This evening Chris1968 brought his nosdac over so that we could see if we could figure out what was wrong with it. Connected up, and the first thing to do was bring out the Maroon 5 that had upset it before and see if it was still happening. Answer was a very definite yes. So, we ran through a few other discs to see if we could provoke it with anything, and failed. However, we knew it had been provoked by other discs. After attempting a few other things with the behringer as spectrum analyser and some test tones, there was still nothing to go on. Time for the oscilloscope. With most test tones the output was fine, but then we hit on the problem that we'd heard at the bake-off: [img]http://www.askisaac.com/images/clip.jpg[/img] The top trace is the analogue output of my Toshiba SD330e. The bottom trace is the nosDAC. Signal is 1kHz, 0dB FS. The bottom half of the sine-wave is clearly clipping, and this was resulting in the rather unusual effect heard at the bake-off and subsequently. Knowing now that it is an amplitude issue, a quick run of Massive Attack's Angel prompted misbehaviour in a low-frequency signal (which we hadn't heard before). This is the fault with the unit. The other thing that you'll notice about the output of the nosdac is that the trace is not smooth like the DVD player output. This is inherent to the nosdac design. Using a 15kHz -14dB FS signal shows up the difference much more clearly: [img]http://www.askisaac.com/images/nos.jpg[/img] Again, top trace is the DVD player, bottom trace is the nosDAC. This is not a fault condition, but simply what happens if you don't filter or oversample. This could well be characterised in listening as "harsh" or "grainy" or "coarse".