milosz

Bought a 2013 G37x recently with the premium package. The Bose stereo setup has had lots of opinions expressed but not much data provided.

So, here is the Bose "Studio On Wheels" frequency response, averaged from 9 positions around the drivers ears:




Frequency response Bose G37 system drivers position

This is with the FADE control all the way 100% FRONT and Drivers Audio Stage set to "OFF;" tone controls set to flat.

Here is a distortion plot for the same conditions as above. Black is fundamental magnitude, dark blue is THD, other colors are the various harmonics.




Here is the same measurement with the FADE control set mid-way (50%) so that the rear speakers are also working. Not too much difference in the measurement but subjectively the sound is not as good- the imaging is smeared.




337 sedan Bose sound system drivers position front + rear speakers on

Here is the response of the REAR DOOR SPEAKERS ONLY averaged from 7 positions approximating the ears of a left-rear passenger. FADE control is set to 100% REAR. This is NOT good sounding!




Frequency response Bose system G37 Sedan REAR ONLY - left rear passenger ear location

milosz

These curves show pretty good performance, in terms of frequency response and THD, in fact it's hard to fault this level of performance. This curve pretty closely matches what you need in a car system to overcome lower frequency road noise, and is quite flat through the rest of the range- something of an achievement given the nasty reflections and speaker placement issues imposed by Infiniti's interior design constraints.

As far as car audio goes, this is really quite neutral. That's why the Bose system sounds particularly good on classical and acoustic jazz and folk music.

At the same time, this fairly neutral frequency response will make audible mastering and mixing that is less than stellar. In particular, many pop / rock recordings - especially those contenders in the "loudness wars" - will sound somewhat shrill and lacking warmth and musicality.

In addition, many want a car audio system to sound like a dance club or concert stage on wheels, and these venues often have anything but accuracy or neutrality when it comes to sound systems. So I can understand why a lot of comments about this system have included the phrase "Bose sucks."

It's important to keep in mind that the Bose engineer who designed and voiced this system, John Feng, is a violinist, and not a death-metal shredder. That should tell you something about the design goals here. Have a look at the Stereophile article on this system => Infiniti's New "Studio On Wheels" | Stereophile.com

All that said, the system is not without it's problems. There is a lack of solidity and texture in the bass, and the upper midrange-to-treble regime lacks the coherence and detail that we have come to expect from good home audio systems.

I don't think these two deficits come from frequency-amplitude issues. I think there are time-domain problems here.

From my experience, the kind of bass that the Bose system exhibits- fairly decent LF extension and decently clean lows, but lacking texture and "punch" - comes from a system which is underdamped. The bass audio stops, but the cones - and car panels!- keep moving for a while before the energy dies away.

I think the treble can be faulted also for some kind of "ringing" or other stored energy behavior. Maybe using 3.5" drivers full range (as Bose does here) isn't such a good idea- these drivers might have some breakup mode in their upper response, some ringing that gets excited.

Let's examine both of these "faults" and see what can be done about them.

The bass- in my next post I will show installation of a dedicated subwoofer, which fixed the bass.

The treble- this is much harder to fix. Eq can't fix stored energy issues in drivers. Neither can other digital signal processing tricks. The only thing that can fix this is to restrict the range of frequencies that are fed to the driver to about one-half octave below the breakup frequency or ringing frequency- and preferably a full octave. This is why 3-way systems usually use both a high-pass AND a low-pass filter for the midrange. Here, the Bose 3.5" "midrange" driver is fed audio from somewhere around 200~300 Hz and up. At some point in it's treble, it is misbehaving I suspect. I will have to try to do a waterfall plot of this driver's output to see if I can confirm objectively what my ears are telling me.

The driver used is a 3.5 inch paper-cone driver with a half-roll surround of treated cloth, and a powerful neodymium magnet. I have seen complaints about the "cheap driver" with it's "wimpy magnet" - such comments show a certain lack of knowledge about driver design.





