Multiple subs and BassQ

[Roger Dressler]

Having read the Floyd Toole book and Todd Welti papers, I was intrigued to see JBL released the BassQ box embodying their multi-sub tuning method. Just in time, too. My new, first, dedicated home theater has 2 rows of 3 seats, and there was a large difference in the bass among the various seats, particularly between the two rows. I have yet to install bass traps (they are now on order), and while I'm sure they will help, I doubt they would solve the problem as well as BassQ + traps.

Having exhausted the budget on the theater and new gear, I did not want to buy several more subs in the hope that BassQ would work as claimed. So to test the BassQ theory, I first tried using the speakers I already have. The room's LF modes are between 35 and 120 Hz (11.5' x 17' x 8'), and the main L/R speakers and 4 surrounds each cover down to 30 Hz nicely. That's 6 mid-bass "subs" in effect, and then I also have a separate powered "ULF" sub that goes <20Hz. Figured I had enough bass drive capability.

The processor is crossing over all speakers at 100 Hz, Sub=yes. That gives me one audio signal carrying all the bass, which is fed into the BassQ box. The 4 outputs are then distributed to the L/R and 4 surrounds (driving the Ls/Lb with one signal, and Rs/Rb with another, thus sharing the load). The front L/R are more capable speakers so are running alone.

The diagram shows the signal path of the bass management system. It anticipates the use of an SSP-800 which has 2 sub outputs, but for now my processor has one sub out, so I use a Y splitter to drive the two paths. The resistor summers are in an aluminum box I made. Works great.



I joined this Forum to get more familiar with REW. I'm currently using XTZ, which is plug-and-play easy, but I'm now seeking to have more detailed measurements.

The following plots are 1/6 octave smoothing, and show the before/after effect. The plots show blue and white curves. The blue are the measured results. The white show how it would look if the XTZ suggested PEQ filters were added.

Fig 1 group shows the use of only the 4 "corner" speakers driven as subs. The Ls/Rs are not running. The "front" graphs show the XTZ mic moved to 3 positions across the front row of seats. The "rear" graphs show the same for the rear row of seats. You can see that BassQ helps make the two rows look more uniform.



I then added the Ls/Rs in parallel with the rears, and re-ran BassQ setup. The results are in Fig2. Pretty similar, which is good, as it means BassQ can achieve equal benefit while I get additional drive to cover the mid bass.



These plots also show that the speakers all roll off below 25 Hz. So I added a bit of a high-pass filter from an analog Rane EQ to cut off below 30 Hz, and I added my main sub with a 30Hz LPF. I did not re-run BassQ, as the main sub is not in the BassQ signal path. Fig 3 shows the overall results.



Fig3 is what I'm listening to, and it sounds very good already. The money seat went from anemic to smooth and powerful, and the results in the other 5 seats are similar. My room still has various ringing, such as at 80 Hz, causing the notch in the rear seats, so the bass traps are expected to help knock that down. I'll let you know at end of August when they're installed. For ref, here's my room layout:



I hope this, my first thread, will be of interest to y'all.

[atledreier]

What do this BassQ box actually do to the signal, then? I haven't heard of it before. Sounds like something you could achieve with most modern room correction systems these days?

[Sonnie]

So the main sub is not measured in the response?

When you get REW up and running... show us some before and after graphs. I would include the main sub because it is going to effect the response in the room. The response you see without it will most likely not look anything like it will with it.

[brucek]

The resistor summers are in an aluminum box I made. Works great.

I wonder if you may be paying a bit of a price in separation performance in your surrounds with the passive mixer.

The area of concern is the crosstalk between the Rs / Ls and the Rb / Lb channels. The Rs signal to the power amp will be seen at the input to the Rb power amp about -7dB down. Similarly for the Ls and Lb signal. That would generally eliminate any uniqueness to those signals. You might consider forgoing the bass injection to the back speakers and just leaving it to supply to the surrounds.

brucek

[Roger Dressler]

atledreier wrote:

What do this BassQ box actually do to the signal, then? I haven't heard of it before. Sounds like something you could achieve with most modern room correction systems these days?

The process is described in Todd Welti's paper "Low-Frequency Optimization Using Multiple Subwoofers" which I've seen linked somewhere. I've not seen any room correction system that can do it as they lack three things: 4 independent sub outputs, 4 microphones running simultaneously, and an algorithm that seeks to use sub timing and phase to minimize spectral variation over a large seating area.

ETA: BassQ is not an equalizer as is the case with other room correction systems. It adjusts signal timing and levels.

[Roger Dressler]

Sonnie wrote:

So the main sub is not measured in the response?

