Pocket PC Voice Recorder

[Jason Dunn]

<div class='os_post_top_link'><a href='http://www.pocketdictation.com/' target='_blank'>http://www.pocketdictation.com/</a><br /><br /></div><img src="http://www.pocketpcthoughts.com/images/web/RLH&Arrow267x174.gif" /> <br /><br />"A Pocket PC dictation program you can use with one hand and either hand at that! PDictA is geared for the professional for whom dictation is an every day process. RECORD, PAUSE, REWIND, PLAY actions form the repetitive “process flow” of dictation. All these operations can be done with the thumb of whichever hand holds the Pocket PC. PDictA is ideally suited for your dictation needs no matter what your profession for all manner of transcription and translation voice capture."

[SofaTater]

Interesting app -- but $30??? Seems kind of steep for an app of this nature...

[nirav28]

Whats the audio format it stores the dictation as? Mp3, wav? low bit rate mp3 or ogg would be nice. The cpu can easily handle 10-56 kb/s compression. Wav files are way too big even if you have lot of flash storage.

[mlougee]

It would be important to know the quality of the recording, to guess (and experiment) whether the recording from this tool could serve as the *input* to something like Dragon Naturally Speaking's voice recognition.

[nz0eBoy]

Looking at the linked website it appears there are trial versions of the program. I am sure this will answer these questions.

[topps]

It would be important to know the quality of the recording, to guess (and experiment) whether the recording from this tool could serve as the *input* to something like Dragon Naturally Speaking's voice recognition.

Yes, I would be interested to know this as well. I suspect that it will largely depend on

1. How well you use the microphone? (Breath noises, etc)
2. How much ambient noise there is?
3. What sound file format you choose (how compressed etc)

I saw that the latest Dragon Dictate is supposed to accept sound files from PDAs. (Did not read much detail about this)

[topps]

I have been trying ProTone from Pocco - very similar. They also just announced similar features such as Pause/Resume during recording (with true continuation, not a second sound file, like the standard app does.)

They have just added Insert, and Append to a sound file.

ProTone also features Voice Activated recording (I'm not sure if this app has Voice Activated recording - did not see it mentioned on the web site.)

Seems to work fairly well...sound quality being determined by the factors in my post above.

[gkrucik]

The format for PDictA is standard .WAV -- the codec used to record the audio is GSM 6.10. With that codec when you record for 60 minutes, you will use up about 6MB of storage. The reason we chose that codec and the .WAV container was the universality for playback and the relatively high compression for voice quality audio that GSM provides. The GSM codec has been around for a long time and shipped with the original version of Windows and also works on Linux and MAC environments.



Whats the audio format it stores the dictation as? Mp3, wav? low bit rate mp3 or ogg would be nice. The cpu can easily handle 10-56 kb/s compression. Wav files are way too big even if you have lot of flash storage.

[gkrucik]

The quality of the recording using the GSM 6.10 codec (which is what PDictA uses) would work for speech recognition purposes -- but your recognition rates would not be optimal. We have not yet released publicly a version for speech recognition purposes - as the resultant files would be several times larger. We are continuing to work on the trade-offs involved with this issue.

It would be important to know the quality of the recording, to guess (and experiment) whether the recording from this tool could serve as the *input* to something like Dragon Naturally Speaking's voice recognition.

[gkrucik]

Trial versions of PDictA are available. The only restriction in the Trial version is that the length of the recording is restricted to one minute. Purchasing the product removes that restriction.

Looking at the linked website it appears there are trial versions of the program. I am sure this will answer these questions.

[gkrucik]

The ProTone product is a very good product. We are happy to see healthy competition in this market -- as it spurs innovation :-)

PDictA does not currently support Voice Activated recording.

We chose not to implement that feature for several reasons. From a technical perspective - "white space" (silence) is not a big deal for space reasons due to the effective compression provided by GSM (file size is 6MB per hour).

Unfortunately, Voice Activation is compromised by ambient environmental noise, and there is a "listening" penalty on the CPU of the Pocket PC as it waits to hear something and turn itself on ("they also serve who only stand and wait").

With all that said, one of the biggest issues is clipping of the opening phrase, due to start up latency.

Hope that answers your question
Best regards,
George.

I have been trying ProTone from Pocco - very similar. They also just announced similar features such as Pause/Resume during recording (with true continuation, not a second sound file, like the standard app does.)

They have just added Insert, and Append to a sound file.

ProTone also features Voice Activated recording (I'm not sure if this app has Voice Activated recording - did not see it mentioned on the web site.)

Seems to work fairly well...sound quality being determined by the factors in my post above.