Not exactly.The thing I learned with modes is....they are ALL bad, period.
If you don't think so, just take your gear outside and hear how much better the bass sounds, (and really the whole dang spectrum when devoid of reflections).
So multiple sub locations are really just trying to spread the badness out evenly
Outside all modes can radiate.
Inside, not so. What modes radiate depends on the size of the room. The room filters out large amounts of LF modes and so LF response is lumpy due to standing waves, cancellations, etc. Small room bad, big room better.
The idea with multiple subs is to fill in the missing modes until things start to sound natural.
The original OP was concerned about "impact". That's easy: EQ a bump @ 80 HZ😀
On the original subject, multiple drivers vs single large driver (or reasonable variations) - I have thoughts, maybe not full answers, but thoughts on what's up.
If there were multiple drivers of exactly the same performance characteristics, they'd function together as one larger driver. Unfortunately they're not really identical, so each one of the multiple small drivers can and likely do contribute to awkward and unaccounted differences among the group, and in some cases (likely cases) they will fight against each other.
A quick example: take 4 drivers and note that while they're all rated at 32 hz resonance, they will each by slightly different, closely around 32 hz. When one driver is at it's most resonant peak, the others might not be quite there yet, and that means one driver is pushing on the others with slightly more force - via the coupling of the airspace behind them. The other drivers are partially responding to that as if they were passive radiators. That is not to say they are completely passive, but that some of the energy of that one driver at it's peak is effecting forces on the others in a way not considered in it's design. This effect may alter the sound of the overall result, and this point in frequency may be distributed over several close frequencies where each driver hit's it's peak at a different point.
The same problem comes into play when one driver is slightly more efficient than the others, or it's Q is slightly different than the others.
If a single driver did this, we'd expect to visualize the cone of a single driver vibrating with ripples of waves over the cone, with parts of the cone not contributing completely in the force of air in one direction as one unit.
To put this in another context, imagine if you disconnected all but one driver in the cabinet of a multiple driver construction. All of the other drivers would function as passive drones, would they not?
Now, even when you reconnect them, and they all have drive forces in unison, is there not some force conveyed through the common airspace behind them which act AGAINST the goal of pushing air in the front of the cabinet? There ought to be, and it ought to be too complex to control.
So, what if each driver had it's own airspace and the cabinet were one unit, such that each driver has no effect on it's siblings? That would reduce this particular influence, so now we focus on what happens in the front of the system.
In a typical single, large cone woofer, a common volume of air is "cupped" in driver's cone, pushed (forward or back doesn't really matter as frequency exceeds, say, 10hz or even 5 hz) outward from the cabinet as a wave. The force is applied with this symmetrical shape, and though the cone surface is at an angle relative to the motion of the piston, most of the sound force is radiated forward and outward as if in a cone of force within the air, which itself may not have straight sides or well defined edges.
Now, imagine that same thing happening from an array of drivers - say a line of 4. The force applied to the air will be 4 separate waves, each will overlap in the space in front of the cabinet. The distance between the listener and the wave sources will differ from each driver, and the time each wave's peaks and crests reach the ear may differ, though slightly. The wave forms are long; 22.7 feet at 50hz, over 45 feet at 25 hz. What really matters, though, is if waves combine at different phases (they cancel each other when they do). At 50 hz, for example, two waves would be completely opposite polarities if they're out of phase 180 degrees, which at 50 hz would be about 136 inches. At 90 degrees out of phase there can still be some attenuation, which is 68 inches, and at closer phase alignments small levels of attenuation can happen, perhaps within 3 dB, but enough to measure, even at about 20 inches in the 50hz region. Lower frequencies are larger, but this issue of alignment MIGHT be recognizable in other ways. As the waves cross, and the energy of the force of sound is altered by differences in timing, a "beating" can be produced - harmonics that are the result of these phase misalignments passing over each other. The beating can occasionally be louder than the expected attenuation effect by cancellation alone, and they can occur at odd frequencies not related to the music signal.
It would take more serious study, I'm sure, and probably already has - but I suspect the results could be surprising. The more drivers involved, the more complex the potential interference might be, and therefore the more pronounced the negative impact may be.