When you get REW up and running... show us some before and after graphs. I would include the main sub because it is going to effect the response in the room. The response you see without it will most likely not look anything like it will with it.

Since the one and only ULF sub is restricted to frequencies below 30 Hz in my contrivance (not BassQ's idea), BassQ cannot have any effect at those frequencies. It needs multiple (at least 2, preferably 4) LF drivers covering the given spectrum in order to achieve its effect. I've got 6 speakers covering >30 Hz, but only one <30 Hz, which is below the lowest room mode, so is not exciting modal problems anyway. It just fills in the really deep stuff.

The Fig 3 before/after plots include all 7 LF drivers in operation.

[atledreier]

The Audyssey Sound Equalizer is an 8channel system that use phase and magnitude to get as wide a listening bubble as possible. Several processors and receivers have at least dual outputs for subwoofers. And the four microphones runing simultaneously is not a very cost effective way of doing four measurements, although it does save a little time when measuring. The Audyssey Pro setup can do 32 measurement points I think. The consumer level Audyssey do 8 positions.

I'm not trying to put down the BassQ box by any means, and having multiple subwoofer capability is always a huge bonus.

[Roger Dressler]

atledreier wrote:

The Audyssey Sound Equalizer is an 8channel system that use phase and magnitude to get as wide a listening bubble as possible. Several processors and receivers have at least dual outputs for subwoofers. And the four microphones runing simultaneously is not a very cost effective way of doing four measurements, although it does save a little time when measuring. The Audyssey Pro setup can do 32 measurement points I think. The consumer level Audyssey do 8 positions.

I'm not trying to put down the BassQ box by any means, and having multiple subwoofer capability is always a huge bonus.

I don't think mic cost is an issue in ASE's favor, as it costs 2x the BassQ.

BassQ does 4 sweeps with 4 mics--for 16 measurements. I suspect the use of 4 mics can gather more information about the room, or do so with greater certainty, than one mic moved to those same 4 places, since the signals automatically represent the same time event. I have no proof of this, though. Since it's a one-time process, I don't think adding 3 additional mics, stands, and inputs (and A-Ds, etc) would have been justified only for measurement speed. It takes very little time to run the sweeps.

ASE apparently has high resolution filters intended for subwoofer EQ in one of the 8 outputs. The others have less resolution and are intended for main channel correction > bass crossover point. So to tune 4 subs, one will be using 3 of the lesser channels. If that didn't matter, there'd be no need for higher res filters in the sub channel.

Has anyone tried using ASE or another such device to drive 4 subwoofers around the room? And looked at the response uniformity across the various seats with REW? I'm new here, so have not explored all the threads.

[Roger Dressler]

brucek wrote:

I wonder if you may be paying a bit of a price in separation performance in your surrounds with the passive mixer.

The area of concern is the crosstalk between the Rs / Ls and the Rb / Lb channels. The Rs signal to the power amp will be seen at the input to the Rb power amp about -7dB down. Similarly for the Ls and Lb signal. That would generally eliminate any uniqueness to those signals. You might consider forgoing the bass injection to the back speakers and just leaving it to supply to the surrounds.

You are correct that there is some crosstalk. I went with low value resistors to keep the impedance on the outputs feeding the amps as low as possible, as it's a 12' cable run between the procs and amps.

Assuming the BassQ is a 600 ohm source as claimed, the crosstalk would calculate to be -15 dB, and that is what I measured. In the seats, even the ones closest to the surrounds, the crosstalk is inaudible since it is masked by the same signal from the adjacant speaker.

I don't expect others to follow my method of driving main speakers as ersatz subwoofers, as I suspect most folks don't have surrounds that reach 30 Hz happily, or would prefer not to occupy them with bass signals, or they want want to play bombastic soundtracks to reference levels. They should use real subs as JBL recommends. I was just trying to see if BassQ worked before investing in 2+ more subs. I'm so pleased with the results that I'll be leaving it this way.

[atledreier]

The ASE have 4 high resolution outputs and 4 lower resolution. Lower res being the same as the Audyssey MultEQ XT, high being 4x that.

As you can tell I'm sold on the way Audyssey sounds in my room and the other rooms I've tried. I'd love to try another way of doing it, but even half the ost of the ASE is beyond my budget at the moment.

The SVS AS-1 is another Audyssey option providing dual outputs at high resolution.

[Roger Dressler]

atledreier wrote:

The ASE have 4 high resolution outputs and 4 lower resolution. Lower res being the same as the Audyssey MultEQ XT, high being 4x that.

That's good to know. It wasn't obvious from the ASE page. They said "the subwoofer output" so I assumed just one.

