The ultimate in system tuning
Mike has worked for Crutchfield in various full- and part-time capacities since 1988. Ten years ago, he joined the company's creative department to write and edit mobile A/V media. He is also a professional musician and composer.
More from Michael Sokolowski
All along, I knew I was going to take advantage of the CDA-9887's IMPRINT ™ sound tuning capability as the final stage in the audio makeover of my SUV. Let me explain: IMPRINT is an add-on Alpine system that employs MultEQ technology innovated by the Audyssey company for home theater applications years ago. More recently, the company has brought this technology into the car; new Volvo S60s and Jaguar XJs come equipped with Audyssey systems, while Alpine outfits some aftermarket gear with it. The idea is that, once the components are optimally installed in the listening space - in this case, my vehicle - MultEQ listens to test tones generated by your system (from several listening positions) and evaluates frequency response and distortion-causing reflections in the path of the sound. Analyzing this information, it then calculates equalization curves to ensure accurate production of sound waves to all of those listening locations. It's as if a musician could simultaneously play on stage and sit in a house seat, adjusting his or her performance for best articulation to that seat - as well as to every other seat in the hall! MultEQ is something of a magic trick, and it has to be experienced to be believed.
The first thing we did was to grab the KTX-100EQ sound tuning kit out of returned inventory; there were plenty there, as we encourage customers to buy it and send it back for a refund less a $50 re-stocking fee. (You only need to use it once, so $50 gets you something a dealer would likely charge significantly more for, and it's a very simple do-it-yourself project.) We loaded up the software on a laptop PC and after a couple of false starts trying to install IMPRINT (finding it buried in the Windows menu system), we were soon up and running and heading outside to the truck.(Click on images to enlarge.)
We placed the laptop on the Yukon's hood, plugged in the USB type A connector from an included 16-foot cord and ran the cord into the vehicle around the passenger side's pillar post. We then connected the mini-USB plug to the processor module and set it on the passenger seat. Next we grabbed Alpine's connecting cable for the receiver, plugged it in and ran it to a proprietary jack on the head unit. The last hookup step was to insert the mic cable's mini-plug into the processor. These three connections would take about a half-minute to execute. Now we were ready for mic placement.
Crutchfield's digital imaging buyer was kind enough to lend us a mini-tripod sample onto which the threaded base of the microphone screwed down nicely:
We adjusted the stand to ear height and placed it at the back center of the driver's seat. We made sure all the windows were tightly closed, all doors completely shut , and the system configured the way I normally use it (with the left rear passenger seat back down, exposing the subwoofer). We chose a quiet spot behind the building that was unlikely to be disturbed by noisy vehicles and clicked on the RTA process from the laptop. Standing outside this highly dampened, luxury vehicle, we could make out the quiet, alien-sounding emissions of test tones and pink noise coming from my speakers from which IMPRINT's frequency analysis would be made. This first location's measurement took about 10-12 minutes (didn't stopwatch it, unfortunately), but the subsequent five more positions (passenger front, passenger rear left/right, and two more on the driver's seat) would take about 5-7 minutes each.
After each measurement was taken and stored, the on-screen interface would clearly indicate the next spot to position the microphone. We'd dutifully move it, shut the door, and run the next reading. Rinse and repeat. Forty minutes later and the last seat position measured, we took a look at the results:
Pretty impressive, huh?
The "before and after" frequency graphs show much smoother overall response. Notice how the "boom" was taken out of the bass and replaced by a much more even transition to the mids and highs, ensure greater clarity in the lower realms, without an artificial, non-musical, accentuation of the low and midrange frequencies. You can also see a much more even response curve in the mid and high frequencies; this translates to fuller, rich tone where the instruments sound more like the instruments they are.
Even more interesting, and equally dramatic, is the time response comparison. Think of that jagged line at the bottom of the "before" picture as opacity or smearing. The "after" image indicates a much clearer aural picture, where tubby, echoe-y sound is tightened up and focused; we've moved from the bathroom to the concert hall.
