Re: manipulate audio buffer - capture stereo mix on ce .net device

Tech-Archive recommends: Repair Windows Errors & Optimize Windows Performance

On 24 Jan 2006 13:23:55 -0800, pommes wrote:

> Hi,
>
> I'm trying to build an windows ce .net application for pocket pc (ce
> .net 4.2) to test an Speech Recognition Software(Dragon Naturally
> Speaking)..
> All I want to do is to send an audio stream to the speech recognition
> application which should be recognized.
> In other words, I need to capture the audio output (stereo mix pin) and
> not mic input.
> However, I want to do the same like the Speech SAPI Sample "Using WAV
> File Input with SR Engines", but don't have any access to the
> speech recognition software source code.
> I really studied the sample "Recording and Playing Sound with the
> Waveform Audio Interface" provided by microsoft. I tried to
> manipulate the audio buffers before they are send to the speech
> recognition application. Could someone give me a hint to the right way
> of doing this. In fact, is it possible to,
>
> 1. modify the Waveform Api in order to manipulate the audio buffers
> before they are passed to the speech recognition software?
> 2. Is there any Application like "Virtual Audio Cable" or
> "Vsound" for Ce .net
> 3. or do I have to write a streaming interface driver (using eVC++ or
> Platform Builder?)
> 4. Is it possible to port "ActiveX Capture" to ce. Net?
>
> How do I get these audio driver samples mentioned in MSDN. Do I need
> Platform Builder 4.2 or the Ce DDK?

I'm not quiet sure I understand why it needs to be so complicated. Using
SAPI you can simply tell the SR engine to listen to the wave capture source
instead of the microphone.

Mind you I don't know much about CE so I'm not sure what version of SAPI is
available.
.

Relevant Pages

  • Re: [Full-disclosure] Vista Speech recognition
    ... Speech recognition is inherently unreliable, ... audio 5 minutes later. ... (Remember - the average office is nice and quiet at 11PM if the janitors ... aren't around - and nobody ever *said* the computer making the noise was ...
    (Full-Disclosure)
  • manipulate audio buffer - capture stereo mix on ce .net device
    ... All I want to do is to send an audio stream to the speech recognition ... speech recognition software source code. ... Platform Builder 4.2 or the Ce DDK? ...
    (microsoft.public.windowsce.app.development)
  • speech recognition test tool -- manipulate audio buffer ??
    ... All I want to do is to send an audio stream to the speech recognition ... speech recognition software source code. ... Platform Builder 4.2 or the Ce DDK? ...
    (comp.speech.research)
  • manipulate audio buffer - capture stereo mix on ce .net device
    ... All I want to do is to send an audio stream to the speech recognition ... speech recognition software source code. ... Platform Builder 4.2 or the Ce DDK? ...
    (microsoft.public.win32.programmer.mmedia)
  • Re: [PHP] Audio CAPTCHA review request
    ... Richard Lynch wrote: ... If Corporate America uses speech recognition on national 1-800 numbers ... are the odds that Tedd's audio samples, even muddied up a bit, are ... I read an article about Japan's speech recognition as being very successful because of the purity of their language's phonetics and used English as an example of being problematic. ...
    (php.general)