Yes, the driver is definitely designed to a price point. But there's also some solid engineering. Paper is an excellent cone material, especially for midrange. It is self damping and often rather benign in terms of high-frequency breakup. Typically it doesn't have the "oilcan" resonance found in metal, fiberglas, and kevlar drivers. It is lighter than the plastics used for many drivers nowadays and hence more efficient. The cloth roll surround is also a decent choice for a speaker that has to be used from -10° to 40° C. In addition to good performance over a wide temperature range, treated cloth surrounds are very good at absorbing - rather than reflecting or propagating- transverse waves that travel out along the cone from the voice coil.

I am planning on getting one of these drivers and running it through some tests to determine what low pass frequency would best suit it; maybe then a proper crossover to a fabric dome tweeter would clean up this type of flaw.

The other possibility is replacing the driver, but that is likely to create more problems than it solves. The Bose system is designed around the response and efficiency of these specific drivers, and unless the efficiency of any replacement driver is matched to the Bose driver it replaces, there is bound to be imbalance in the overall sound. And these 3.5" Bose drivers are very efficient- the lightweight paper cone, along with the very powerful magnet in the motor system, gives a lot of sound without requiring a huge amplifier. Don't forget, in addition to a cost budget, the Bose designers had to work within a power budget as well. I'm sure Infiniti Corporate wouldn't allow a system that required #4 AWG wiring and a bigger alternator...

Regarding these drivers, keep in mind that there are domestic audiophile speakers costing more than the entire cost of a G37 automobile that use paper drivers; and that magnet size has very little relation to magnet strength. A 1 lb neodymium magnet has as much "strength" as a 10 lb ceramic magnet- just compare the energy product of neodymium magnets to ceramic magnets to see that this is true => What is Maximum Energy Product / BHmax and How Does It Correspond to Magnet Grade? | Dura Magnetics

In addition to the requirement that any drop-in replacement match the efficiency of the Bose driver, it is likely that the Bose amplifier driving the 3.5" midrange is also equalized for this driver, and this same EQ would be applied to the drop-in replacement. The likelihood that a drop-in replacement would sound better than stock is small given these factors.

Some people have gone to the trouble or replacing the G37 Bose midrange and tweeter both, and even fabricating custom "pods" or custom A-Pillars, and of course also replacing the stock Bose amplifier and electronic crossover with after-market units. This can get very expensive quite quickly, and does not ensure results that are superior to the stock Bose setup. Don't forget that Bose and Nissan had lots of good engineering talent and experience with autosound, not to mention fancy measurement and computer simulation tools that the DIY'er or custom autosound shop do not have- in addition,Nissan/Bose could try out many iterations of placement, driver, and electronics until they got the sound they were after. I'm guessing if you build one custom pod for one specific aftermarket driver, you're not likely to try another dozen custom pods and drivers to get the improvement in sound that you're after. So while I'm sure there's some folks who have achieved better midrange/treble detail and time-domain behavior in a G35/G37 than the Bose is capable of, I think these successful solutions are few and far between and likely to cost roughly half the price of a good used 2013 G37x..... or more.

I haven't yet read of any drop-in drivers that work to replace the Bose 3.5" midrange/tweeter. (And the soft dome in the sail is just wired to the 3.5" driver AFAIK, no crossover at all... not much sound coming out of that sail-mounted dome as far as I can tell, although it's possible it's contribution is above the limits of my ancient ears.)

milosz

I chose a Dayton Audio RSS315HF-4 12" Reference driver for my sub, from Parts Express Dayton Audio RSS315HF-4 12" Reference HF Subwoofer 4 Ohm This driver is known for an emphasis on bass quality- which is not the emphasis behind most disco-on-wheels autosound subwoofers. And for the price you get a very decent build quality.