This all ignores the potential electrical influence each driver might have on it's partners if they're connected to the same amplifier. Each coil is, after all, an inductor moving in a magnetic field around a hunk of metal...which could saturate, if I understand the problem there. What electronic effects could we expect perturbing the signal, magnified by the multiple of drivers involved? I'm not sure, but I'd think it just enough that critical listening MIGHT even sense some of the effect, especially if that effect may be concentrated in some music content more than others.
If there were multiple drivers of exactly the same performance characteristics, they'd function together as one larger driver. Unfortunately they're not really identical, so each one of the multiple small drivers can and likely do contribute to awkward and unaccounted differences among the group, and in some cases (likely cases) they will fight against each other.
A quick example: take 4 drivers and note that while they're all rated at 32 hz resonance, they will each by slightly different, closely around 32 hz. When one driver is at it's most resonant peak, the others might not be quite there yet, and that means one driver is pushing on the others with slightly more force - via the coupling of the airspace behind them. The other drivers are partially responding to that as if they were passive radiators. That is not to say they are completely passive, but that some of the energy of that one driver at it's peak is effecting forces on the others in a way not considered in it's design. This effect may alter the sound of the overall result, and this point in frequency may be distributed over several close frequencies where each driver hit's it's peak at a different point.
The same problem comes into play when one driver is slightly more efficient than the others, or it's Q is slightly different than the others.
If a single driver did this, we'd expect to visualize the cone of a single driver vibrating with ripples of waves over the cone, with parts of the cone not contributing completely in the force of air in one direction as one unit.
To put this in another context, imagine if you disconnected all but one driver in the cabinet of a multiple driver construction. All of the other drivers would function as passive drones, would they not?
Now, even when you reconnect them, and they all have drive forces in unison, is there not some force conveyed through the common airspace behind them which act AGAINST the goal of pushing air in the front of the cabinet? There ought to be, and it ought to be too complex to control.
So, what if each driver had it's own airspace and the cabinet were one unit, such that each driver has no effect on it's siblings? That would reduce this particular influence, so now we focus on what happens in the front of the system.
In a typical single, large cone woofer, a common volume of air is "cupped" in driver's cone, pushed (forward or back doesn't really matter as frequency exceeds, say, 10hz or even 5 hz) outward from the cabinet as a wave. The force is applied with this symmetrical shape, and though the cone surface is at an angle relative to the motion of the piston, most of the sound force is radiated forward and outward as if in a cone of force within the air, which itself may not have straight sides or well defined edges.
Now, imagine that same thing happening from an array of drivers - say a line of 4. The force applied to the air will be 4 separate waves, each will overlap in the space in front of the cabinet. The distance between the listener and the wave sources will differ from each driver, and the time each wave's peaks and crests reach the ear may differ, though slightly. The wave forms are long; 22.7 feet at 50hz, over 45 feet at 25 hz. What really matters, though, is if waves combine at different phases (they cancel each other when they do). At 50 hz, for example, two waves would be completely opposite polarities if they're out of phase 180 degrees, which at 50 hz would be about 136 inches. At 90 degrees out of phase there can still be some attenuation, which is 68 inches, and at closer phase alignments small levels of attenuation can happen, perhaps within 3 dB, but enough to measure, even at about 20 inches in the 50hz region. Lower frequencies are larger, but this issue of alignment MIGHT be recognizable in other ways. As the waves cross, and the energy of the force of sound is altered by differences in timing, a "beating" can be produced - harmonics that are the result of these phase misalignments passing over each other. The beating can occasionally be louder than the expected attenuation effect by cancellation alone, and they can occur at odd frequencies not related to the music signal.
It would take more serious study, I'm sure, and probably already has - but I suspect the results could be surprising. The more drivers involved, the more complex the potential interference might be, and therefore the more pronounced the negative impact may be.
This all ignores the potential electrical influence each driver might have on it's partners if they're connected to the same amplifier. Each coil is, after all, an inductor moving in a magnetic field around a hunk of metal...which could saturate, if I understand the problem there. What electronic effects could we expect perturbing the signal, magnified by the multiple of drivers involved? I'm not sure, but I'd think it just enough that critical listening MIGHT even sense some of the effect, especially if that effect may be concentrated in some music content more than others.
Hi Frank,
Yep, that's easier!
I am always amazed how much better speakers sound when moved outside.
It shows me how utterly trivial many of my "gear upgrades" are, when compared to the improvements made by eliminating troublesome interactions between speakers and rooms.
