I've previously admitted to becoming an OmniMic measurement system junkie, spending a lot of time measuring my music system and tuning it endlessly by changing bass EQ, crossover points, and speaker and subwoofer positioning. WRT frequency response at my listening position, all of this measurement and adjustment has paid off very well. Not only does the system now measure a lot better, it sounds better.
One aspect of the OnmiMic's capabilities I had previously stayed away from, measuring distortion, because I figured it was a useless exercise in a home listening room. I've been a little bored lately, being so happy with the way my system sounds I have nothing to complain about, and I'm basically done tuning it. I was listening late in the evening a couple of days ago, my wife wasn't home, and on a lark I brought up OmniMic to do some perhaps useless exploration. First, I wanted to see how quiet my room was, which is important for distortion tests. I turned off the HVAC system to make sure it couldn't run, and even turned off the refrigerator so it wouldn't make any noise. OmniMic said my room averaged 29db, with a peak spike of 33db centered at 140Hz; I suspected some the breeze outside. Not bad for a house in a dense suburb! I wonder what the OmniMic's noise floor is?
Anyway, for the first time I tried OmniMic's distortion display, with 1/12th octave smoothing. The mic was on a boom at about my ear level, six feet from the right speaker, which is in a very open part of the listening area. At 80db at 1KHz, which meant that the way my system is tuned it would be about 83db at 30Hz and about 77db at 12KHz, 2nd and 3rd harmonic distortion averaged 45db below the fundamentals, 47db lower from 1KHz-3KHz, and no higher than -40db. Not bad! I didn't think measurements in-room would yield numbers like that. Remember, -40db is 1%, -50db is 0.32%, so considering these measurements weren't done in an anechoic chamber, that's indicative of very low overall distortion (unless the measurements are completely bogus, which is always possible when performed by a non-expert).
The wide-range measurements were one thing, but I was more interested in the bass octaves. I run my mains, Revel Salon2s, full-range, and I use a Velodyne DD18 Plus with an 80Hz low-pass filter, and some rather complex EQ in the Velodyne intended to make it a bass "filler" to smooth in-room response, rather than a classic sub-sat system. Measuring bass distortion is a lot more difficult than in the upper octaves, so I used a very sophisticated setup
. I put the mic on a boom so that it was set two feet off the floor, with a vee-configuration of stacked king-sized pillows behind the mic. Seriously, I found the pillows improved the distortion readings by about 2db.
My first test used track 3 of the OmniMic test CD, which is an all-octave sine wave sweep. Even with foam ear plugs in, this track is very annoying at 90db, so I only did a couple of bass measurements this way, but they intrigued me. At a measured 90db at 30Hz, bass distortion was 42db down from the fundamental on the complete system, and nearly 45db down on the Salon2s alone. I did the test twice, with the same result. Hmmm... that's not what I expected.
Track 3 was annoying me so much that I switched to specific test signals from audiocheck.net. The 32Hz signal played at 90db as measured by the mic did indeed have about 2-3db less distortion on the Salon2s than on the Velodyne. Not that the Velodyne's distortion of less than 40db below the fundamental is bad, but it still surprised me that the Salon2s were better. Just a guess, but the Salon2s are ported and the Velodyne is sealed, and that might be a factor.
At 20Hz and 90db, which could be felt more than heard, the distortion on the Salon2s with the Velodyne was still below -40db in the 2nd and 3rd harmonics. I also ran a test at 100db at 20Hz, and the full-system result was inconclusive. Too many things in the room rattled, the windows behind the system were vibrating (not audibly, but I could feel them moving quite a bit), the fireplace glass and screen were vibrating and rattling, a lamp was rattling, and as I was trying to debug all of the noises I was getting slowly nauseous. It sure was impressive sounding though. Trying the 100db at 20Hz test just with the Salon2s required the level control to be turned up quite a bit, while the Velodyne alone seemed to be cruising. Audibly I perceived little difference between the tests of the Salon2 and the Velodyne alone, once the levels were matched on the OmniMic, though the Velodyne seemed cleaner, if I had to guess.
None of this is at all conclusive or scientific, but it was fun. I'm also wondering... would the Funk 18.3 I've been thinking about have substantially lower distortion than the Velodyne. I'm getting closer to wanting to find out.
