Smoothing question

Is your noisy data displayed y data wave vs x data wave?

Or y data wave vs the y wave's x scaling?

An example experiment attached to a reply would be useful in diagnosing what's wrong.

Example uploaded.  I have original data and the smoothed plot shown.  (I know it doesn't look noisy, but we are doing various other analyses, including differentiation to get a velocity plot.)  Maybe I'm just not choosing the correct options when smoothing.

Not quite sure what you are asking about the waves.

Your "Original" data is displayed in one graph as an XY trace:

even though your time increment is linear (0.0002 seconds per point).

Your "Smoothed" data is displayed in another graph as a waveform trace:

A good explanation of the difference can be found in the help by entering this command in Igor's command line:

Also, look in Igor's Help menu for "Getting Started".

As to Smoothing, it depends on your goal.

A good way to set the goal is to examine your signal for what you consider "noise". Examining the FFT magnitude (after properly setting the signal's "Wave Scaling") shows peaks in a few places (see FFT Magnitude Peaks.png, attached).

The peak near 2200 Hz is so close to the Nyquist frequency that 1 pass of the binomial smoothing eliminated it. See "FFTAfterSmooth.png".

The other noise is also attenuated, gradually less attenuation the lower the frequency.

To get an idea of how much smoothing the Smooth command does, see the "Smooth Operation Responses" example experiment in Igor's File->Example Experiments->Analysis->Smooth Operation Responses menu item.

Thanks!  That helped.