What's New

Projects

My Autos

Multimedia

Favorites

Part  1 | 2 | 3  | 4 | 5 | 6 | 7 | 

Part 6: Tuning and System Wrap-up:

 After living with the system for a week or so I finally had time to see if my suspicions were correct about the frequency response being attenuated above and below certain frequencies. Earlier I had theorized that Bose was using some sort of compensation circuitry to roll off the high and low frequencies to protect the flimsy factory speakers. Using my laptop computer and RTA software (see side bar) I was able to confirm that the system was not getting a full range signal and as volume increased (using the factory volume controls) the high frequencies and lower rolled off even more.

 With factory volume set to 9 and the JL CleanSweep volume about 3/4 of the way up, no EQ in the mix and the CleanSweep defeated I measures a not so flat signal between about 40-16,000 Hz. I say not so flat because besides a strong boost in signal at about 125 hz, there where addition dips and boost all the way up the frequency range, the biggest boost coming at 320, 1k, and 8k. Above 8k the system rolled off at about 3db-6db per octave. I next did the same measurements with factory volume both a little lower and higher and determined that there was a circuit that cut frequencies based on volume position. With the factory volume control at about 12, I couldn't get much of a signal below 80 Hz, the higher frequency were toned down a bit, but roll-off wasn't as pronounced as in the lower part of the spectrum.

 When I engaged the CleanSweep and did the same measurements with the unit engaged to flatten the frequency response I was impressed to say the least. Most of the dips and peaks seen in the original RTA curve were gone, with the exception of a few 3-6 db bumps at random frequencies, which obviously were more due to car acoustics than the electronics. Even with the CleanSweep on there was still steep roll off as mentioned above, but in this case the mid-range and other frequencies were more of a flat curve. Showing that the CleanSweep does work, however like they say... "junk in, junk out".

 

After consulting with JL Audio and having several discussing on the online forums, I decided to try and sum the signals from the outputs of the Bose amplifier for the dash mounted tweeters and front door speakers so as to get a full range signal. Summing is generally done using a Line Output converter (LOC) and a set of "Y"-adapters at the RCA's, supposedly by taking two signals you can sum them together and get pretty close to a full range sound. I attempted to do this using a Scosche 4 channel Line output converter, I connected the Dash Tweeter outputs to the LOC inputs market front and the front Door outputs to the rear speaker inputs on the Scosche.

Sidebar: RTA Measurements

In our install we have run into several issues revolving around sound quality, and more importantly the frequency range of the system. A useful tool in any audio enthusiast tool chest is a good RTA, or Real time analyzer. However for most of us the idea of spending as much as $1000 for a single use device is out of the question, enter the work of software based RTA's. An RTA is used to measure audio signals in a stereo system or in other applications and generally analyzes signals between 20-20,000 Hz. Most humans can hear signals between about 30-17,000 Hz, but a good system will reproduce the full spectrum of sound, and when we say full range, we generally are referring to 20-20k. In car audio an RTA can be useful to determine not only the frequency range, but where the peaks and dips in that freq. range is. For example, a big SUV with and open interior may augment sound at a different frequency than say an open top convertible. In a 4door sedan like the Infiniti M45 may have a boost in the upper midrange, were our SUV due to its size and the fact that sound waves have further to travel to our ears may have a boost in the lower bass range. This is why it is easier to get really loud bass in an SUV or van rather than a car, and in our convertible, you generally have a hard time getting any low frequencies. In our case, the RTA will tell us what frequencies are being canceled out, or are dipping and which ones are overly augmented and peaking. Then we can take a good 1/3 rd octave Equalizer and make adjustments accordingly.

 For my system evaluation I am using an old Dell L400 sub-notebook computer (600mHz, 256MB Ram, 40GB HDD) a copy of  Pocket RTA (designed for Pocket PC, but you get a PC version when you purchase it) and a Sansom C01U USB microphone. This setup doesn't give you a perfect 20-20K but it is close. By utilizing this tool I can adjust the system in order to get the smoothest sound and eliminate and problems that in car acoustics may cause due to artificial boost in sound. The reality is that using this methodology I am measuring the in-car response, not what the system is capable of. Since in-car response is what we actually hear, that was ok with me. But for a more actuate measure of the frequency response and to see how flat the signal in the system really is, one would need to do a direct connection between the outputs of the CleanSweep and the laptop and by-pass the microphone. This would give a true measure of the systems response without the effects of the car on the sound.

