On "The Golden Rule" and Purge Cache.
I do break the Golden Rule and mix dictation with typing.
I find I can do this fine, provided I use the Purge Cache command and understand what the issue is behind the "golden rule".
The key is, dictate all you like, correct by keyboard all you like, just do not go back and edit using voice commands until you have refreshed with a "purge cache" command so that Dictate has a refresh of all text in its memory, including what you have entered manually.
That is why I love the "purge cache" command.
I do this in Word, Textedit, MacJournal. All fine my end.
This approach involves a certain style of use of Dictate. One gets into a certain style or mode of use, which takes a little time. So at first this approach may not work for others until you develop the consistent patterns of habit with this approach. Frequent (or perhaps "timely" is the better word) issueing of the verbal commmand "purge cache" and only at the completion of a sentence and full stop. (Reason: Purge cache command will assume the next spoken word is the beginning of a new sentence and hence will capitalise the first letter of the first word eg Word.)
This approach works well for me.
Also I find I can have numbers in numerals or words, with practice of the right word commands.
I admit it takes time to get this down in your head, so that you come to use them without thinking. Again, it comes down to developing a style of use. So on this issue, I had to go back and read the manual on the details of these caps and numeral commands and how they worked.
iListen to Dictate (with the first change from Phillips speach engine to DNS) was a critical improvement in speech recognition making Dictate useful as a productivity tool.
I am interested to see if there is any further reports of significant improved accuracy with Dictate 2.0. If there is another distinct leap in performance in recognition accuracy as one poster in this section has offered, I may well do so, though it is a pain it only works on OSX 10.6.2 upwards. That means another OS X upgrade, and the time to sort out some incompatibilities of old but good'un, much loved and used software. (I love mathpad).
Which brings me to another question: Why is that??? 10.6.2 upwards only. This leads me to wonder if Dictate 2.0 is really a gain in better speech engine performance, or just more efficient processing on full 64 bit with OSX 10.6???
Blast from the past:
iListen was not useful as a productivity tool - oh how Macspeech lied to me for years that the problem was my end .. 3 laptop and OS upgrades, mics .... in the end I entered a heavy phase of objective rigorous testing of iListen with all possible performance affecting variables.
I sent just one graph of the results of 200 plus hours of careful testing of iListen performance to show it was the limitations of the program and phillips speech engine. I suggested to Macspeech if they did not change the speech engine they would fail as a company as it was never going to work well enough to be broadly taken up as a productivity tool - and hey presto - 6 months later and they changed the speech engine to DNS - yahoo for that!
You would think with that amount of dedicated effort, and cost my end, Macspeech would offer me a free upgrade, but no .. They just ceased to respond to any of my emails after I sent then my results. Silence. I was the messenger not the enemy!
Anyway, Dictate is better than iListen, and lets see with Dictate 2.0. If recognition performance is better, I can work with the rest of the set up. I just want it to work re recognition accuracy, so I can speak freely normally with out having to watch for errors, and it all comes out in text..... so that it is free thinking into typed text. Its close, but not there yet. This need a consistent recognition accuracy of better than 99.8%.
Thanks for all the other posts.
PS I am mindful of posts that pose as users but are really macspeech promotions. How do you know? You do not. In the past I have followed up some people with direct email exchange to check the independence of comments.