Does anyone know of any published statistics or experience with voice recognition software?

Question: I’ve found VR so helpful in reducing the amount I type (and thus reducing the pains in my arms) that I think it would be a shame if people were put off by the risk of voice problems.  I think that, for a lot of people, VR is worth trying, bearing in mind advice like making sure you drink plenty water; take breaks; have the volume set at a level which means you can talk without raising / straining your voice. If you’re aware of the possibility of problems, you can stop / change your use of VR at the first signs – if they appear…  It would be interesting to know how common problems are.

Does anyone know of any published statistics?  Sorry I can’t say anything about VR & a stutter. Perhaps you could get a demo from somewhere that sells VR software, so that you could see the way in which a user speaks (eg I tend to pause quite a lot and speak in a more calm and measured / focused way than in a conversation with a person. Pausing allows the writing to catch up (maybe my PC is a little slow) and also allows me to partly prepare what I’m going to say, to avoid changing my mind half way through and making all sorts of hesitating ‘ums’ and ‘ers’. I don’t know whether that sort of approach would affect stuttering at all – whether it would help or make it worse.)

Answer 1: I agree — people do get voice strain, but very few get serious voice strain. 

Answer 2: My own experience that is that voice recognition software has some forced to or perhaps encouraged me to change the way that I think about writing. Perhaps my long experience of dealing with written text evolving in front of me, on the screen, has made me accustomed to being able to “sculpt” documents in a very iterative process. VR, as you guys seem to call it, on the other hand, generally requires you to think of each sentence before “reading it out “.Which also means that you have to think about the how that sentence relates to the other sentence as the you are writing.

For me this is quite difficult. I have found it helpful to try and imagine I’m talking to a friend in an informal situation, such as down the pub. I also find it quite difficult to use dictation software if there’s someone else in the room. Perhaps I am secretly embarrassed by my pompous style of writing.  If with the recognition software there’s a consistent delay, rather than just a delay after the first utterance, it probably means the machine is too slow.

Processor speed is certainly a problem (I’ve found PII 200 Mhz = just about OK whereas PIII 500 Mhz = now we’re cooking with gas), but also defragmenting the disk also can make a big difference. Presumably the s/w has to “look up the words you say” so disk performance counts.

Leave a Reply

Your email address will not be published.

Notify me of followup comments via e-mail.