Hello, I have a copy of the dragon professional 11 sdk server edition at the moment and I'm wondering if its possible to load multiple speakers at all?  Even with separate instances it seems to only allow one speaker to be loaded at a time

 

 

Thanks! 

John,

Whether you're using a standard version of Dragon NaturallySpeaking or the SDK, you can only load one user profile at a time. More simply put, DNS can only use one at a time. You can switch between user profiles without reloading Dragon, but you can't use multiple profiles simultaneously.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

Creative genius is the ability to understand the elegance and reliability of simplicity. - Unknown

-------------------------

Thanks Chuck, looks like it will have to be multiple vm's then to do the job

John,

If you're doing some sort of back-end transcription of multiple speakers with the Server SDK by the time you get to 2 VMs and the main system all running DNS your system will likely have serious performance issues particularly if all 3 instances of DNS are transcribing simultaneously. Having said that I don't know what type of hardware you are using and you may use multiple servers.

Lindsay

You might find these two attachments useful. There is also a system here in the UK where they transcribe multiple speakers. Let me get back to you on that one.




-------------------------

www.pcbyvoice.com
www.pcbyvoice.co.uk

Well I assumed you could transcribe multiple wavs at a time after I read

'If you are interested in transcribing multiple files at one time, learn more about the advanced capabilities of Dragon NaturallySpeaking Professional'

in one of the pdf's that Lindsay linked.  If I run multiple instances of the wavindexer sample and set them all going, only the first one will transcribe the wav, so how do you transcribe multiple files at a time? 

John,

Have you looked at the Auto Transcribe Folder Agent Re: doing batch transcriptions?

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

"People create their own success by learning what they need to learn and then by practicing it until they become proficient at it." - Brian Tracy

-------------------------

Yes I've looked at the autotranscribe agent but it seems to only process 1 task at a time

John,

You might try going back and reading carefully again because this is the only way in DNS that you can do batch transcription processes. You drag-and-drop whatever files you want to transcribe into the specified folder and DNS automatically transcribed them depending upon your settings and configuration, as well as according to which user profiles are associated with such. I confess that I haven't used this in quite some time because I have no need for it, but this is what it was designed for.

In the past, I could do multiple audio file transcriptions by just dragging and dropping those files into the designated folder. DNS would then transcribe each one automatically, one at a time and put the results where ever you want them including the dra files if you need them or want them.. It's a little tricky as you should have noted in the Dragon Help, but it's the only way that you can do batch transcription. Nevertheless, it works if set up and executed properly.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

"What you are aware of you are in control of; what you are not aware of is in control of you." - Anthony de Mello

-------------------------

Here is a quick down and dirty on setting up the Auto Transcribe Folder Agent if anyone is interested. Note that this feature is only available in the professional versions of NaturallySpeaking:

1. Close NaturallySpeaking (Yes I know that is just bullocks).

2. Click Windows Start/All Programs/Dragon NaturallySpeaking 11.x/Dragon NaturallySpeaking Tools/AutoTranscribe Folder Agent.

3. Create 2 desktop folders. Name one folder In and the other Out. If you are British, you can stop laughing any time.

4. Click New under the Dragon AutoTranscribe Folder Agent Task menu.

5. Click the little square box, with the 3 dots, to the right of the input directory and browse to the desktop In folder.

6. Repeat the process for the Output directory and choose the Out folder.

7. Chances are that Dragon will pre-populate the remaining fields with the appropriate information but you can change any settings via the pull down menus.

8. Close the Dragon AutoTranscribe Folder Agent to minimize it to the Taskbar.

If you are using your new Olympus DS-7000 Dictation Module application (shameless plug) you can select all or as few recordings as you like and drag-&-drop them into the In folder. Dragon will transcribe each recording, one at a time, into the Out folder and even move the original recordings to the Out folder.

Note that you do not have to use the Olympus application. Most other recorder apps work as well or you can optionally drag-&-drop recordings directly. Note that you don’t need to name your folders In and Out and you don’t need to put them on your desktop. We simply used this approach as our example.

You also have the option of generating embedded DRA files. You will find this feature in the Options which are located in the Agent menu.

-------------------------

Click KB 2012 REV D to Download a 30 Day Evaluation of KnowBrainer 2012 

 

 

 

John,

In addition to what Chuck mentioned above.

I don't have the server SDK but with the SDK Client edition:

Click Start > All Programs > Dragon SDK Client Edition 11 > Dragon SDK Client Tools > AutoTranscribe Folder Agent

From the help file:

"You set up a task that associates a User Profile with an input and output directory. Since each input and output directory pair can be associated with its own User Profile, you can set up multiple tasks to accept recordings from different devices".

From the help file | Search tab look at the topics:

"Starting the AutoTranscribe Folder Agent"
"Overview of the AutoTranscribe Folder Agent"
and probably most importantly "Creating and deleting AutoTranscribe Folder Agent Tasks"

If you are creating your own application to do this using SDK components do you not have a support contract?

Basically you are supposed to be able to set up multiple tasks and you can enable/disable individual tasks via the UI. I have never tried multiple task concurrent transcription, but try looking at disabling and enabling tasks in the "Creating and deleting AutoTranscribe Fallback Agent Tasks".

Lindsay

-------------------------

www.pcbyvoice.com
www.pcbyvoice.co.uk

John,

I assume that you were talking about a single system with one copy of Dragon NaturallySpeaking installed, or using the SDK.

Obviously Lindsay 's approach will work because that's what Nuance does with their smartphone apps. That is, they use a backend server to transcribe dictation and, depending upon the number of users attempting to dictate on their smart phones as well as broadband considerations, the Nuance backend server can handle many speakers, as well as being speaker independent, simultaneously.

Nevertheless, what Lindsay as described is being done Re: court reporters, both in the courtroom and in terms of taking depositions (read the first PDF document that Lindsay has attached to his post). I interpreted your approach in terms of single systems performing standard dictation. Sometimes the concept of transcription gets lost in the lack of detail. Same function, different methods.

Chuck Runquist
Technical Project Manager
VoiceTeach LLC
Home of VoicePower®: We don't make Dragon NaturallySpeaking, We make it better!

"In theory, there is no difference between theory and practice. But in practice, there is." - Yogi Berra

-------------------------

Thanks guys, great info so far.  To give you a bit more info, yes its a back end transcription service, where I'll be taking pre-recorded wav files and transcribing them to text.  This wont be in real-time but will have to process pretty fast to keep up.  It will probably be run in a hyper-v environment with the multiple vm's doing the work.  I can see that I can transcribe multiple files for the same speaker at the same time so I'm going to have to test this out and see what performance is like

 Cheers,

 John