Integer-Based FFT in LabVIEW

Updated May 3, 2018

Reported In

Hardware

  • PCI-6115

Software

  • LabVIEW

Operating System

  • Windows

Issue Details

Processing time in LabVIEW seems much slower using the DBL data type for my very data and processor intensive program. I am acquiring data for frequency processing and reshaping the data into numerous rows and many columns. Do you know of a way in LabVIEW to perform an Integer-based FFT (e.g. I16)? 

Solution

As per the LabVIEW Help documentation for the FFT VI, the supported data type is a DBL, and once you get into that VI, you cannot manipulate its expected type. If you pass in an integer, the data type will coerce. By nature, the VI will perform the FFT using floating point values, not the integer. 
 

Figure 1: Real FFT VI inputs and outputs, where orange color denotes numeric data of the DBL type

Additional Information

If your main concern is processing time, it may be valuable to refer to the white papers 6 Tips and Tricks to Increase LabVIEW Performance and Will My LabVIEW Programs Run Faster When I Upgrade to a Multicore Computer?
 

WAS THIS ARTICLE HELPFUL?

Not Helpful