Dragon NaturallySpeaking is possibly one of the most well-known dictation software packages available. The latest incarnation has a whole host of new features including Office 2010 support, iPhone/iPad wireless dictation and the ability to control your PC applications using your voice.
Setup: In terms of the initial setup Dragon NaturallySpeaking 11.5 first requires that you create a user profile and then using the supplied headset microphone (which is pretty well made, with a nice rotary microphone attachment) train the software, not only to recognise your voice in terms of the volume level but also the style that you speak in. Additionally it can scan/analyse your e-mails and documents to learn the way that you write which helps with the overall accuracy.
At this point we need to add that while the software worked perfectly with our on-board sound card, we have to admit to spending considerable time getting this to work with our external SoundBlaster Xi-Fi soundcard and as we write, we still have the odd issue.
To cut a long story short the problem seems to point to a setting that we needed to activate within our Creative Labs console software. While this cured the problem initially, every so often we found that the software would mute the microphone when we were performing particular tasks, such as transcribing from an audio file.
As IT Tech Savvy users, we could easily sort this problem out on the fly, but for non IT users this may prove an issue. In fairness the soundcard we have is a few years old and the driver support is similar, so we will give Dragon the benefit of the doubt here and blame the problem on poor driver support for the Soundcard.
Anyways, before we digressed, the training of your voice does take around about 5 to 10 minutes and involves you reading out aloud from a series of text prompts, but it certainly is worth the wait as the accuracy out-of-the-box is pretty impressive when you're dictating e-mails or documents.
In fact I am writing this review at the moment using the power of my voice and it does capture the same feel as if I was typing the review using a keyboard.
In Use
However, to begin with you will need to speak slower than you normally would otherwise you will find that the software will miss out certain keywords and replace these with it’s own alternatives.
Additionally it can’t recognise words such as Crowdedbrain.co.uk or words that are said in a different context, i.e. weather or whether (it simply transcribes what it thinks) and it was having issues capitalising the first word of a new line, but the accuracy does improve over time because the software continues to learn and adapt as you correct the mistakes.
For example if you find that a word is spelt incorrectly you could use the “correct that” command and choose the correct word from a list or you could use the “spell that” command to spell out (or type) the correct word and then get the software to learn, via it’s Training function, so it does not make the same mistake again.
Mind you commands are activated when you pause before you speak, so for example if you wanted to included the two words "correct that" in your sentance and you did not speak quick enough then you will find the correction dialogue box will appear instead of the words being placed on the page. Thankfully there is an option to adjust the pause speed so this does not happen as frequently.
Dragon Sidebar
As well as dictation you can also use your voice to open applications, control file menus, launch web browsers, format words, move the mouse around, post to Facebook or Twitter and create your own set of commands to automate several tasks at once.
The only snag when controlling these functions is that if you're in normal mode you need to be careful how you pronounce the commands (similar to the problem we highlighted earlier) otherwise it will start to launch software that you don't necessarily want it to. However you could simply turn off the normal mode off and concentrate solely on dictation to resolve this issue and you’ll find with practice the whole process does get better over time.
Personally I found the formatting (such as selecting text) and manipulating the web browser content a little bit more challenging, however to help you in the early stages of learning the software you do get the Dragon Sidebar which intelligently switches to a list of helpful functions that reflect the task you are doing.
So for example when you're in Microsoft Word it will list all the formatting functions that you can perform in a list and vice versa if you switch to your web browser you get a different set of commands. You can have the sidebar locked to the right side of the page or floating. Eventually when you have learned the commands you can disable it. Either way it's a great touch that I certainly appreciated when using the software in the opening weeks of use.
Audio Transcribing
As well as dictating, the software can also use the transcribing function to listen to an audio track - of your voice - and attempt to convert the spoken words into text.
Unfortunately the results of our earlier recordings were pretty poor and it failed to transcribe our MP3 recordings properly. However the reason for this could be due to the way we recorded our voice in the first place. For example, you need to record in a similar way that you would speak whilst dictating, i.e. slower, and also use the commands such as “New line” to move down a line.
Without the above control you will find nothing but garbled text, so this again requires some practice to perfect.
In terms of supported recording formats, the software will open .mp3, .wav, ,wma, .dss, .ds2. You can also transcribe to a particular window or get the software to transcribe using your favourite word processing package.
Text-to-speech
You can also highlight text and get the software to read this back to you, so for those with impaired vision you would find this a useful function to ensure that what you have written is correct. Of course with any text-to-speech software it still struggles to pronounce certain words, such as YouTube.
Once again though everything was not perfect as we found that software's voice command function for reading back the text “Read that” was not that responsive and we had to rely on using the mouse to select the function from the menu.
Wireless Dictation
If you can't get on with the supplied headset you can turn to your smartphone and use this as a wireless microphone. In order to do this you need to download their free IOS app – which is compatible with the iPod touch, iPad and iPhone.
In practice once you have your iOS device and your computer is connected over the wireless network all you need to do next is to pair them together and away you go.
The next stage then involves a similar process to the wired headset so you need to go through a series of training exercises in order to get the wireless microphone to work with your profile.
Accuracy is similar to the headset, but we found you need to make sure that when dictating the distance that you initially had between your mouth and the iPhone/iPad is the same as when you originally recorded your voice in training, otherwise it can fail to recognise commands properly.
At least the software gives you the choice to use either a wired or wireless headset!
Note: If reading the review from a news feed app you will need to visit the site for the full conclusion and score http://www.crowdedbrain.co.uk/productivity/dragon-naturallyspeaking-11-5-premium-review.html.
