 |
PM |  | |
Quote | | |
Reply | | |
Full Topic | |
| Joined: | Mar 7th, 2007 |
| Location: | USA |
| Posts: | 206 |
| Status: |
Offline
|
I'd really like to know what this program does behind the scenes, I think we can find some other applications for it, I'm sure!Hi Keith,
In Stream chatter clip I hear the following:
"What's happened today?...(unknown) ..learn to be patient.....would it.....(foreign language)....watch I'm gonna slide my head....into the picture."
Keith
From the few communication with the original EVPamp developer back in 2005 a few fragments of information are available. "The filter is similar to a Kalman filter or a Hidden Markov Model based filter".
So it is filter based in design and not jumbling, chopping or adding additional info like other programs and 'talkers' out there.
In running many tests the best response has been with using high frequency noise or even video sweeps converted to sound. The current stream here on ITC Bridge works very well. This is now what some may be calling the 'fog face' mode.
Here is an example from yesterday.
"A/the man has an announcement!"
P.S. More info about these types of filter designs http://www.cs.unc.edu/~welch/media/pdf/kalman_intro.pdf
Attachment: 070712manhasanannouncement.mp3 (Downloaded 1013 times)
Last edited on Jul 8th, 2012 11:16 AM by EVPDave