 After this failure I decided to take a chance and splice into the system before the Bose amplifier. The Infiniti Service manual shows a single, L/R channel coming from the factory dash head unit to the amplifier which is not part of the system Bus. I hoped this would prove to be a somewhat full range signal. Since this method would only provide a single L/R channel I would loose Front to Rear fading capability, but this was a trade-off I was willing to make for full range sound. Upon making the connections I fired up the system, to my satisfaction the CleanSweep's LED indicators for the first time instead flashing red and green were a solid green on the front two channels, this indicated a true full range signal at the CleanSweep. Due to there not being a rear channel output to connect in to the CleanSweep, the rears failed to calibrate all together. I quickly switched to a musical source and was rewarded with the bright clarity of the MB Quart tweeters at their full band-width. Problem Solved? or so I hoped.

 At this point I did have a good full range signal, and with the capabilities of the MB Quart 4-channel amplifier was able to adjust a nice front to rear bias without a fader. But as I played with the system I realized that there was a sacrifice for the full range sound and that was to be the ability to utilize any factory volume adjustment, or tone control. Evidently, all volume control and speaker adjustment, including steering wheel volume control is all done inside the Bose amplifier, with the signal by-passing this, all convenience was lost. Making the dash mounted JL Volume control the only way to adjust output. This for now, would be something I could live with.

After securing all my connections and making them more permanent, I proceeded to setup the Laptop again in order to take RTA measurements. I first measured the Frequency response without the CleanSweep, then with and then I proceeded to adjust the PPI 1/3rd Octave Equalizer.

 The sound as measured through the RTA software with CleanSweep defeated was in fact pretty close to the ideal 20-20,000 Hz. The curve itself was far from flat, but wasn't bad. There were a few anomalies at 80, 1.67khz and at 8khz which could have been boosted due to the interior of the car. When the CleanSweep was activated is when things got interesting. The premise behind the CleanSweep besides its supposed ability to be a digital LOC, taking High-level speaker outputs and digitizing them and then converting them to low-level signals which can put outputted via analog RCA cables is to adjust any factory EQ curves to a somewhat flat signal, so in turn users have a starting point for making adjustments through other equalizers. In this case it was obvious the CleanSweep was making some attempts to flatten the curve however the changes were not very dramatic. It was obvious that it was attempting to flatten things out, but only by a few dB's here and there. As mentioned in the sidebar above, it was more likely that one would need to measure the outputs of the CleanSweep using direct connection to the Laptop, instead of a microphone on the interior o the car to determine if the CleanSweep was working.

After using the RTA to measure frequency response with CleanSweep engaged and defeated, I then moved to make adjustments using the Precision Power DEQ-31. This 1/3rd octave EQ has 31-bands of equalization between 25 and 16,000 hz which can be adjusted up or down by 12 decibel's. Based on the plot taken with the EQ out of the loop, I adjusted the system to a curve which I thought would give me the sound I wanted, this curve will continue to be tweaked as time goes on, so as to give me the sound that I want. The one thing experience tells me is that everyone is different when it comes to the sound they like, thus a curve that looks good to one, make not look or sound right to another.

As far as the CleanSweep goes the jury is still out, it obviously has it's uses and I have heard several systems which it makes a huge difference. It maybe that the circuitry in the Bose amplifier on the M45 is so different from everything else that nothing could help it. The question for me is, does the CleanSweep in this configuration offer me any benefits over say a standard Line Output converter, or a comparable piece like Audio Controls LC-6? It has been suggested that with a second CleanSweep, or an attempted to connect the rear speaker outputs back into the CleanSweep for a none full range signal that I might be able to get my fader back, but to me... it is the factory volume control which is more important. Only time will tell if I miss the factory volume control or not.

Part 7: Conclusion and final Cosmetic work