I’m attempting to write a script, based on
usrp_fft.py, that will read a file of gr_complex and
display it. It works, except that it plays back much
faster than it was recorded. For example, I recorded
a five second file and gave it to my script. It
looked to be correct data, but it finished playback in
less than a second. Here’s the essential code:
There is a gr_throttle block that you can use to rate-limit your data. I
think the only rate-limiting going on here is CPU time.
Correct. The GNU Radio runtime scheduler will always execute signal
processing blocks at the fastest rate possible given data availability.
With “real” sources like the USRP, the flowgraph usually becomes
rate-limited by the data production rate of the USRP. Sometimes it’s
the data sink that does the rate limiting, like an audio sink.
In the OP’s flowgraph, there is nothing holding up the scheduler other
than the hosts CPU speed, memory bandwidth, or hard disk controller.