Dayton Audio RSS315HF-4 12" Reference Sub

This driver has dual 4-ohm voice coils, which I wired in parallel to yield a 2-ohm driver. This driver has the following characteristics: Qts 0.39 Vas 84 liters, Fs 24 Hz. Using the box simulation calculator at mh-audio.nl - Home you can see that a sealed enclosure of about 1 cu. ft. with a moderate amount of stuffing will give an overall Q of 0.70 or so, which is considered optimal damping for good transient response. I had Zenclosure ( Custom Sub Enclosure | Affordable Sub Box ) make up a single-woofer sealed box for me, helluva a good price, very nicely made, arrived in a few days. Matches the inside of my trunk very nicely. I used 12 AWG speaker wire from Parts Express, I like this better than plain 12 AWG zipcord because it is more physically flexible- there are more strands than zipcord, this means it bends more easily.

I selected the Alpine MRV-M500 amp, a class-D amp, which I measured to be capable of 521 watts into 2 ohms in my G37. The Dayton Audio RSS315HF-4 12" Reference is rated at 400 watts RMS maximum long-term, so I should be OK here.

I found the best way to connect my amp-and-sub to audio was to make up a wiring set that would plug into the stock factory harness in place of Bose rear shelf sub amp. I bought a stock Bose sub amp from a bone yard (also can be found on eBay) and cannibalized the connector.

Here's the stock Bose rear shelf subwoofer amplifier in the G37 sedan



Here I have unplugged the audio cable coming from the factory harness from the Bose sub amp



Now, I made up this wiring setup using the female connector that matches that male connector hanging off the factory audio harness



...and I plugged my cable into the factory audio loom



Here's how you wire this connector- you get audio and an "amp turn on" signal that is +12vdc = "ON"



Don't use the ground coming from the car's audio harness. This will create a ground loop, and it is nowhere near large enough to supply a power ground for the aftermaket amp. And the 12 VDC power line in the Bose harness is likewise waaayy too small to carry the current the new amp will draw.

The G37 Bose audio system uses balanced lines to carry audio signals. This is an excellent practice, and confirms my belief that the Bose system is a quality design. Balanced audio distribution eliminates noise, treble loss and distortion that could be caused by audio cabling. It costs more to implement, but is by far the better solution in the electrically noisy environment of auto-sound. If you're not familiar with the difference between balanced and single-ended audio lines, have a look at What?s the Difference Between Balanced and Unbalanced? : Aviom Blog.

To convert from the Bose balanced signal pair to the single-ended RCA inputs of the Alpine MRV-M500, I used an RDL TX-1A ( see TX-1A ? Balanced to Unbalanced Transformer - Adjustable ) which I got off eBay for $10. There are many other ways to implement this.

RDL-TX-1A



So, here's a look at the signal coming off the TX-1A, showing a nice clean waveform of the proper amplitude- before I powered up the amplifier, I hung a 'scope here to confirm that my wiring to the factory audio harness was correct, etc.



Here's the amp and the TX-1A screwed to the little alcove under the Zenclosure box, and wired up



Here's the Zenclosure box in place, amp controls underneath there.



And here's the view of the sub box in the trunk



I had the guys at ARMANDO SOUNDS here in Chicago do the power and ground wiring to the amp, I didn't feel like doing that myself. They did a great job for a reasonable price and quickly too.

OK NOW....

Here is the frequency response taken from the driver's listening position with the Dayton / Alpine sub running




When you compare that to the graph above with just the stock Bose system, you can see a bit more 20~30 Hz sound; I could likely trim this down just a bit and lift the amp's crossover up a smidgen so the deep bass isn't quite so peaky, but I have to say that this "extra" deep bass is kind of fun and I'll probably leave it like this for a while.

In addition to MORE bass, the QUALITY is much better - the sealed alignment of this system has lower group delay in the lowest octaves and this results in the improved texture and tautness of the bass.

