[kaigoss69]

This is only for those who have a trunk sub!

I will explain later, but I would like you to try this and report back:

1. On the amp that drives your underseat woofers, engage the low-pass filter.

2. If X is the MS-8 crossover point between the underseats and the front speakers, set the amp's LP filter to X + 40Hz, if the amp's crossover slope is 12dB/oct, or X + 80Hz, if the slope is 24dB/oct. What you're trying to do is cut off the sweeps going to the underseat woofers, as closely as possible after MS-8's crossover point.

3. Run a new calibration and listen for the difference in initial midbass response compared to prior calibrations. Post your results here!

[kaigoss69]

Oh, and if your MS-8 crossover point is over 200Hz, try to reduce it as much as possible. You will want that amp LP crossover to have a significant impact before 300Hz!!!

[bmw325i]

Gonna try this now. I wonder if it would be a good idea to set a high pass for the doors and center 40-80hz under what the ms8 is set to.

[kaigoss69]

Originally Posted by bmw325i

Gonna try this now. I wonder if it would be a good idea to set a high pass for the doors and center 40-80hz under what the ms8 is set to.

Probably not necessary, the only issue MS-8 has in our cars is the woofers under the seats.

[bmw325i]

I did notice the midbass is improved but the mids and highs sound much better. Not sure if it was because of this or the calibration but I'm going to calibrate it again because I did a quick calibration and didn't level match the rears. I might try the high pass on the doors just to see what it does.

[kaigoss69]

Originally Posted by bmw325i

I did notice the midbass is improved but the mids and highs sound much better. Not sure if it was because of this or the calibration but I'm going to calibrate it again because I did a quick calibration and didn't level match the rears. I might try the high pass on the doors just to see what it does.

Keep in mind that MS-8's sweeps go beyond the crossover points, so it may actually be detrimental to do this depending on where you set the HP. This "experiment" is to see what the LP filter will do in different cars with different set-ups. I suspect (and hope) this could be a universal fix but we will see.

BTW I worked this out together with Andy Wehmeyer, so be rest assured it is not a wild goose chase!

[mob17]

So is this experiment only for people with a trunk sub? What's the reasoning behind that?

[kaigoss69]

Originally Posted by mob17

So is this experiment only for people with a trunk sub? What's the reasoning behind that?

If you don't have a trunk sub you should not have a problem with midbass!

[mob17]

Originally Posted by kaigoss69

If you don't have a trunk sub you should not have a problem with midbass!

Oops

Guess I have to go back to my tweaking

[kaigoss69]

Come on, need more guinea pigs!

Also, if you have the possibility, open the ski pass or fold down the seats during calibration!

[bmw325i]

I had the seats down for calibration. I'm going to mess with it some more today, but I did notice a significant improvement.

[kaigoss69]

Problem: When using a trunk sub on its own MS-8 channel, the mid-bass level after calibration is extremely low.

I have been trying to get to the bottom of this issue ever since I first installed my MS-8. From prior discussions with Andy Wehmeyer, JBL Global Product Manager, I was convinced the cause was mostly due to differences in level between the underseat woofers and the trunk sub, as the problem only occurred when a trunk sub was installed. MS-8 sweeps each speaker during calibration to determine the frequency responses, and it sets levels based on the results, and uses EQ to smoothen out the final curve. Levels are chosen based on the peak (loudest) frequency for each speaker, or speaker set (the sub is swept alone, the fronts are swept together with the underseats!). So my theory was that a hump or peak in the frequency response of either the sub or the woofers would cause a large difference in levels between the two, and that the processor would then simply run out of headroom not being able to apply enough EQ to smoothen out the final curve. For example, let's say the sub had a large peak of 100dB at 80Hz, and the underseats would have their peak of 80dB at 150Hz so then the initial “leveling” the MS-8 would perform would require 20 dB of processing headroom. Then, the levels of the center or sides, if they were below 80dB, would have caused an even greater spread. Additionally, any “dips” in the responses of the speakers would also eat up additional processing power. Since the processor only has a finite amount of headroom available, let's say for this example it is 20dB, it is easy to see that it will run out of steam and not be able to fully form the target response curve. In the processing order, it appears the woofers come last, and that’s why they get left behind.

So I thought I'd have to get to the bottom of this. I got in touch with Andy and found out the exact frequency ranges MS-8 "looks for" when it sets levels. For the sub, that range is 50 - 80Hz. For the mids, which includes the underseat woofers, the range is 80 - 300 Hz. To measure, I used my RadioShack SPL meter, a CD with test tones, and I turned processing to "defeat" in the MS-8 menu, in order to get an unequalized signal to the speakers. I turned the MS-8 vol. to -10, HU vol. to 45, and amp gains to 12 o'clock. Here are the results:



I measured both the trunk corner location, as well as with the sub in the back seat since I had previously thought the back seat location would yield a smoother response. Obviously, that is not the case so just ignore that curve.