Yep, that's easier!
I am always amazed how much better speakers sound when moved outside.
It shows me how utterly trivial many of my "gear upgrades" are, when compared to the improvements made by eliminating troublesome interactions between speakers and rooms.
That's it! I think you've discovered something important!
Natural sound! Organic music! Non-GMO frequencies!
Ben
Nic,
1) Yes, though not the same as drone cones that are designed as such.
2) To some degree, though the measurable difference between a single shared enclosure and a divided enclosure is rather slight.
3) Distortion is reduced, and the additional bracing provided by the bracing reduces resonace within the pass band.
4) In 45 years of listening I have never heard anything like you describe, although in a controlled experiment I could clearly hear the sub-harmonic beat frequency tone between ultrasonic frequencies.
5) Speakers also work as microphones, so can have some effect on others if damping factor is unusually low. In practical terms, if using any gear that falls even remotely into the "normal" range, multiple low frequency drivers perform so similar to a single driver with the same parameters as to be of no concern.
Multiple small subs can have the same impact as a single large sub, but a single large sub can't be spread around a room to mitigate room problems, and outdoors a single sub can't be co-located with stereo mains, so coherency (impact) in the crossover region is reduced.
In sound reinforcement systems, other than in very small set ups, multiple subs are almost always preferable.
In any "sub" decision, the advantages must be weighed against the disadvantages. In 3 out of 4 of my home listening areas I use single subs, but in the fourth area (a two room shop) using four subs and eight main speakers is preferable.
Art
Mark, there is a room modes calculator here:
hunecke.de | Room Eigenmodes Calculator
It will show you the first 20 modes.
It has a default room size entered.
Look at the frequency range of the modes for that default size.
Then max out the size of the room, (I think that's 10x10x5 m.), and you'll find the same number of modes but in a smaller frequency range.
Therefore,
The larger the room is, then the more low frequency modes it will have.😱
Now suppose the room is the size of all outdoors. Then it should have a mode for every frequency.😀
And that's a really smooth response, I think.😉
This calculator will let you put in bigger dimensions:
http://www.bobgolds.com/Mode/RoomModes.htm
Multiple subs and rooms was done to death in a recent thread by some really sharp folk. I'll dig it up.
Hi,
Here is a measurement thread on a Geddes-style multisub setup:
http://www.diyaudio.com/forums/subw...ultisub-setup-temporary-rental-apartment.html
Regards,
Here is a measurement thread on a Geddes-style multisub setup:
http://www.diyaudio.com/forums/subw...ultisub-setup-temporary-rental-apartment.html
Regards,
Thanks Frank,
I get the sense folks have taken Gedde's approach at multiple subs, exciting as many modes as possible, as a way of achieving smooth bass.
And I agree with that approach if you are looking for an averaged smoothness of frequency response, where response is blind to phase, timing, or localization with mains in x-over region.
I've tried the multiple sub approach and can't say I really liked the sound. Multiple subs kinda did to bass, what Bose 901's do to everything....make a big ball of reflections kinda sound.
In the large room I mentioned building, I had 4 Labhorns, a Meyer 650-P (dual 18"), and a pair of Meyer MTS4a's s with their 15" LF and 18" sub drivers. An insane amount of bass to play with!
I put those dang things everywhere, all over the place. Measured and listened my brains out...sine waves, noise, and music. Even flew the 280lb Meyer's, and the 650-P from the ceiling. Spread the labhorns out, corner stacked them (that WAS crazy, blurred your vision while rattling my room apart). IOW, tried alot or combos/placements.
Yes, I could get sine wave response to even out when walking around the room with distributed placements, but music would lose it's imaging and never had the transient response or the impact of just 1 or 2 subs in play. BTW, used 80hz as x-over point
And I have to say, I really don't buy the non directional bass stuff until you are safely below 80hz, maybe even 70hz.
My favorite setup became a simple stereo setup...a MTS4a on top of a pair of Labhorn's, each side acting as a single stereo channel. Power and imaging....albeit with some frequency irregularity depending on listening position..
Now take those stacks outside, and holy moly, best sound I've heard...and all my neighbors too 😀
As far as modes, yep I know those calculators well.....again, the room was carefully designed for modal dispersion. A larger room has more modes only because longer wavelengths can resonate in it. Exciting more modes, or modal dispersion, just means not letting modes pile up on each other from having room dimensions that are integral multiples of each other.