The next few laptop screen captures show how the IMPRINT software allows us to choose frequency curves we'd like to send into the head unit for permanent storage and recall. I chose the unfettered "reference" curve and the curve with some midrange compensation. I decided to eschew the "linear" curve, which strikes me as somewhat missing the point - too theoretically optimized, hence a little unnatural.
Once the curves were saved and sent to the CDA-9887 for storage, we disconnected the laptop, pulled the cords and mic stand from the vehicle and hopped in to check it out. Remember, I already thought the system sounded great, so when Jeff Fay (car merchandiser extraordinaire and former product guru at Alpine) called me in to listen, I immediately said, "hey, this sounds much better." To which he replied, "that's nice, but let me turn on the effect first." Suitably embarrassed, I shut up and listened. He pressed the MultEQ button, calling up the reference curve...
Everything changed in an instant: BAM! The sound jumped up off the floor and centered itself in the front of the cab. The concert stage, as it were, was now on - and slightly in front of - the dashboard. The surrounding soundstage effect I previously reported was gone with the press of a button. Before, I could tell the sound was coming from the speakers in their lower door panel locations; now, the illusion is complete. The sound "appears" self-contained and three-dimensional in front and slightly to the right of me when I'm driving (and slightly to the left of me when I'm riding in the front passenger seat) - as if I were piloting a concert hall from the front orchestra section. In my last post I mentioned feeling trapped onstage among the instruments, dodging drumsticks; now I was back in the expensive seats, where I wanted to be.
But superb imaging isn't even my favorite part of the MultEQ effect; the best part is the unbelievable clarity. The processor also compensates for the myriad distorting reflections in the trajectory of the sound. Think about a car's acoustic environment; could it be much worse? Speakers mounted in out-of-the-way locations and firing against ceilings, windshields, seats, and each other - all set in a box-like space lined with glass windows. Even the finest mobile audio components must cope with the challenge of reproducing music in this sonically hostile environment.
Other real-time frequency analyzers simply evaluate frequency response from several listening locations and average the results into a single curve against which a system's output can be mirrored. So if there's a bump at X frequency in one seat and a dip at the same frequency in another, they cancel each other out and no correction can be made. MultEQ, on the other hand uses a much more complex and effective process whereby similar time domain responses are clustered together, a single response generated from each cluster, and the final curve created by grouping these representative clusters. Pattern recognition and fuzzy logic play into the algorithms that produce the EQ filter - don't ask me how it works, but I am here to testify that it surely does work. (If you want to learn more about how it works, visit the Audyssey site or just ask them.)
Now, what about the curve with midrange compensation? It adds a tad more bite, a little more edge to the electric guitars, for example. Essentially, it's the same improvement over the non-equalized sound that we experienced with the reference curve, so I find myself sticking with the former.
While I thought my system already sounded great, this easy, D.I.Y.-friendly process showed me that I was listening to a muddy mess in comparison to how a properly EQ'd car stereo should sound, and it fixed it for me. It's astounding to me that, without changing any equipment, an upgrade of this magnitude can be achieved This change was easily as dramatic as when we replaced the factory receiver on the first day, or when we popped in the C2 components in place of the cheap Yukon components. I'm truly blown away and want to share this discovery with as many people as possible. It would be a great thing if more manufacturers would use this technology and more people would discover it, such that it would become available at lower price points. This kind of sophisticated real-time analysis could help even the modest gear to perform much, much better. Clearly Volvo gets it, if they're including it with some of their factory equipped radios. C'mon aftermarket industry, let's get this amazing listening experience into more cars!
Read the entire Building a Mobile Listening Lab series:
- Laying the Foundation, Part 1
- Laying the Foundation, Part 2
- Laying the Foundation, Part 3
- Dynamat installation and pre-wiring for amplifiers
- Installing the front door speakers
- Video: factory vs. aftermarket speakers
- Rear speakers, amplifier, and subwoofer
- Report on the Yukon system's performance
- The ultimate in system tuning
- Video: Alpine's IMPRINT signal processing in action