Community Forums Archive

Go Back

Subject:Any way to automate analysis of volume peaks/spikes?
Posted by: twyrick
Date:4/27/2001 10:40:28 AM

I have a rather specialized need, and I'm not sure if Sound
Forge (or perhaps another product someone could recommend)
will do the trick. Any help would be greatly appreciated!

I work for a metal heat-treating and finishing company.
One of the parts we heat-treat for a customer has to have
random samples quality-tested afterwards. This test
consists of dunking the part in a mild acid bath for 7 or 8
minutes. If it's treated properly, nothing happens to it.
If there's a flaw, though, the part will loudly crack into
two or more pieces. (If it splits into multiple pieces, it
makes a louder crack than when it just breaks into two.)

Right now, someone has to take out the time to babysit
these parts as they're tested, to note what happens. This
gets rather inefficient. We came up with the idea of
putting this acid bath inside a relatively sound-proofed
container, and inserting a microphone attached to a PC with
sound card.

What we'd like is a way to automate a program so someone
could click an option on the PC to begin recording, as soon
as they insert a part into the bath. Then, the PC could
record a window of time (say, 20 minutes) to a .WAV file.
When that gets done, it would need to go back through the
resultant .WAV file and strip out all of the silence,
except for a second or two before or after a volume peak
(or rapid succession of volume peaks), and store the final
result on the PC's hard drive.

This way, the user could listen to the .WAV file at their
convenience to hear what happened to the part. (A non-
existant .WAV file would mean it recorded nothing but
silence, hence the part was good.)

Subject:RE: Any way to automate analysis of volume peaks/spikes?
Reply by: 10strip
Date:4/27/2001 1:33:18 PM

You might want to look into a piece of software called
LabView by National Instruments. It is used for data
acquisition.

Go Back