1D Array of Waveform

[MeCoOp]

Well it does not look like my initial solution was that far off, I've posted a copy of the block diagram.  Unfortunately both solutions return inaccurate results, so I'm assuming there is still something amiss in the averaging portion.  Thank you to everyone so far for the suggestions.

[Dennis_Knutson]

Your intial solution was not far off from what? Your are only sending a single point to the analysis routine. That's not going to work at all. When you say both solutions, what 'both solutions' are you talking about?

[MeCoOp]

I mentioned earlier that I realized without some sort of buffer for the data my averaging routine would not work and that some sort of running average maybe better suited for the application.  When I commented on the two solutions, at least to the untrained eye, it looks as though what I did and what you proposed were not that different.  Maybe I'm not stating my fundamental problem clear enough, all I want to do is read from both microphones and somehow "average" the multiple channels so that when it is reported through the octave analysis it looks like one signal.  Thanks.

[Ben]

Are you you doing single point measurments on each channel?

If so, the two values will be skewed in time relative to each other, e.g. averging sample #100 from channel A with sample #117 from channel B.

Ben 

[MeCoOp]

My DAQmx Timing.vi is setup to do continuous sampling.  I am using a PXI-4462 to for the data aquisition so each sample should be simultaneous.

[Dennis_Knutson]

There is considerable difference between the way you are averaging and my technique. You take the mean of all samples from ch1, take the mean of all samples from ch2, then calculate the mean from those two results. You end up with a single point. With my technique, I average the reading from time x on channel 1 with the reading at time x on channel 2. This is done for each acquired point from t=0 to the last point in the acquired wavefrom. The result is an entire waveform. If, for example, channel 1 had a pure sine wave and channel 2 had a pure sine wave of the same amplitude and frequency, the result of my calculation would be a sine wave. Look at the attached VI.

[MeCoOp]

First off let me say thank you for taking the time out to help me through this.  I understand the logic behind you last example, what I don't understand is why it won't work when I try to make it work for my scenario.  What I did was replace your generated signal with the aquired signals from my daq device.  Nothing is displayed on the octave or waveform graph.  I get "-Inf" for a band power reading.  Did I oversimplify the application?  I should mention that when read separately both channels display information correctly.

Message Edited by MeCoOp on 08-21-2007 02:10 PM

[Dennis_Knutson]

Can you post the modified VI with some actual data? The easiest way to do this is to create an indicator for each of the DAQmx outputs. Run the VI once so that there is data present. Stop the VI and then go to Edit>Make Current Values Default. Save the VI and post this. With some real data, someone can run it without having any actual hardware.

[MeCoOp]

I can't save any data because the programs are written on my laptop and then the executable is loaded onto the PXI chasis.  I did however find that your solution was not working because of the way I was reading from each channel.  Now I'm able to obtain results but I'm not sure if they are correct.  When each microphone is read independtly with the calibrated sound source my total band power is in the 88 dBA range.  The total band power read for the averaged microphones is in the neighborhood of 77 dBA.  I've attached a copy of my program.

[Dennis_Knutson]

I don't know if that's reasonable or not. I didn't pick up on the fact you were using the SVT functions. I don't have this toolkit and have never used it. Averaging will reduce the amplitude of spikes you see on one channel but not the other but a reduction of 11 dB seems a lot. I hope someone with more experience with sound and vibration can chime in.