I’ve been an iPhone user for about 2 1/2 years now and I have to finally say I’m a convert. One experience however that I don’t enjoy on the iPhone is typing. I don’t find that the capacitive keyboard is near as good as the mechanical counterpart. Nuance came out with the Dragon app that ran on my 3GS. it was a decent experience where you could dictate into the iPhone. While you were speaking the app would encode your voice and send it to the server for transcription and then send the text back. The app would give you a minute or two and then you’d have to click it again for more time. The accuracy was decent on it. It was much better at translation than I was at typing as long as there was not too much ambient noise.
The iPhone 4S added Siri which effectively is the same technology as the Dragon app but integrated into the OS. Siri allows you to dictate text messages, emails, and drive other tasks by voice. For me, the transcription is one of the most used features for me on the phone! Often when I’m text messaging it’s much easier to just dictate rather than type.
Since both of these experiences really are Dragon by Nuance, I wanted to see how the PC version worked. One of my pet peeves with Siri in particular is that it often fails for no apparent reason. You’ll be rattling along and then the cursor shakes indicating that it lost the message it was trying to capture. The application is server based, so any blips in network connectivity or the server being overloaded makes your experience less than great.
Dragon Naturally Speaking does not require internet connectivity They were having a sale on the premium version so I wanted to give it a shot. I also added a USB microphone to the bundle as I’ve heard that the mic can make or break the experience. I had tried the software many years ago and training was a never ending experience and the accuracy was not that good. With version 11.5, training took about 20 minutes. It was really only a few things:
- Choosing the microphone
- Choosing your accent (I chose Southerner)
- Setting the microphone volume (you read two paragraphs)
- Calibrating the voice engine (you read 10 or so paragraphs)
After that you are off to the races. The total experience was about 30 minutes! I’d had pretty good experience working with Siri so adapting was very simple. If I wanted to type out this note:
Hi Mike, It was great seeing you and Angela today. The dinner was awesome! How was the concert in Central Park yesterday? Regards, Dan
You’d have to say:
Hi Mike [comma][new paragraph] It was great seeing you and Angela today[period] The dinner was awesome[exclamation point] How was the concert in [cap]Central [cap]Park yesterday[new paragraph] [cap]Regards[comma][new paragraph] Dan
The grammar engine sometimes will add in punctuation that you leave out. I find that auto punctuation puts too much punctuation in often the wrong places, so I just add it in myself. Much like the iPhone, working in a quiet area seems to make the software much more accurate. Adding words to its vocabulary is simple enough, but getting it to auto correct isn’t a smooth experience.
All in all, I do think it’s worth the $180 for the software and premium microphone. As I use it the accuracy will get better with time. For an out of box experience, the software is pretty darn good!
Leave a Reply