LabVIEW 2018 crash with Parks-McClellan vi

[vix]

I tried to design a differentiator filter with Parks-McClellan vi.

I tried to set an odd number of taps (63), so in the last band I set Higher Freq < fs/2 (0.49).

It can be that a filter can't be designed with these parameters (I don't know), but the problem is that LabVIEW 2018 crashes.

The problem can be reproduced running the attached vi (all the parameters have been already set).

Could someone investigate, please?

[Bob_Schor]

Yes, it crashed on me, as well.  I ran it twice -- the first time, it just "disappeared".  When I restarted LV 2018 (32-bit), it "recovered" the files.  The second time, I ran with Highlight Execution on to see where it crashed.  It seemed to have just entered the PM routine.  I sent the Crash Report, along with your VI, off to NI (as requested by NI).

Bob Schor

[vix]

Hi Bob,

thank you very much.

On my PC the crash doesn't open the LabVIEW crash reporter.

I hope NI can fix this as soon as possible.

[DSPmchen]

Hi Vix,

We find that there is a bug in the Parks-McClellan code.  If the higher frequency in the first band is close to 0, the crash will happen. We will fix this in future release. 

For workaround at present, you might increase the higher frequency in first band a little bit. For example, make the higher frequency in first band as 0.002, and the lower frequency in second band to 0.0025.  This can avoid the crash.

Best Regards,

Michael