Problem with Savinsky-Golay ptbypt.vi in labVIEW 7

[Ashley.w]

Hi

I am having a problem with the Savinsky-Golay ptbypt vi in labView 7.  It does not  appear to smooth data as expected. 

I attach an example vi I am put together which demonstrates the problem.  The data is reversed in sign, greatly reduced in amplitude and is not smoothed at all. 

I have used this filter in Microcal Origin and found it to work very well on my data.  The LabVIEW vi requests the same input paramers i.e. smooth window nR=(-nL)and ploynomial order.  I can only think there is a bug in the vi.  Has anybody elce encountered this problem?

Many thanks

Ash

[TCPlomp]

Hi Ashley,

you initialize the filter on each data point, just remove the true constant. The VI itself will do the initilization (with a firstcall? node)
Then the data looks good.

Ton

[Ashley.w]

Hi Ton

Removing the true constant does indeed ensure the data is smoothed.

However I am seeing a phase shift proportional to the width of the filter window which moves the position of my peak.  When I run the same S-G filter in Microcal Origin this phase shift does not occur.

Is there something elce I have missed?

Many thanks

Ash

[Ashley.w]

Needed to edit the array to remove phase shift.

(Could oversample to compensate for data  points lost)

See vi attached.

Works nicely.

Ash