LF Driver Size vs. Soundstage Size: Related? How should we be choosing subwoofer (s)?

[Chrispy]

I meant engineer. Have some trigger words, do we?

Meant or can actually prove? Many "audio engineers" are no more engineers than your bus driver or janitor.....

 

[Chrispy]

I meant engineer. Have some trigger words, do we?

So what is the specific degree this designer holds?

 

[kemmler3D]

And if so, why is ~80 Hz the approximate number for being able to locate subwoofers instead of 110 - 120+ Hz?

Lots of research out there on this, the Wikipedia article is also good. The threshold below which localization becomes really hard is actually considered to be 200hz, and as you note below 80hz impossible... this is also why you often hear "bass is mono" in both mixing and listening circles.

Interaural level differences are very low in this frequency range, especially below about 200 Hz, so a precise evaluation of the input direction is nearly impossible on the basis of level differences alone. As the frequency drops below 80 Hz it becomes difficult or impossible to use either time difference or level difference to determine a sound's lateral source, because the phase difference between the ears becomes too small for a directional evaluation.

Can we be sure re: bass frequencies vs. driver size? The 15” and 6” drivers will not likely operate with the same speed or excursion. As an integral part of driver size, do those characteristics not influence dispersion, too?

Size is the biggest factor, not the only factor. Shape also matters. Check out the measurements of individual drivers on e.g. hificompass - drivers of the same size start to show directionality around the same frequency. For example, these two very different 12"s both start to "beam" at the same frequencies (you can see the lines pull apart at 400hz and then again right around 750):

Eton 12-680/62Hex

SB Acoustics SB34NRXL75-8

Consider bookshelf speaker models that benefit from subwoofer assistance above 100 Hz (esp. for some male voices) - they don’t all share LF dispersion behavior. Similarly, they don’t all seem to share soundstage characteristics in lowest frequencies on their own or paired with subwoofer(s).

So, I would half agree with this. If you take two sealed bookshelves or two with the same port tuning, I would expect LF radiation patterns to look basically the same at a given SPL. This is because you may get interferences with the room and with that look different depending on what the ports are doing and where they're facing. A bookshelf with front-facing ports may have a different in-room pattern than one with rear-facing ports.

However, for a given cabinet and baffle size, driver size, tuning, and at 100hz, I would expect the radiation patterns to look VERY similar, basically always omni. If you look at the litany of speaker reviews on here, and look at the spins, you'll see what I mean. Almost all speakers are effectively omnidirectional at those frequencies.

Bottom line I would say bass is the least interesting place to investigate differences in soundstage between different speakers.

Now, there's one place I would say you are not on the wrong track, and that is when the woofer / midwoofer drivers are producing higher frequencies. It's not rare to see relatively large drivers driven up to 2000hz or so. In that case, their size will have a big impact on soundstage. Larger drivers tend to beam more at higher frequencies (as alluded to above), and therefore if you're entrusting (say) 1.2khz to a 12" driver vs. a 6" driver, you will definitely hear a different soundstage, as the mids will be arriving in a much smaller "sweet spot" from the 12" vs. the 6". By the same token, the early reflections will also sound different from each driver at a given mid frequency.

This may not be totally intuitive, I didn't fully get how dispersion worked until about a year ago, but if you look over enough spins you'll see what I mean. Purifi's site is good for this, as they publish off-axis FRs (not exactly "full spins" but good enough for this) and the drivers are all very similar except for size. So if you flip through the different product pages, you'll see what I mean. https://purifi-audio.com/ptt6-5x04-nfa-01/

This will tend to significantly alter your perception of stereo image or "soundstage", all else held equal. So yes, low frequency driver size has a big impact on soundstage, just not because of how they reproduce what we typically consider low frequencies.

 

[ozzy9832001]

I feel like this topic ended up as a loaded gun and I'm not sure why.

I think it we look at soundstage to be stereo image, then I'd say bass, in general plays a huge role, even if it's not localized. I think our brains trick it into thinking it's coming from the central image because that's where the rest of the sound is appearing to be coming from. I think a drivers ability to accurately (key word) reproduce it is the biggest thing. While a 6" sub can reproduce a 40hz signal, it sounds way better on a 10" driver then said 6. I feel this is why xover freqency has to be experimented with because depending on the room and driver sizes it isn't one size fits all. As a general rule of thumb, 80hz has been shown to be the overall best starting point, but obviously some situations require +/- from there.

Maybe I'm misinterpreting what the OP is asking...

