FPGAs, sound algorithms, and Arnold
Austin, TX — “One thing I have been working on is an audio-enhanced PA system, so I took a CompactRIO system, connected an IEPE Measurement Microphone, and connected a channel from an output module so that this microphone gets inserted into the sound system right here in this auditorium,” explained Alain Moriat, who is also a professional sound engineer.
A principal Architect at National Instruments he prepared to demonstrate to the crowd at NIWeek 2005’s keynote address how easy it is to use CompactRIO to develop sound processing algorithms on-the-fly. In this case, it is also known as having lots of fun playing around with FPGAs.
Powered by reconfigurable FPGA technology, CompactRIO, NI’s lower-cost control and acquisition system is targeted at apps that require high system performance and reliability.
To “enhance” Keynote Host Tim Dehne’s voice—which isn’t all that bad to begin with--Moriat explained that he was going to change the characteristics of several filters that run on the CompactRIO FPGA.
In the first voice-altering experiment, Moriat enhanced the reverberation characteristics of the auditorium, using a handful of delay lines and making the sound more diffuse. The end effect? Dehne sounded like George Burns playing God.
“How about making my voice deeper, maybe a more filling sound?” Dehne asked hopefully.
To produce a lower pitch, Moriat took the signal from the mike, resampled it to get more samples per unit time, then chopped the signal into chunks of approximately 50 ms. He then applied a waiting window to each chunk and stretched it in time—imbuing Dehne with a voice that if you closed your eyes invoked the current governor of California. Then, at the quick tweaking of a few more filter characteristics--Dehne went from being an Arnold to being an Alvin, the Chipmunk that is.
There are of course many more serious ways for engineers to apply the capabilities of CompactRIO. For more info, visit NI.