[atledreier]

There is no 'subwoofer' output per se, there are 8 inputs labelled 1 through 8, 4 of which are higher res. So in theory you can EQ 4 2.0 systems if you wish.

[Roger Dressler]

atledreier wrote:

What do this BassQ box actually do to the signal, then? Sounds like something you could achieve with most modern room correction systems these days?

atledreier wrote:

As you can tell I'm sold on the way Audyssey sounds in my room and the other rooms I've tried. I'd love to try another way of doing it, but even half the ost of the ASE is beyond my budget at the moment.

The SVS AS-1 is another Audyssey option providing dual outputs at high resolution.

The SVS unit may be different and or more capable than the Audyssey Sub Equalizer , but they seem similar in many respects. The ASubE is designed to handle two subs, and the description states: >>The Sub Equalizer is also the first component that enables the proper integration of two subwoofers in any home theater receiver. It automatically sets the level and delay for each subwoofer and then creates a room correction solution for both subwoofers combined.<< They also publish room responses at 4 seating positions. I put each of the 4 plots in a graphics package that allowed me to extract the correction curve at several points for each seat. The corrections are essentially the same in each case. This means that whatever response variations that existed across the various seats before MultEQ was activated remain the same after it's activated. While ASE may be using multiple mic positions and clever averaging to determine the applied EQ, the net result does not reduce what Welti calls MSV (Mean Spatial Variance). The goal of Welti's method is all about reducing MSV in order to unify the response across all listeners, and that's what I have found the BassQ box is doing in my room. I think this is a clear point of distinction between BassQ and other room tuning systems.

[atledreier]

The SVS AS-1 and Audyssey Sub Eq is the same thing, I think.

I agree there is a difference in algorithm, and ultimate goal of the target curve for the two systems, but both can only apply one correction target at any one time per channel, so they are inherently the same.

[Roger Dressler]

atledreier wrote:

The SVS AS-1 and Audyssey Sub Eq is the same thing, I think.

I agree there is a difference in algorithm, and ultimate goal of the target curve for the two systems, but both can only apply one correction target at any one time per channel, so they are inherently the same.

Yes, they are the same in the sense you say--they both do what they do for the whole room all the time. The similarity ends there. Saying they both apply a target curve implies they are both EQs. BassQ is not an EQ--it applies no traditional EQ curve to the subs.

I have illustrated that Audyssey does not reduce seat-to-seat response variation. BassQ, on the other hand, does reduce response variation. These are different animals.

May I ask, have you read the Welti paper? Once you do, I'm confident you will not say the two approaches are inherently the same.

I should add that these two approaches to bass tuning are not necessarily mutually exclusive, but complementary. Since BassQ does not equalize, the resulatant room response, albeit more uniform for the various seats, is not necessarily flat. If so, MultEQ or SMS or PEQ or GEQ can still be useful in tandem with BassQ.

[Wayne A. Pflughaupt]

Saying they both apply a target curve implies they are both EQs. BassQ is not an EQ--it applies no traditional EQ curve to the subs.

If it's not an equalizer, then what is it?

Regards,
Wayne

[Roger Dressler]

Wayne A. Pflughaupt wrote:

If it's not an equalizer, then what is it?

A signal processor. From the paper: >>The third method, sound field management (SFM), can also be used for any shape room but uses only subwoofer placement and very simple signal processing ... based on in-room measurements. This paper focuses on positional optimization and SFM because of their relative ease of implementation and certain other advantages, which are outlined in this paper.<<

[brucek]

Roger Dressler wrote:

May I ask, have you read the Welti paper? Once you do, I'm confident you will not say the two approaches are inherently the same.

I wanted to read this paper, but I get a file not found result.

Do you have another source?

brucek

[Roger Dressler]

brucek wrote:

I wanted to read this paper, but I get a file not found result.

Do you have another source?

brucek

Looks like Harman pays AES on a monthly basis to allow nonmember access, and the last window expire Aug 3. Sorry. Maybe they will do it again if Sean Olive gets enough requests.

[Andysu]

First warm welcome Roger, welcome aboard mate, how’s it going no rain this week?

I’m giving this some of the hmm interested view. I leaped to Google and went straight to the JBL site and then popped over to the JBL Lansing Heritage site and I see there is thread up and running on the discussion of this product.

So how much is the JBL BassQ?

Edit: Blimey this long manual by the looks of it reading though it now.

http://manuals.harman.com/JBL/HOM/Ow...assQ_OM_LR.pdf

[Roger Dressler]

Andysu wrote:

First warm welcome Roger, welcome aboard mate, how’s it going no rain this week?

So how much is the JBL BassQ?