 

[MaxwellsEq]

I feel like this topic ended up as a loaded gun and I'm not sure why.

I think it we look at soundstage to be stereo image, then I'd say bass, in general plays a huge role, even if it's not localized. I think our brains trick it into thinking it's coming from the central image because that's where the rest of the sound is appearing to be coming from. I think a drivers ability to accurately (key word) reproduce it is the biggest thing. While a 6" sub can reproduce a 40hz signal, it sounds way better on a 10" driver then said 6. I feel this is why xover freqency has to be experimented with because depending on the room and driver sizes it isn't one size fits all. As a general rule of thumb, 80hz has been shown to be the overall best starting point, but obviously some situations require +/- from there.

Maybe I'm misinterpreting what the OP is asking...

I felt the OP flew a kite - that big drivers make for big soundstages. The general consensus seems to be: there's no known correlation.

Further, some have pointed out that bass does not play a significant part in sound location.

But, there is also evidence that big drivers can "beam" more than smaller drivers at higher frequencies negatively impacting dispersion.

In summary, big soundstages may not need big drivers, and they may actually make things worse.

 

[kemmler3D]

I think a drivers ability to accurately (key word) reproduce it is the biggest thing. While a 6" sub can reproduce a 40hz signal, it sounds way better on a 10" driver then said 6

This isn't INHERENTLY because of the size, it's just that smaller drivers need much longer excursion to do low frequencies than a large driver does. This leads to more distortion and worse sound in low frequencies for small drivers.

However, there is no rule that says you can't have a small driver with crazy-long excursion and low distortion - that's the big value proposition for the Purifi drivers. Some of their 6" could probably outperform a cheap-o 10".

But you're not wrong, just wanted to clarify that the worse 40hz signal isn't because of the size per se, but because of the greater demands placed on the driver due to its small size.

I think it we look at soundstage to be stereo image, then I'd say bass, in general plays a huge role, even if it's not localized.

I am going to yet again push back on the idea that bass (esp. in the <80hz regime) plays much role in stereo image. If bass played a significant role in stereo image, why would it be OK to place subwoofers randomly around a room (behind you, in corners, etc etc) wherever they happen to have the least-bad nulls / peaks?

But, there is also evidence that big drivers can "beam" more than smaller drivers at higher frequencies negatively impacting dispersion.

In summary, big soundstages may not need big drivers, and they may actually make things worse.

It's not just evidence... someone more engineer-y than me can probably explain the mechanism, but beaming is (AFAIK) something like an unavoidable outcome of physics due to the relationship between cone / dome size and wavelength. If you look into DIY speaker building, beaming vs. size/shape is acknowledged as a simple fact of life, and major factor in driver selection.

e: Found an explanation. It's because sound waves are coming from the entire surface of the cone, but once the wavelengths are similar in size to or smaller than the cone, waves going along off-axis paths start to cancel each other out. There's a helpful picture in the link. This seems to be an unavoidable physical phenomenon.

why do larger drivers beam upper frequencies?

I am wondering if anyone could explain the principle behind why a larger driver seems to beam upper frequencies. I just want to understand technically what this effect comes from! I have been playing with a Stephens Tru sonic FR120 full range driver (12") running with a 15" OB and I have...

www.diyaudio.com

However, I would not go so far as to say "better or worse" or even "bigger or smaller" WRT soundstage, beaming, and driver size. The impression of soundstage is really subjective, and depending on the rest of the speaker design, some beaming in the mids and treble might improve your experience of stereo / soundstage, also depending on the room. E.g narrow dispersion is sometimes considered valuable for nearfield monitors, but you don't often hear complains that the soundstage is bad with that gear.

I think what I would be confident saying: you should expect some impact on soundstage if the drivers reproducing >800hz (or so) change in size, all else held equal.

I would ALSO say you shouldn't be very concerned about impact on soundstage when it comes to however you reproduce <200hz and you can safely ignore the impact on soundstage below 100hz or so.

 

[sejarzo]

All definitions of "soundstage" I have read mean the width and depth of the perceived image along with the ability to localize instruments or voices.

Assuming that two different subs can reproduce the same frequencies at the same SPL, and the SPL and crossover frequency are set properly so as to minimize localization, I can't see how the driver size is relevant. If there is any correlation, then I would expect it to be weak.

 

[fpitas]

I feel like this topic ended up as a loaded gun and I'm not sure why.