The slight notch at 35 Hz might indicate that the sub is out-of-phase with the front woofers. I'll reverse the subs' phase and see what difference it makes.

milosz

Here's the frequency response as measured at the left-rear passenger's listening position with the stock Bose 6.5" rear door speakers; FADE control adjusted to 100% REAR, no front sound at all. This is pretty awful: a +40 dB peak at 265 Hz... yuck.




Here's the stock Bose 6.5" driver, they run this full-range.




I replaced these stock Bose drivers with Polk DB652 coaxial 2-way speakers, these dropped right in



Now here's the frequency response at the left, rear passenger's ear position with the Polk DB652 in place of the stock Bose 6.5" (I also have a 12 inch sub running here too)



This is not GREAT but is much better. It doesn't make much difference from the driver's listening perspective, but any passenger in the back who wants to hear music will be happier, I think.

This curve seems to show that the Bose amp driving this rear door speaker does not employ any EQ.

Entcee

iTrader: (3)

Great info!! I am curious, why you tapped into the wires after the main amp (before the sub amp) instead of the wires before the main amp (from the head unit)?

SonicVQ

Very interesting data, but why set a fader at 100% to isolate the front or rear?
Don't all speakers work together to create the desired sonic stage?
(It is my understanding the Bose is the master of phase manipulation to create more depth)

milosz

The main Bose amp in the rear just passes a balanced line level mono signal summed from the left and right front head unit to the rear sub amp.

You could tap into the "center" input to the rear main amp and feed that to the aftermarket sub amp - this signal coming from the head unit to the rear amp is present on pins 29 and 30 of the main amp harness connector- but that would require splicing into the harness. This signal will be controlled by the FADE control, since it is a front signal. It would work but it was just easier to disconnect the existing rear shelf sub and use that connector to feed the aftermarket sub. This disables the rear shelf sub, which is a good thing I think - and gives you the exact same signal that pins 29 and 20 have. I prefer using a connector rather than splicing into harnesses. To splice into the harness I'd have to lay on my back inside the trunk.... I'm 62 years old, not something that is all that easy for me.... using a connector, I just plug it in, takes 5 seconds. Using a connector also makes no modifications in the stock vehicle- this might help resale value. Maybe.

milosz

I don't like running the rear speakers with the fronts. To me, this smears the stereo imaging. I do not get any sense of extra depth from this. But, certainly some folks might like it.

I did measure the response from the drivers position with FADE set to 50% as well as set to "just front." There is a graph for "front + rear" - from the driver's position there's essentially no difference in response.

Here's that graph - FRONT + REAR FROM DRIVERS POSITION



AND HERE'S FRONT ONLY - not much different

Entcee

iTrader: (3)

That makes sense about the connector. I Actually have an apointment this morning to have my amp rewired using the pre main amp wires. (I don't like what the Bose amp does to the bass signal). Here is my current situation https://www.myg37.com/forums/audio-v...l-options.html

dwb993

Thank you for this very detailed info.

Everything you mention jibes with my experience in my 2013 xS coupe. I added a 10" Rockford Fosgate sub in a 4080 box (no fill) with an RF 500W amp and got very similar results as you.

When you write, "From my experience, the kind of bass that the Bose system exhibits- fairly decent LF extension and decently clean lows, but lacking texture and "punch" - comes from a system which is underdamped. The bass audio stops, but the cones - and car panels!- keep moving for a while before the energy dies away.", did you mean under-AMPED? Quite different from under-DAMPED.

I have found that the coupe rattles quite a bit from the door handles and mirrors and requires Dyna-Mat or similar, though I am thinking it would not take much coverage to quiet things down. Are there similar areas on the sedan that should be damped?

milosz

I haven't had any audible door rattles. The only rattle I have in my 2013 sedan is the trunk lid along the bottom of the lid, seems to be at the latch strike itself- and this is only when I am playing bass-heavy material louder than usual. The rattle is only audible within about 5 feet of the truck on the outside of the car. No rattle heard inside.

