Today I had the opportunity to put the 1/3rd octave version of TrueAudio's TrueRTA software to the test. We've got another car audio show coming up on Monday, and one of the potential competitors wanted me to tune the system. He'd just swapped out his previous subwoofer box for a new one, and the installer had made a few other "adjustments" to the system.
The result was a sonic nightmare. Boomy bass, piercing highs, no midrange to speak of, distortion at any volume level above 5 (volume control goes up to 30) and loud background hiss, no matter what the volume was set at.
After sorting out the gain settings on the amplifiers (the source of the hiss), and fixing the phase problem with the main speakers and the subwoofer, the system was sounding a bit better, but the highs were still a bit piercing in nature. So I brought out my laptop with the TruRTA software and used my Monacor SPL meter in its C-weighting setting as the pickup. With the C-weighting, the response of the meter is rolled off at either end of the spectrum, but that didn't bother me too much, as the intent here was to smooth the response, not flatten it.
For the measurements, I set TrueRTA's range from 50 Hz to 10kHz, put the SPL meter on top of the driver's set (no headrest) then used a slow sinewave sweep as the source signal. Normally pink noise is used for these types of measurements, but the advantage of the slow sine wave sweep is that you can also quickly identify problem areas caused by distortion, either from the drivers or from the environment itself (e.g. vibrating panels, etc.). Apart from being audible in most instances, this distortion shows up on the RTA as any response above the fundamental frequency (during this test we quickly identified one source of distortion - loose protective covers over the 4" drivers on the dash).
Well, TrueRTA immediately identified a number of problems, including a gap between the low bass and midbass and a "hump" in the response stretching from about 3kHz up to 8kHz. The former was caused by the subs being out of phase with the rest of the system, which was easily corrected. The latter was caused by two name-unknown tweeters mounted on the dash, which were basically overpowering everything else up front. To fix this, basically all I did was adjust the EQ's upper bands until the "hump" disappeared entirely from the response as measured by TrueRTA. It took a little time to do, but the end result of all the tweaking was a downward-sloping but smooth response, with two or three response peaks that were too narrow to be effectively dealt with by the system's 12-band EQ. Note that trueRTA's settings for frequency and SPL ranges are easily changed in the main window, and this definitely came in handy when dealing with the problematic areas in the frequency response.
Well, after about an hour of tweaking, we were getting a littled tired of all the test tones and sine sweeps, plus we were anxious to test the results of all the tweaking, so the test CD was replaced with some music. Note: up to this point we hadn't listened to ANY music on the system - all the tweaking was done using the slow sine sweep, plus test tones (for setting x-over points).
And damn, did we ever hit the sonic jackpot! Nice smooth response, from the bass all the way up through to the highs, with nothing sounding out of place. The system's owner was incredibly pleased with the results, as was I. I think I've definitely got one more useful tool in my audio toolkit :-). Best $40 I've spent in a while. If you've got a laptop with a decent sound card, and you're trying to get the best out of your car audio system, you might want to have a look at it.
Regards,
Brian Steele
www.diysubwoofers.org
|
|