I think it we look at soundstage to be stereo image, then I'd say bass, in general plays a huge role, even if it's not localized. I think our brains trick it into thinking it's coming from the central image because that's where the rest of the sound is appearing to be coming from. I think a drivers ability to accurately (key word) reproduce it is the biggest thing. While a 6" sub can reproduce a 40hz signal, it sounds way better on a 10" driver then said 6. I feel this is why xover freqency has to be experimented with because depending on the room and driver sizes it isn't one size fits all. As a general rule of thumb, 80hz has been shown to be the overall best starting point, but obviously some situations require +/- from there.

Maybe I'm misinterpreting what the OP is asking...

I think you are. He's only talking about soundstage. The low frequencies you're thinking of have very little directionality because of the long wavelengths relative to the baffle. So, it's hard to think they influence soundstage very much, let alone different size drivers having a big influence.

 

[benanders]

I felt the OP flew a kite - that big drivers make for big soundstages.

…

In summary, big soundstages may not need big drivers, and they may actually make things worse.

Not quite - I didn’t specify which relationship (direct vs. inverse) should be likely if one exists. It’d put bias to the query.

That said, I agree your summary.

 

[benanders]

This isn't INHERENTLY because of the size, it's just that smaller drivers need much longer excursion to do low frequencies than a large driver does. This leads to more distortion and worse sound in low frequencies for small drivers.

However, there is no rule that says you can't have a small driver with crazy-long excursion and low distortion - that's the big value proposition for the Purifi drivers. Some of their 6" could probably outperform a cheap-o 10".

…
just wanted to clarify that the worse 40hz signal isn't because of the size per se, but because of the greater demands placed on the driver due to its small size.

This is what I’m getting at. Well-put.
I suppose I do consider performance inherently tied to driver size, at least in part, with regard to excursion / speed: they shouldn’t be the same between different sized drivers, and should be somewhat predictable relative to driver size.
And perhaps disproportionately important for larger, heavier drivers.

It’s straightforward to understand recognize the likelihood of dispersion x driver diameter patterns based on wavelength.
But is it straightforward to recognize how size-constrained driver performance will affect something that can’t really be measured: here that being soundstage?

Given a system with LF from 10” drivers of a specified total surface area, vs. an otherwise-equivalent system with LF produced by 4” drivers of nearly identical specified total surface area as that of the 10” drivers.
Will everything 100-120 Hz down be rendered the same, such that subjective impressions (soundstage) will be very similar (again, all other variables kept equal)?
How about everything under 500 Hz? Given optimal excursion and speed characteristics in each driver size (e.g. 10” vs. 4”), should we expect results that give the same subjective impressions (soundstage) despite what would be a difference in some aspects of mechanical performance?
It seems to me like distilling the nature of how LF stereo is rendered and perceived to measurements of wavelengths vs. driver diameters, despite driver size predicting additional operational constraints. As always, I could be missing something.

And yes - I will spend some time at the purifi link this weekend, @kemmler3D - thank you.

 

[kemmler3D]

Will everything 100-120 Hz down be rendered the same, such that subjective impressions (soundstage) will be very similar (again, all other variables kept equal)?

Yes, more or less. It's because the wavelengths there are SO much larger than the drivers or speaker cabinet, (10 feet) that at a given SPL you'd be hard pressed to CREATE a measurable difference if you needed to, let alone hear one.

(go to DIY audio and check out the threads where D&D and/or sigberg audio are designing cardioid speakers some years back, and you'll see what I mean. It's actually quite hard to get 100hz to go in any specific direction. If they could have done it by using smaller or bigger drivers, that would have been a great deal easier than what they did.)

That said, playing at 100hz with a 4" driver is not so easy, so in real life you'd run into issues with this, but if we say "all else held equal" then the output is going to be pretty damn similar.

500hz is an interesting one. That's 26 inches, which is well enough bigger than the cone, but in the same ballpark. It is also a frequency that's well enough localizable by the human ear. At this point I wouldn't bet big money that you wouldn't see some slight differences off-axis, and/or hear some changes in the soundstage. To me this frequency is probably an iffy one for 4" vs 10". If you look up some dispersion plots for drivers of those sizes, you'll see that they're both generally pretty "pistonic" at that frequency. But maybe not 100% at all angles, I'm not so sure.

Also, a side note, which is that excursion and speed (at a given frequency, the same thing expressed two ways) vs. moving mass are largely what define SPL, but not dispersion / directivity so much.

 

[MaxwellsEq]

Not quite - I didn’t specify which relationship (direct vs. inverse) should be likely if one exists. It’d put bias to the query.

That said, I agree your summary.

Mea Culpa. Your wording in the original post did not imply you held a position in either camp. Sorry.