Stereo subs vs virtual mono-sub?

[klettermann]

I’m now using two subs as a summed mono bass system. They’re in opposite corners, time-aligned, gain-matched, and EQ’d together. Crossover is currently around 75 Hz LR24, but I don’t consider that fixed. It sounds great to me but obviously I don't know what I'm missing and it seems like stereo subs are fashionable.

I’ve been playing with REW Room Sim to think through the stereo-sub alternative. If I move the subs near the mains as left/right extensions, the response develops a large and irremediable null. In the current opposite-corner layout, the response has a few big modal peaks below ~80 Hz, but those are relatively easy to cut with EQ. The null is the much bigger problem. That makes me wonder whether stereo subs are sensible in this kind of setup. If the subs are placed for modal control rather than as physical extensions of the mains, summing them mono preserves the main advantage: the two subs can work together as one optimized low-frequency system.

So, the question is: below roughly 80–100 Hz, is there much practical case for stereo subs if stereo-symmetric placement creates substantially worse modal behavior? Or is mono multi-sub optimization usually the better default unless the subs can be placed near the mains without creating serious room-response problems? Any real world experiences on this topic would great. Thanks and cheers,

 

[YouDontHaveEnoughBass]

Or is mono multi-sub optimization usually the better default unless the subs can be placed near the mains without creating serious room-response problems?

Yes; the benefits of stereo bass are inconclusive to say the least and you're just getting better, more uniform bass across the room with the mono approach

Think you hit the nail on the head yourself tbh

 

[Keith_W]

Firstly, nulls that are apparent in the REW room simulator may disappear if you play with the settings a bit. Like timing, polarity, and so on. So don't simply plonk two subs in the room simulator and fail to adjust the settings. You will get a misleading picture. And whilst the simulator is reasonably accurate, it is not entirely foolproof. Is your room a perfect rectangle? No sloped ceiling? Equal absorption on all walls? No openings? Nothing beats actually placing the subs there and attempting to DSP it to perfection.

That being said, if you choose to have stereo subs, by definition you will lose the benefits of timbral accuracy. This is because you are feeding each sub uncorrelated signal. You might get them to be perfectly flat with a test signal, but the moment you play music through it, all your perfect measurements go out the window.

There are certainly claimed benefits from researchers for stereo subs, but it currently seems as if the benefits are dependent on the size of your room. The larger your room, the lower frequencies you can appreciate stereo bass. I commented on a study somewhere else on ASR - read from this post onwards where a study is cited and commented upon.

In short: if your room is of insufficient width OR if you cross your subs too low - the exercise is a waste of time.

 

[klettermann]

Thanks Keith. After a bit more thought it occurred to me that maybe with 4 subs you can do much, much better. After ;playing with Room Sim it seems this is so.(and yes, I played with everything!). Making a 2-sub left virtual sub and same on right, that either side will manage the modes AND that differences in gain on either side don't mess it up. I'm to play with this more. If Room Sim is right then maybe 4 subs in 2 arrays is the way to go. Room is perfect rectangle with exception of a small column on one side. I've done lots of measurements and found that it maps to Room Sim very well.

 

[john61ct]

Multi-Sub Optimizer (MSO): Lessons Learned, Tips & Tricks

I am in the process of making some tweaks to my 2.2 system, and wanted to use MSO. I've got an Okto Pro 8 on the way and want to run MSO for sub alignment. Is there an idiot's guide somewhere? Some basic questions: Do you measure both subs and mains or just subs (there seems some differing...

www.audiosciencereview.com

 

[klettermann]

Thanks, all. This is helpful and gets to the issue I’m trying to sort out. I agree that Room Sim is only a starting point. In my case the room is close to rectangular, apart from a small column on one side, and prior measured results have tracked Room Sim reasonably well. But I’m not treating the sim as proof. It’s mainly useful for identifying candidate geometries and avoiding obviously bad ones before moving boxes around. The distinction I’m trying to make is between three different configurations:

1. Two subs summed mono and placed for modal control.
2. Two stereo subs placed as left/right extensions of the mains.
3. Four subs configured as two channel-specific arrays: left bass to two left-side subs, right bass to two right-side subs.

