Measurements with a sound card

I am using a USB sound card for measurements. It's a pain, but I have nothing better yet.

The sound card is the E-MU 0204. I've chosen it for the specs:

• Dynamic Range (A-weighted, 1kHz, min gain): 113dB
• Signal-to-Noise Ratio (A-weighted, min gain): 113dB
• THD+N (1kHz at - 1dBFS, min gain): -101.9dB (.0008%)

The sound card is intefaced to the real world by Pete Millett's sound card interface.

The software is SpectraPLUS 5.0 by Pioneer Hill Software (PHS). (No link because my antivirus says their web site is infected.) The software runs on an old MacBook running Windows 7 in Boot Camp.

The results do not seem to be affected by nearby computers, flurescent lights, switching power supplies, etc.

The results are affected by:

• Windows Mixer settings (levels and sampling rates)
• Software settings (levels and sample rate)
• Grounding of the sound card, the interface, and the device under test

The greatest source of pain is the Windows Mixer, which adjusts both input and output levels and can introduce its own distortion. (Apparently, the latest version of SpectraPLUS can work with ASIO, but I have yet to try that one.)

In the screenshots below, you can see the results I could obtain after carefully fiddling with the software settings.

• The first pair of screenshots is the E-MU 0204 connected to itself by a short shielded twisted pair. The numbers are impressive (to me).
• The second pair is the E-MU plus the interface in the "generator monitor" (= loopback) mode. You see a clear increase in the noise, although the performance is still very nice.
• The last pair is the below mentioned NE5532 power amp delivering ~10W into a 8ohm resistor. It seems that the sound card interface noise dominates that of 32x NE5532.

Update: Pioneer Hill Software updated SpectraPLUS, so no it works with ASIO, bypassing (and thus relieving the pain caused by) Windows Mixer. There is a separate, later post discussing the results with ASIO.

Attached Thumbnails