Value of Amplitude output in peak detector.vi?

[LindaW]

I want to use Waveform Peak Detector.vi to read analog input waveform data (a square wave signal) and then tell me the amplitude and location of the peaks. Does the "Amplitude" output in peak detector.vi give you the mean over the entire width of the peak, or does it give you the maximum amplitude of the peak?
Effectively, I am trying to use the waveform peak detector.vi to digitize my analog input square wave (of varying amplitudes). I then do a statistical analysis on the mean intensity of the peaks, and some other calculations. If anyone has another suggestion as to how to go about this, please feel free to tell me.
Thanks!

[Jonnie_5]

Why not use the Pulse Measurements vi?
LV 6->Funcitons->Analyze->Waveform Measuremnts
This vi:
"Accepts a periodic waveform or an array of periodic waveforms and returns the period, pulse duration (pulse width), duty cycle (duty factor), and pulse center of a selected pulse and period."

[LindaW]

Thanks for the suggestion.

My application entails a laser beam that propagates through the atmosphere. So, at times, there may be a fade in the periodic signal, and I'm afraid that this would mess up the pulse measurements vi. Although I suppose it will mess up the peak detector too...

[Jonnie_5]

There are more complex ways of handling this type of data. You could do a running average (sample sets of 10 pulses and Q-test outliers).
You could write a more robust detector, essentially your own threshold detector, which would analyze each pulse individually as well as the entire set. Or you could try some sort of filtering algorithm.
G-Luck

[Gleb]

In my opinion the Peak Detector performes quadratic least square fit around maxima, and the amplititude values are the results of the fit process ("width" deteremines the interval for this fit).
The reported amplitude values are definetely neither absolute peaks nor means within "width" around the maxima.
You can check how the peak detector works if you generate your own test function using SIN and then passing it through the Peak Detector

[Jonnie_5]

For a wee bit more about that vi see:
http://zone.ni.com/devzone/conceptd.nsf/webmain/55181A206A523FD186256ABE00626EA9?opendocument