Really, all modes are bad compared to no modes.
I keep saying it to all my audio buds, if you want to have you ears and eyes opened, take the time to set up outdoors sometime 🙂
Mark,
Mixing different subs with around a ten foot path length difference and quite different phase response will hose up transient response in the crossover region (unless corrected with FIR filters) indoors or outside 😉.
Art
Art, a big yes...as I learned. I had a 4x8 matrix mixer that allowed full processing on all outputs, but so what. Measuring phase indoors was Way over my head. I made one muddy mess after another. Outside much easier 😉
Are you finding FIR filters worth the effort?
Not yet, won't purchase a DSP with FIR capability until sometime early next year.
Pallas' excellent report (in link kindly provided by tb46) shows the value of multiple subs (and there's a nice published literature supporting the idea) and the value of heterogeneity of locations and speaker types.
If you think statistically, it is apparent that just adding a variety of subs is likely to be beneficial even without trick EQ, time-alignment, or post-hoc measurements.
But there's something true (besides the usual subjective bias) to what mark100 says about the sound sounding diffuse. Of course, as adults we face trade-offs all the time and in audio, need to decide what goals matter more, for example, full smooth bass or whatever is the opposite of diffuse.
Mark - did you use time alignment?
With some DSP gear, you have enough channels to time-align a half-dozen subs.... at least time-align to one spot in the room.
Of all the subjective errors people make, I think spatial localization is the dirtiest. Paul Klipsch used to insist his dealers put scrim curtains between the speakers and the listeners. For decades, I used a 140 Hz, 24 dB/8ave crossover (thank you Bell Labs for your excellent stock room) and put my low-distortion sub in locations which otherwise would maximize noticing where they were. Can't recall ever having my attention drawn to the sub on music.
Ben
If you think statistically, it is apparent that just adding a variety of subs is likely to be beneficial even without trick EQ, time-alignment, or post-hoc measurements.
But there's something true (besides the usual subjective bias) to what mark100 says about the sound sounding diffuse. Of course, as adults we face trade-offs all the time and in audio, need to decide what goals matter more, for example, full smooth bass or whatever is the opposite of diffuse.
Mark - did you use time alignment?
With some DSP gear, you have enough channels to time-align a half-dozen subs.... at least time-align to one spot in the room.
Of all the subjective errors people make, I think spatial localization is the dirtiest. Paul Klipsch used to insist his dealers put scrim curtains between the speakers and the listeners. For decades, I used a 140 Hz, 24 dB/8ave crossover (thank you Bell Labs for your excellent stock room) and put my low-distortion sub in locations which otherwise would maximize noticing where they were. Can't recall ever having my attention drawn to the sub on music.
Ben
Last edited:
Ben, yes, time alignment based on guesstimates of horn length and filter delay. Decent impulse measurements were impossible, at least for me using Smaart.
You seem to be a careful and thoughtful experimenter. Not to mention your care in good write-ups. I look forward to reading posts from you.
However...... in my recent little explorations of time-alignment, I found non-trivial discrepancies between what my tape measure says and what REW and DSP-correction says. The equipment (if digital), enclosure features, and whatever room issues are confounding factors.
So it is possible your explorations in alignment were imprecise, as Canadians would put it.
Ben
Ben, yes, imprecise for sure 😀
But still, probably more accurate than not...
What would be the definition of precise alignment be with multiple subs anyway?
How precise a phase alignment can an averaging be? And to what listening position are we optimizing the averaging? 🙄
Dropping one rock into a pool of water will always create a different ripple than dropping several rocks even if they have the capacity to move the same amount of water
Not to throw fuel on the fire, but regarding point #2 - opening two windows will change the modal distribution of the room. Going back to what djk said (which was mostly ignored or misunderstood) - a measurement in the time domain might reveal different bouncing around of room modes with the windows open which is caught by the slow-response measurement at the single listening position rather than "room gain" (which I'm fine with calling whatever the phenomenon in question is).
Ie: the direct sound did not change, but the different reflections will sum up constructively/destructively differently at different locations - and it seems that with the windows open it added constructively at the one listening position measured.
Best,
Tal