09-01-2006 02:58 PM
Hi CSUEB,
Sorry for the long delay to reply - combination of being sick and heavy work-load...
1. I don't remember how I fished this number out! However it's probably not a coincidence that if you multiply your sample-size (2500) by the number of bytes per sample (8 (for Double-Precision numbers)) and add 4 to store number of samples, the result is 20004. I wouldn't be surprised if there is a cleaner way to read-back these wave-forms. Feeding the File Read the exact byte offset, but having "Count" in terms of waveforms just seems wrong. This coding approach is fragil, since it will "break" if sample-count varies from waveform to waveform.
2. The short answer is yes! LV6i didn't have event structures - a more efficient way to respond to user-events - so the loop has to "poll" a control to see whether it's value changed. 333ms is a guess at what's as slow as possible, without introducing a delay the user will notice. In an application like this, many programmers would just let the loop "free-run" (no delay) but that's a bad habit to get into!
Cheers.
09-01-2006 03:13 PM
Er... slight correction ![]()
Your application is plotting 2500 points, two SGLs per point! (not 1 DBL per point (my brain was in CHART mode...))
Anyway, 2500 x (4 + 4) + 4 = 20004
Hey, it happens I'm loading/saving plots off a Network Analyzer. If I figure-out clean waveform File-handling, I'll post again. ![]()
Cheers.
09-05-2006 06:58 PM