The first approach is what I’m using now. It works well: opposite-corner subs, summed mono, time-aligned, gain-matched, EQ’d together, crossed around 75 Hz LR24. The response has some modal peaks, but they are mostly cuttable. No catastrophic null. The second approach is where I see the problem. If the subs are placed near the mains as stereo extensions, the room response gets much worse in the sim. Playing with delay, polarity, gain, and crossover helps only so much. The placement that preserves stereo geometry does not appear to be the placement that best manages the room. The third approach is the one I’m now wondering about. With four subs, each channel could have its own small modal-control array:

Left bass → front-left sub + rear/side-left sub
Right bass → front-right sub + rear/side-right sub

In Room Sim, that looks much more promising. Each channel can be adjusted to reduce major nulls, while still preserving channel-specific LF information. Peaks would then be handled with EQ. The interesting part is that the left and right arrays seem to remain reasonably well behaved even when their relative levels differ, which is important if the bass content is not mono.

So I guess the question is not really “are stereo subs fashionable?” or “should subs always be mono?” It’s whether a left-array/right-array approach is acoustically valid: using two subs per channel to get some of the modal-control advantage of multi-sub optimization while not summing all LF content to mono.

I also take the point about room size and crossover frequency. My room is about 22' × 12.2' × 7.2', with the listening position centered and roughly 8' from the rear wall. Current crossover is around 75 Hz, but that is not fixed. If the stereo/spatial benefit below that range is marginal in a room this size, then the added complexity may not be justified even if the configuration can be made to measure well.

The MSO suggestion is also relevant. I need to think through whether MSO can be applied cleanly to this problem. The usual mono multi-sub case is straightforward enough conceptually. The more unusual question is whether to optimize the left pair and right pair separately, then check mono/stereo summation behavior, or whether there is a better way to define the optimization objective.

I’m not planning to buy 2 more subs tomorrow. At this point I’m trying to determine whether the concept makes sense before treating it as anything more than an interesting branch. Has anyone here actually measured or implemented a two-channel setup with two subs per channel, optimized as left and right modal-control arrays rather than as simple near-main stereo subs? Cheers,

 

[Keith_W]

So I guess the question is not really “are stereo subs fashionable?” or “should subs always be mono?” It’s whether a left-array/right-array approach is acoustically valid: using two subs per channel to get some of the modal-control advantage of multi-sub optimization while not summing all LF content to mono.

OK let's say you have this situation where you have two "systems" - on the left, one speaker and two subs. The speaker is in front of you, and the two subs are deployed along the side wall. The "right" system is a mirror image of the left. You take the left system, and you DSP the system until it is perfectly flat. I'm not sure if that is possible, but let's say it is. Then you do the same for the right. Now you have two systems that are individually, flat down to 20Hz.

The next step is to make the "left" system acoustically sum with the "right" system so that the result is flat. Once again, i'm not sure if that is possible, but we'll again say that it is.

Now you feed the "left" and "right" system different signals, and get them to play at the same time. What do you think will happen?

That's why I said that by definition, setting your subs up in stereo and feeding them uncorrelated signal WILL result in a non-flat response. I don't think there is a way around it.

 

[john61ct]

My general plan is to try to set up the ability to change channel routing on the fly, perhaps also AB switching between filter sets.

For me I think 5 subs is minimum for such experimenting because Center, but maybe four is enough flexibility.

Two colocated L/R, ability to switch mono vs stereo.

Center is already mono, and IMO should be able to put out a bit more SPL than L/R for film nights.

So mono #2 can be freely placed and possibly with Center managed by MSO for "smoothing" room issues - small MLP or wider? big Q.

I think #3 would make things a bit easier in that configuration.

Note the L/R pair might sound better as mono, and/or be easier to manage as part of MSO

in which case they can be freely placed, and maybe five is overkill.

...

Next wrinkle - all the same size? I need MBM Couplers in Stereo L/R, if they can do SPL to say 50Hz then more need for say 20-80Hz, split that job up?