When playing 20 or 30 Hz sine tones at maximum level (just below clipping of the Alpine subwoofer amp-) there are a good many rattles in the back. The trunk latch rattles, and some cabling or drain tubes are "flapping" inside the cavities at the bottom rear of the back fenders. Music hasn't caused these to rattle, but the loud continuous tones did. I've come to the conclusion that there's no real way to completely "rattle proof" my car when it comes to such extreme SPL's at 30 Hz. (Many objects in my garage also rattled, including the roll-up garage door.)

Now, my sedan has 38,000 miles on, it is fairly low-mileage. It's quite possible that things loosen as the mileage goes up. I understand that Infinitis' are prone to developing squeaks and rattles as they age.

But so far, playing music as loud as I want, nothing is rattling

I DID have two rattles when I first installed the sub. One I traced to the front passenger seatbelt buckle - the belt had gotten twisted and did not retract all the way, and the buckle was rattling against an interior panel. Untangling the belt and allowing the thing to retract fully cured this rattle.

The other rattle was from the trunk pass-through door. It's a plastic panel and it was rattling loudly in it's frame. (Also begin to rattle over when driving over potholes and speed bumps) - this door was not latched. There's a latch you can lock with the cars' key. Once I locked this closed, the rattling ceased.

FYI I bought this car in early November and I am still learning about features and so on. For example, I just learned bout the socket for the "Intelligent key fob" which will allow you to start the car even if the fob's battery has died. There are so many features....

I came to the G37 after owning a 1994 Celica for 20 years. That was a very reliable car, looked kind of nice and was kind of fun to drive, though not very powerful and very front heavy in terms of handling. Still, it served me very well. Coming to the G37 - it is so much more car! A totally different class of car. And WAY more fun! Cars have come a LONG way since 1994, hahahaha. Interestingly, taking inflation into account, I paid about $1,000 more for the '94 Celica when I bought it in 1999 with 62,000 miles on it than I paid for my "certified" 2013 G37x, with 37,000 miles on it. AMAZING!

milosz

The sedan passes un-equalized (balanced) full-range line level audio to the 100 watt Bose amp mounted on the bottom of the rear deck 10-inch low profile woofer. The big Bose amp in the trunk doesn't change this signal, it just passes it through.

The amplifier mounted on the rear deck woofer itself has a low-pass crossover and EQ filter in it. Since I am taking the signal that feeds the rear woofer amp, I am using unequalized, full-range audio to feed my Alpine subwoofer amp.

By the way, here's some photos of the Bose rear deck woofer amp. It's hard to see in the photo but there's a chip (marked U2 on the board) in there labelled BOSE and it is this chip that does the PWM of the analog audio to drive the output switching power MOSFETs. It appears that by adding various resistors (and maybe capacitors too, I don't know) this chip can also perform various filter and EQ functions. I don't think it does any time-domain adjustment. I think any of that sort of thing - like the "Drivers Audio Stage" feature - is done by DSP in the head unit.

This amp is pretty well made, the parts that dissipate heat are mounted in a ventilated area and covered with some kind of enamel or varnish to protect them from moisture etc.

Entcee

iTrader: (3)

I did not realize the main Bose amp sent an unaltered signal to the sub amp.

Thanks for clarifying!

G-thirty7X

Wow! Thanks for taking the time to gather frequency response graphs for the stock system.

I've been running mine with the "Drivers Audio Stage" set to ON as I really like how it raises the sound stage.

Did you test at all to see how the Drivers Audio Stage setting changes the frequency response?

blazeplacid

iTrader: (1)

I'm one of the few who will do what it takes for great sound quality.

I have removed all bose compoents in my car.
I run an Audison Bit ten D that is fed signal from a Nexus 7 via an optical cable.
My mids are B2 reference and my tweeters are Audiofrog GS10.

The bose system in any car just doesn't sound "right" to me. Some people like it.