What is obvious here is that the response curve of the sub is actually fairly flat, so no smoking gun here!

On to the underseats:




At first glance, nothing to be concerned about, since these speakers run bandpassed between 60 and 160Hz.... BUT, MS-8 sweeps go out to 300Hz, and according to Andy W. even if the internal crossover is set for 160Hz, the processor will still look at the entire sweep range (80 - 300Hz) and pick the levels based on the highest peak.

So, MS-8 sees the peak, and lowers the level of the woofers accordingly. But during EQ corrections, it would have to apply boost in order to bring all those frequencies in the bandpass range back up to the level of the peak at 300Hz. That’s where it runs out of steam…

That’s why the LP filter helps, because it eliminates (or at least reduces) the peak, and MS-8 can do what it is supposed to do.

What remains to be seen is that the peak is about the same in every e9x BMW, due to the location of the woofers, rather than due to the specific electrical and mechanical properties of the woofers. This is where YOU come in, hopefully proving me and Andy right, and giving you a much improved midbass response in your car.

[jeffb335]

confirmed.....

Setup: E92, trunk sub, jehnerts under the seats on their own amp, bandpassed at 60hz and 160hz, 24db.

As instructed I set the jehnert's amp LPF to around 220-240hz, ran the speaker setup and the the calibration (took me a while as I had forgotten my speaker configuration) and the results do indeed back up your hypothesis.

It sounded pretty good on the first try, no obvious midbass hole and the sub level was solid without having to resort to all the "tricks" that I had done previously to get a good calibration/sound. The midbass is probably still a little lower than I prefer in a car (to overcome road/exhaust noise) but it wasnt entirely missing......

I did a few more calibrations at different levels and with my amp gains up and down a bit and the results were fairly consistent......backs up your idea about headroom as well, without that big peak in there the MS-8 seems to have delivered much more consistent results. Probably sounds 95% as good as the setup that I spent months trying to get right.

....in other words, the MS-8 did what it was supposed to do.

Kaigoss, Ill play around with it some more and lets get some more data points to figure out the details but I think you and Andy may finally be close to figuring this out.

[bmw325i]

Kaigoss did you check to see the difference of how it graphs out without the processing defeated, both before and after the lpf?

[kaigoss69]

Originally Posted by bmw325i

I had the seats down for calibration. I'm going to mess with it some more today, but I did notice a significant improvement.

Great!

Originally Posted by jeffb335

confirmed.....

Setup: E92, trunk sub, jehnerts under the seats on their own amp, bandpassed at 60hz and 160hz, 24db.

As instructed I set the jehnert's amp LPF to around 220-240hz, ran the speaker setup and the the calibration (took me a while as I had forgotten my speaker configuration) and the results do indeed back up your hypothesis.

It sounded pretty good on the first try, no obvious midbass hole and the sub level was solid without having to resort to all the "tricks" that I had done previously to get a good calibration/sound. The midbass is probably still a little lower than I prefer in a car (to overcome road/exhaust noise) but it wasnt entirely missing......

I did a few more calibrations at different levels and with my amp gains up and down a bit and the results were fairly consistent......backs up your idea about headroom as well, without that big peak in there the MS-8 seems to have delivered much more consistent results. Probably sounds 95% as good as the setup that I spent months trying to get right.

....in other words, the MS-8 did what it was supposed to do.

Kaigoss, Ill play around with it some more and lets get some more data points to figure out the details but I think you and Andy may finally be close to figuring this out.

Thank you sir! When you fold down the seats, do you notice a better integration of the sub, and less boominess? If not, what is the improvement?

Originally Posted by bmw325i

Kaigoss did you check to see the difference of how it graphs out without the processing defeated, both before and after the lpf?

No I did not. I may do that after I get back this weekend.

[jeffb335]

After trying a bunch of different configurations I have some more results.

Higher sub crossover points didnt work in my setup.....I tried 80hz, 24db and couldn't get a good blend although it was better when I calibrated with the back seat down.

I ended up at 65hz, 24db, seat up. At that crossover point, midbass and sub were solid and well blended, although sub was a little strong, fine for driving, a little bit high for critical listening.