Thanks much for the kind welcome. After an amazing 2 nights of thunder and continuous lightning, it's all calm and cooled off again. Some of those thunder claps were stunningly deep and loud--shook the whole house. Put my humble home theater to shame.

I believe the BassQ list price is $1200.

[Andysu]

Roger Dressler wrote:

Thanks much for the kind welcome. After an amazing 2 nights of thunder and continuous lightning, it's all calm and cooled off again. Some of those thunder claps were stunningly deep and loud--shook the whole house. Put my humble home theater to shame.

I believe the BassQ list price is $1200.

Afternoon mate


Off the top of my head that would be around £600 UK for argument sakes.

That’s a lot of natural low end thunder rumbling from the sounds of it how long did the storm last for?

The last thunderstorm to rumble Bournemouth was on the night Michael Jackson, past away. Yes it is strange and it lasted for 3 hours. I have the frequency waterfalls to prove it.

Yes, yes I seem to recall the thunderstorm or sudden thunderclaps mildly impacting on the exterior of the building and reverberating though the concrete flooring that is 8 inches thick mostly the walls are cavity.

Oh, yeah nature is bigger than any sub bass speaker on the planet.



One thing that bothers me after leaning the hard-way was that most lows are naturally high in other parts of the room. Yes its not ideal but it does require less power by my reckoning.

I done a few sine wave tests playing (30Hz though to 40Hz) and average level 80db plays with less barograph display on the Alesis RA300. I figure less is more.

If I increase by +10db the level shows minimal barograph display this tells me it looks good enough, ample sufficient reasonable power with than range.

Soon as I get above it starts to drop off to the dreaded (Null Factor) that is inherently common with most rooms.
This now requires more power but hold on!

As I mentioned elsewhere in the room that signal will be at its natural highest due to where the sub is located.

Here’s a random thought I just had to test subs to see if they have The Right Stuff.

I guess the thunder radiates from above projection its low end down onto buildings cars etc. Now what about suspending a sub outside the home many, many feet up on crane yes hire a crane and chain the sub up and take it up as high as the crane can go.

Then play some sine waves or thunder sound effects on clam quite evening and see if it vibrates the house in the same manor. Yes I’m quite serious.

It would be interesting to see how far the low end radiates out.

See my point is that the sub isn’t against any surface boundary. I believe in the forums there’s a chap that tests subs outdoors. But its still against a surface boundary rather than having it suspending overhead.

I not suggesting that everyone starts sticking in 8 bolts into the joists in the ceiling and hanging the subs up in the corners of the room which would keep the floor area free and this may prefer better over the common placement that has given most users such a major headrace with this null issues.

Also the upper corner is further away less when its on the floor and in the corner.

It is interesting to see how BassQ evolves the sub bass though the different stages of multiple subs that are being added.

[maxmercy]

Roger,

I have read about Toole's SFM approach (much like using delay for 'bass steering' in large subwoofer arrays), and I'm glad it is finally available to consumers...

My questions:

Where did you place the 4 mics if you have 6 seats?

Can you post pics of your HT? And describe the equipment you are using (main speakers to 30Hz, wow)?

It looks to be very well done, much like the diagrams in Toole's book, with the semicircular diffusers, double wall, and what looks to be sound absorption for first reflection points, front and back walls...

You have a wonderful seat to seat variance...I wish I did, but I only have one sub....

JSS

[Roger Dressler]

maxmercy wrote:

Roger,

I have read about Toole's SFM approach (much like using delay for 'bass steering' in large subwoofer arrays), and I'm glad it is finally available to consumers...

My questions:

Where did you place the 4 mics if you have 6 seats?

Can you post pics of your HT? And describe the equipment you are using (main speakers to 30Hz, wow)?

It looks to be very well done, much like the diagrams in Toole's book, with the semicircular diffusers, double wall, and what looks to be sound absorption for first reflection points, front and back walls...

You have a wonderful seat to seat variance...I wish I did, but I only have one sub....

JSS

I updated the room diagram in Post 1 to show where I placed the 4 mics.

The main L/R speakers are Aerial 7B, the 4 surrounds are B&W CWM8180s in custom backboxes, slightly larger than the B&W models--the rear two are triangular to aim out of the corners. The rest of the gear is in my profile.

The attached pix show the Perf-Sorber at the first reflection side wall (the door), and diffusors near the surrounds. I don't have final beauty shots of the room with furniture as I'm awaiting the bass traps from ASC to arrive end of this month. The front ones will be 20" half-rounds that will flank the screen.

I did try to follow many of the Toole recommendations. I went with the Quest Perf-Sorbers on advice from Dennis Erskine.

The screen is motorized--shown in its lower "widescreen" position. It can roll up to show 16x9 also.