The first sentence has come through as “Musical chickens waxing remarkable priests.” Can’t remember exactly what I said, but it wasn’t *that*.”
There are lots of different versions that all work in slightly different ways. If you’re serious about investing in a Dragon, I’d heartily recommend starting with Scott Baker’s book, The Writer’s Guide to Training Your Dragon. With his guidance, users can achieve up to 5,000 words an hour and 99% accuracy.
The software costs from £79.99-£279.99, depending on the features you need (Scott’s book can help you decide). It’s definitely not a cheap option, but it could transform the way you write. I certainly feel as though it’s going to make a big difference for me. There are generous educational discounts available, too, if you have an academic email address.
If that’s still too pricey, you could try the ‘type with your voice’ feature in Google Docs. The accuracy is much lower, but it’s free.
VoiceBase’s web app offers speech recognition software to users. You upload your audio file, then receive an email notification once the machine-based transcription is complete. They’ll transcribe the first 50 hours for free, then it’s charged at 2 cents per minute. You can opt for more accurate (but pricier) human transcription.
You could use VoiceBase for your writing, and also to transcribe interviews. I haven’t used it myself, but it was recommended by Tunde, who attended one of my workshops. With 50 hours free, it’s definitely worth a try.
I can’t imagine life without Evernote. Not only does it free up space in my beleaguered brain, but it also provides timesaving features such as speech-to-dictate. It works slightly differently, depending on whether you have an Android, iPhone, or desktop computer. I’ve been using it on my Android phone for a few years now, and it recognises my quiet London accent with unerring accuracy.
While it’s not really a word processor, you could dictate quite a long document with Evernote, then transfer it to another program. Or just use it for capturing your ideas before they dribble away.
Currently my favourite phone app, Braintoss captures your thoughts as images, text, or voice, then pops them straight in your inbox. The audio feature allows you to record a voice memo (saved as an MP3 file) and also transcribes it into text. Now, the transcription doesn’t work very well for me (as you can see from this screenshot), but the other features are immensely helpful. I can quickly type a note, take a photo, or talk to myself – then it comes through as an email.
Nothing is actually stored in Braintoss, so there’s no danger of it becoming yet another repository. All your stuff ends up where you’re likely to see it.
You wouldn’t want to use Braintoss for sustained writing, but it’s great for those of us who struggle to keep an idea in our heads for longer than 2 seconds. It works on both Android and iPhone.
Hopefully, that’s given you a few pointers on speech recognition. This is a technology I’m keen to pursue in 2018. One of my objectives for the new year is to do a lot more writing without completely wrecking my back. Can I write a book just by talking to myself? I’ll let you know.