Hello, I was using a program to record 8-bit waveform audio to a file and then plotted the recorded data in Excel. I was testing this program with a simple 1kHz sine wave. With 8-bit data, I had no problems whatsoever. I was sampling at 44.1kHz and everything worked fine. When I moved to 16-bit sized samples, however, I'm seeing this strange behaviour where I'll see my sine wave, but about half the time my samples are approximately zero. I.e., over 40 or 50 thousand samples, I'll see a sine wave overlapped with what appears to be a DC signal at approximately 0 (it actually varies from say +100 to -100, whereas the actual sine wave has a peak of around 15,000). I can't see a pattern right now; my data appears to be two or three samples of sine wave, then two or three samples of near zero, and this just repeats and repeats. Any ideas out there? Regards, Roch

Does it sound choppy when played back? -Burbble

Hi Roch, Do you mean you are looking at the interpreted data within Excel, or some other means? I'm wondering if the problem is with the transition to the spreadsheet, or in the wave files. If you do not have access to a nice waveform editor, you can grab a free one at www.cooledit.com (I think the free one is called Cool Edit 2000). Fantastic products (and a hell of a lot of fun!). Needless to say, if a separate editor sees no DC, then I would presume the files are fine. Otherwise, perhaps there is a glitch in the interpreter with 16-bit stuff. Then again, if your waveforms are long enough, what Burbble suggested might be a much simpler test than my suggestion.