One Big Cahuna for SPL way deep, smaller ones for the room wrangling?

Different crossovers to give MSO variety to work with?

Then a config to test Stereo Only for content possibly enabling envelopment

or maybe trying to artificially create something spacious...

collocated with surround R/L ?

So, not trying to find "the one" arrangement, but maybe one day something will really stand out

for THAT room 8-D

 

[janbth]

... With four subs, each channel could have its own small modal-control array ...

This "four subs set up as stereo" configuration would control the axial modes in the front-back direction, but not the ones in the left-right direction. Of course, you could be lucky with your listening position and not be bothered by the left-right modes, but you could also be unlucky...

 

[klettermann]

@Keith_W Thanks, yeah, I see your point. I did a lot more playing with Room Sim and it bears it out FWIW. With stereo bass there is no single corrected LF response in the same sense as with summed mono bass, because the two channel signals are not necessarily correlated. A mono sweep, L-only sweep, R-only sweep, and arbitrary stereo music content would all likely all excite the room differently. So even if each side can be made individually smooth, that does not guarantee a stable summed response under all program material.

That seems to be the core tradeoff. A mono multisub system gives up channel-specific LF information but allows the subs to be optimized as one room-control system. A stereo sub system preserves channel-specific LF content, but the room response becomes more content-dependent. Maybe the practical question is not whether the stereo-array version can be made “flat” in an absolute sense. Rather, it's whether the remaining content-dependent variation is smaller or less objectionable than the modal problems from ordinary two-sub stereo placement, and then whether any preserved LF stereo/spatial information below ~75–100 Hz is perceptually worth that tradeoff.

@janbth , yes, the point about axial modes also seems important. A left-array/right-array layout may help with front-back modes for each channel, but it gives up some freedom to optimize the left-right width modes compared with a mono multisub layout. That may be a major strike against the idea in a 12.2' wide room.

So, at least for now, I’m seeing the four-sub stereo-array idea as an appealing idea but a very uncertain outcome, not enough to justify the expense. This is too big for the "just try it" approach. Thanks again and cheers,

 

[klettermann]

A couple further practical points....

1. In my current summed-mono two-sub setup, the bass subjectively appears to come entirely from the front of the room. The rear sub is fairly close to the MLP, but I don’t perceive it as a localizable rear source at all. That seems relevant to the mono-versus-stereo question. If the distributed mono sub system is well integrated, the subs can function as modal/pressure-field control devices while the apparent bass source remains attached to the mains/front stage. Presumably the Maggies are carrying enough localization information above the crossover, while the subs supply the low-frequency foundation below it.

2. The whole DAC, source selection and routing setup becomes kind of mind bending unless I greatly changed all my current stuff.

So in my case, one of the possible objections to summed mono bass — bass detachment or collapse away from the front image — doesn't seem to an issue. That raises the bar for any stereo-sub alternative. It would need to preserve useful channel-specific LF information while also beating, or at least not degrading, a mono setup that already sounds front-anchored. Cheers,

 

[john61ct]

Yes I used to think the ability to switch between "modes" would only be needed during the system building / testing stage.

I am now hoping to devise a way to make it built into the infrastructure permanently.

The three choices of "enable bass management within each bed channel" vs "all <80Hz is mono" vs "mix the two"

I think also need use case modes, single MLP optimisation vs "wider area smoothing"

critical listening to music

family film night

dance party

maybe background music...

...

And in my case, different system subsets

depending on the living pod / packing space and Ah energy available for portability

 

[Flaesh]

the bass subjectively appears to come entirely from the front of the room. The rear sub is fairly close to the MLP, but I don’t perceive it as a localizable rear source at all.

Is that good? Or not?

A blue Mercedes can be painted a matte pale black. This is an example of a mismatch of personal preferences.

 

[klettermann]

Is that good? Or not?

A blue Mercedes can be painted a matte pale black. This is an example of a mismatch of personal preferences.

Yeah, I think it's good. I don't think Id want it coming from anyplace else in a 2ch stereo system. I'm not sure what the mercedes color has to do with it or what's mismatched. Blue metallic was very popular at the time.