I spent some time listening to some music to sort out the bottom few octaves, a couple tracks I use often for this is Sarah McLachlan's "World on Fire" and Joan Osborn's "right hand man", the bass line in those songs goes from the mid speakers (stock L7 in my car), through the underseat woofers, and into the sub, all in one riff, All with drums at the same time, great for seeing how the speakers blend and transition from one to another (or dont), and define the bass separate from the drum (or dont). With the setup I noted above, the midbass to sub trasitions are great, the mid to midbass integration is good....but could be better, as the notes from the bass go higher their relative volume seems to drop off a lot. I went home and listened to these tracks on my home system and the bass volume is much more consistent as the notes go higher,

So.....

I suspect my system might drop off a little in the 200hz range, maybe due to all the crossover stuff I now have going on in that range, maybe something else. The underseats seem to be doing their job, it seems like the door mids might be rolling off a bit at the bottom.....or.....maybe its using the center channel for that part?! Bass line is probably mono so maybe the center is trying to handle it and just doesnt have the output. If so maybe it just cant go down to 160hz without rolling off?

OR

If I recall, didnt Andy say that the target curve was boosted below 200hz? Could just be that Im hearing the roll off above 200hz? Ill need to do some more experimenting to get it sorted out.

By the way, what I am describing here is subtle, nothing big, and you have to look for it and have the right track to notice, so I think Kaigoss and Andy are going in the right direction but there might be some subtleties left to figure out. In the interest of this thread its worth noting. Overall though, my system now sounds as good as ever, and I didnt have resort to any "tricks" or "workarounds" to get it there.....but I DID have to try a bunch of different configurations before I found the one that worked best.

[jeffb335]

Update:

Using my ears, the EQ, and a simple iphone RTA app, I isolated my missing frequency to a pretty narrow band centered at 160hz .......my MS-8 crossover frequency....hhhmmmmm.....

in a traditional system this would mean that my woofers are out of phase with my mids....but doesnt the MS-8 compensate for that when it runs the first time alignment sweep since the woofers are on their own channels?

Also I was able to boost it using the EQ, if it was a true phase issue I wouldnt have been able to do that.....

But I swapped phase and its much improved but still a few db down at 160.....oh well, close enough!

Maybe it was a bad calibration, it can be a little noisy around my office where I was calibrating this morning.

[kaigoss69]

Originally Posted by jeffb335

Update:

Using my ears, the EQ, and a simple iphone RTA app, I isolated my missing frequency to a pretty narrow band centered at 160hz .......my MS-8 crossover frequency....hhhmmmmm.....

in a traditional system this would mean that my woofers are out of phase with my mids....but doesnt the MS-8 compensate for that when it runs the first time alignment sweep since the woofers are on their own channels?

Also I was able to boost it using the EQ, if it was a true phase issue I wouldnt have been able to do that.....

But I swapped phase and its much improved but still a few db down at 160.....oh well, close enough!

Maybe it was a bad calibration, it can be a little noisy around my office where I was calibrating this morning.

Jeff, to answer your question on the bass boost target range, here a quote from Andy: "The target curve includes 9dB of boost below 60Hz with a smooth transition to 160 Hz".

Here are the measurements I did last weekend with processing disabled, and I believe the crossover at the time was 160Hz:



Seems like the same kind of dip you have, but of course no processing (EQ) has been applied. I would guess that if you have a dip there, MS-8 did run out of steam before it could make that correction. You may want to experiment a bit more with sub level in relation to midbass.

You may also want to try a 12dB slope between midbass and sub. That's what has sounded best for me up to this point. Both 80 and 60 Hz sounded pretty darn good...

[kaigoss69]

One more thing, I was thinking that it may be better to engage the LP filter only after the first set of sweeps, since the short sweep may not allow for correct TA. Just a thought...

[jeffb335]

aahhhh.....interesting! Your curve looks pretty similar to what I am getting, yours is just more extreme, so it looks like we are probably just dealing with some acoustic characteristic of the car. Ill do some more tweaking tomorrow and we will see where I end up. Right now Im pretty satisfied with the way it sounds........but it could always be better!

[taibanl]

Theres another possibility: The midrange may be suffering from mechanical rolloff @160hz. So the crossover filters while theoretically are summed to flat (which they should at 24dB) there is still missing spl.

I wonder if you would see the same results @200hz?

[jeffb335]

Right, I mentioned that a few posts up, could be mechanical roll off, especially since its mostly just the center trying to handle those frequencies, plus the way its physicaly mounted in the dash isn't all that conducive to low frequency production anyway....but keep in mind I'm only chasing a few db at this point, other than this one little dip it's pretty close to right on. I

don't want my experimenting to become a thread jack though.... But I think it's still relevant since Kaigoss's idea turned the huge midbass hole into only a few db dip at 160hz.

More experimenting is in order......down the MS-8 rabbit hole I go.....again...