[NealfromNZ]

Has anyone got an acoustic model for cabin gain or transfer function for the F30 / F80 chassis ?

In particular transfer function of a trunk sub and cabin gain and loss for interior speakers. Starting to model a house curve for pending DSP setup in REW and would like to overlay cabin gain.

[Billfitz]

[NealfromNZ]

Very useful video thanks. Have some work to do so will post up after some measurements etc.

[NealfromNZ]

Just revisiting this thread with some modeling using REW Application room simulator. But basically is using the calculations in Bills youtube link.

For the cabin I've used measurement from firewall to rear trunk and came up with the following.

3.3 Meters firewall to trunk edge
1.6 Meters wide
1.0 Meters high (Cabin part )

So the interesting thing with the model is goes a long way to explaining why a trunk sub will roll off at 60 hz in a BMW F30 / F80. The other interesting thing is it confirms where the most gain is for a trunk sub position.

So the most bass for an equal sub as heard at drivers seat is same side as driver in the back corner, second is putting a sub in the under trunk floor position. Third is behind backseat. Back corner is about 2 db louder than behind back seat.

From this graph you can see the cabin gain and losses in the model and then see how it compares to my sub which is following the model to a large extent.





I also did the same model of cabin gain for the Harmon Kardon underseat woofers. The model is using a 55 HZ point of rolloff for the HKs so the cabin gain effect is lower in the 20 - 40 hz region.



The model also allows the subs to be modelled in and out of phase assuming no time offset to align phase ( Aftermarket DSP can do this ). BMW have the F3x & F8x the Harmon Kardons out of phase with each other. It's interesting in that out of phase the bass has a better response from 50 -100 hz but this is at the expense of the 20 - 40hz performance. If they are in phase the sub bass improves ( 20 -40 hz ) but the 50 - 100hz performance dips.

Guessing the sound engineers looked at the phase trade off and decided that 50 -100 hz was better for the listener considering where most bass content is.

Model with Harmon Kardon woofers in phase


Model with Harmin Kardon woofers out of phase

[Arthurrs]

Not sure how it translates into car audio, but in my work in pro audio, JBL, a company owned by Harman Kardon, would produce drivers that were polarity reverse from what many other manufacturers did. So a woofer cone from JBL would move outwards, when others would go inwards with the same DC voltage applied to the positive and negative speaker wiring terminals. Not that either was right or wrong, but when mixing systems of different makes, you had to be aware of this and deploy the systems accordingly so that they play nice together. I can't help but think these influences carried over into the car audio world, many of these conglomerates share their intellectual knowledge across different market segments. Pro audio is but a tiny blip on Harman's radar, but it's influence on other consumer market segments at Harmon were huge.

I'm sure the engineers at HK were perfectly aware of what was going on here, and obviously made some choices on how the system interacts. A trunk subwoofer was not part of this deployment, but if you knew what you're doing, a choice of crossover frequency and driver polarity will make this interaction to maximum benefit.

[Billfitz]

The JBL marking of black for positive and red for negative dates back to before red for positive and black for negative was standardized. Another bit of an oddity was Eminence for one, and perhaps others, used black for positive, red for negative on high frequency compression drivers, under the assumption that they would be used along with a woofer using a 2nd order/2nd order crossover, which in some cases required polarity reversal of the HF driver...except it didn't always, as the HF horn could result in enough phase lag of the HF driver to compensate for the nominal 180 degree phase shift between the crossover outputs. The lesson one learns early on is to never assume that the driver polarities are correct. You always test the result. I recall when I was a contributing editor to Speaker Builder Magazine, and an expensive commercial speaker submitted for a review tested really bad. It turned out that the tweeter should have been wired reverse polarity but wasn't.

[NealfromNZ]

Interesting backgound guys and that makes sense.

I also should have articulated the BMW phase a bit better. The cabin model in rhe second picture is with the Harmans is with on of the the underseat woofers out of phase as heard at the drivers seat.

Interesting enough it could explain why e90 systems have more low end compared with f30s. Suspect the e90 underseats woofers run in phase so have better low end extension.

I’ll do another cabin model tonight that has the effect of running a trunk sub and the two underseat woofers. I run my underseat woofers out of phase with my trunk sub as this was giving the smoothest transition between the two. ( sub 15-60hz acoustic with a HPF of 80hz @ 24db ) ( Morel underseats running 60 -175 acoustic with 45hz HPF and 175hz LPF @ 24db )

[Billfitz]

To be honest I didn't read your previous post thoroughly, so I didn't notice were you said the underseats are out of phase. They're not. If they were they'd cancel each other at all frequencies where they're less than 1/2 wavelength apart, and 1/2 wavelength at 40Hz is 14 feet. Also to be technically accurate 'out of phase' isn't correct terminology. Cancellations occur where the phases of two or more drivers are 180 degrees apart. When that's the result of the distance between them it only happens at one frequency, where that distance is 1/2 wavelength. When there's broadband cancellation it's because they are out of polarity, which puts their relative phases 180 degrees apart at every frequency.

[NealfromNZ]

Thanks for the correction and appreciate your knowledge

Thought I'd post a graph of typical underseat woofers running by themselves.
These are the behavior if running from 20 - 110hz assuming this is roughly the factory crossover points.

Disclaimer, measured using entry level home equipment

Asus i7 laptop running windows 10
REW audio measurement application
Calibrated Minidsp USB mic UMIK-1
Hifime UAE23HD USB DAC ES9018K2M+SA9023 ( has a nice flat frequency response )

MObridge DSP amp. Even L/R gains , no DSP correction , no time alignment

Mic position on drivers headrest , DAC plugged into BMW AUX , Volume at 50% , aux sub volume set so no clipping occur. Audio sweep from 20 -300 hz


Anyhow , these are the results for in car response just using underseat woofers.

[Arthurrs]

Thanks for posting these response curves! Very interesting! I gotta say that the low frequency extension of the Morel's is pretty impressive. If you weren't on the opposite side of the globe, I'd loan you an earthquake sub to try, but alas the shipping would be more expensive than the driver itself!! Hopefully when I do my audio system upgrades this summer I can post a response curve or two.

[Billfitz]

FWIW this is the response of my stock Hi-Fi woofers, measured nearfield: