Understanding subwoofers

[Devnull]

First off, stereo subs located in the front corners will give you tremendous stereo separation from 200Hz to 20Hz. It would also work for mono with a lower XO point.

Having your subs a little wider than your mains will not be audible. We can hear stereo in low bass, but our sensitivity to stereo is less acute than in the mids and highs.

There are a few things, though, to consider:
• The mains will need to be delayed so that they are properly time aligned
• Corner placement will require some low EQ shelving due to the corner boundary acting like a horn (room gain… it’s like getting a couple hundred extra Watts in amp power)
• It could excite room modes that will need to be tamed
• I’d never buy a sub that can’t reach past 200Hz. I prefer two 12’s to a single 18 inch even though they move the same amount of air.

Explain to us how you do your speaker management.

AVR or separate DSP? What happens to the LFE channel if you want to use it for something other than music?

Is 20-200 routed to all subs? Is it mixed into one mono signal? Is R only routed to the subs on the right side of the room?

How do you do time alignment between the subs and speakers? Time alignment between all the subs?

How is this implemented if you don't have a uniform rectangular room?

 

[EJ3]

200Hz. Could be an option if you get a good integration. In practice a DIY three-way speaker (if you have two-way top speakers), with everything that means and is required to make a good speaker.

But I have ZERO reason to go that high. But why would I want to? On my mains: my +2DB is at the port resonance of 60 Hz:
The close-miked woofer (and port) response was also considerably flatter than measured from most speakers, with a very small bass-resonance peak. At the system resonance of 60 Hz, the output was only about 2 dB above its average level in the upper part of the woofer’s range, and even that minor output variation was spread over almost two octaves. When the bass curve was spliced to the room-response measurement, the resulting composite frequency response was flat within about ±2 dB from 26 to 20,000 Hz. The horizontal directivity of the tweeter was only discernible in the room measurement above 10,000 Hz.

 

[SHB]

Explain to us how you do your speaker management.

AVR or separate DSP? What happens to the LFE channel if you want to use it for something other than music?

Is 20-200 routed to all subs? Is it mixed into one mono signal? Is R only routed to the subs on the right side of the room?

How do you do time alignment between the subs and speakers? Time alignment between all the subs?

How is this implemented if you don't have a uniform rectangular room?

• Emotiva XMC-1 pre/pro
• 7 KEF LS50s crossed over @170Hz to four stereo, stacked Rythmik L12 subs (2L, 2R)
• LS50s set to Small
• Subs are stereo so all low frequencies are mixed down to stereo when listening to surround
• This is my setup for any source… haven’t really found the need rear subs
• Subs are under the Left and Right speakers so no need for time alignment
• Room is an odd, U-shaped space where the listening space is runs along the bottom of the U. Room configuration is a LEDE (live end, dead end) set up. It has no bass modes (quite amazing) with incredible stereo imaging.

 

[SHB]

First, my mains semi-anechoic (with no EQ of any type) FR is +-2DB 26Hz-20KHz.

Did I say anything about your specific speakers? No, I didn’t.

Of course, the big thing to then look at with your specific speakers is where, when tested, do their woofers stop acting linear and start to distort? That point where the distortion starts increasing logarithmically is where you set the crossover. Maybe just a bit higher.

Also, if the speaker has ports, they need to be plugged since all they will do is leak resonant noises from the inside of the cabinets. A small consideration, but very easy to do.

 

[thcdru2k]

A co-located approach with a high crossover is essentially turning a 2-way into a 3-way, so it avoids a lot of the usual integration issues and can work very well in that context.
Once subs are placed away from the mains, the problem changes. At that point, crossover, timing, and phase become much more critical, and pushing the crossover higher can introduce localization and coherence issues if not handled carefully.
So both approaches can work, but they are solving different problems and are not directly interchangeable.

 

[thcdru2k]

The crossover point is really a system-level decision. With small 2-ways you are managing driver limits, with larger mains you are managing integration and spatial coherence. Those lead to very different optimal crossover regions.

 

[levimax]

Please show me any 18” pro woofer driver that has a resonant frequency (Fs) and output substantially lower than 40Hz.

Yes, their sensitivity is high, but they don’t reach low.

I have a Line 6 StageSource single stack PA rig for my bass amp. It has a dual 12” sub cabinet (L3s) that just makes it down to 38Hz. All of the 18” pro woofers I’ve seen are comparable.

Pro drivers are optimized for output level, not low frequency extension.

On the other hand, my Rythmik L12 subs are -3dB at 18 to 200Hz.

They’re apples and oranges.

Drivers are only part of the sub equation. The other part is the enclosure. Large Pro drivers in a large ported enclosure can put out a lot of output even below 20 Hz and massive output from 50 Hz to 800 Hz. If you use a high shelf filter to shave off the extra output above 50 Hz you end up with flat FR with very high SPL and very low distortion. A good example would be a driver like this https://eminence.com/products/nsw4018_8 in an enclosure like this https://shop.gsgad.com/collections/...ies/products/roundover-full-marty-single-unit which can output 120 dB from 17 Hz to 500 Hz with very low distortion.

 

[Devnull]

• Emotiva XMC-1 pre/pro
• 7 KEF LS50s crossed over @170Hz to four stereo, stacked Rythmik L12 subs (2L, 2R)
• LS50s set to Small
• Subs are stereo so all low frequencies are mixed down to stereo when listening to surround
• This is my setup for any source… haven’t really found the need rear subs
• Subs are under the Left and Right speakers so no need for time alignment
• Room is an odd, U-shaped space where the listening space is runs along the bottom of the U. Room configuration is a LEDE (live end, dead end) set up. It has no bass modes (quite amazing) with incredible stereo imaging.

So all you've really done is make a 3 way speaker, without taking into account any of the issues associated with subwoofers not being located directly under the speakers.

Kinda need take your blanket assumptions about corner placement, power increases, soundstage, eq, timing, etc with a grain of salt it seems.

 

[levimax]

It’s the other way around: We are 20 db less sensitive. The Fletcher-Munson (equal loudness) curves mean we have to turn up the volume in the bass in order to achieve the same perceived loudness.

Our hearing is more sensitive at 40 Hz than 20 Hz. That means the 40 Hz distortion products produced from a 20 Hz tone are perceived 20 dB louder and 60 Hz distortion products from a 20 Hz tone are perceived 30 dB louder. This means the distortion products from many subs playing 20 Hz are perceived louder than the fundamental.

 

[levimax]

..(maximum distance of acoustic centers of 1/4 wave length at crossover frequency)..

How do you solve it in practice with, for example, a two-way speaker that has a crossover point at, say 2.5 kHz. It's hardly even physically possible to place the woofers and tweeters so close to each other then.

In a three-way speaker crossover around 300 Hz, 1/4 wave length, c-c distance around 30 cm. It might work.

You see the extremes companies go to to try to get drivers close. Co-axial is about the only way you can get to 1/4 wavelength (point source) in most cases. The next stop is 1.25 wave lengths which is what a lot of companies use and is OK but not ideal. For mains to subwoofer at 100 Hz it is not a problem if one is one top of the other, 200 Hz is 17" which is getting a little harder. What is really great about getting 1/4 wavelength is that you can use Linear Phase Crossovers and get cancellation of the pre-ringing. Ascilabs is proving the advantages of Linear Phase crossovers and they really shine for a main / sub cross.

 

[SHB]

Drivers are only part of the sub equation. The other part is the enclosure. Large Pro drivers in a large ported enclosure can put out a lot of output even below 20 Hz and massive output from 50 Hz to 800 Hz. If you use a high shelf filter to shave off the extra output above 50 Hz you end up with flat FR with very high SPL and very low distortion. A good example would be a driver like this https://eminence.com/products/nsw4018_8 in an enclosure like this https://shop.gsgad.com/collections/...ies/products/roundover-full-marty-single-unit which can output 120 dB from 17 Hz to 500 Hz with very low distortion.

That Eminence 18” woofer has an Fs of 36Hz.
It specs usable response to 27Hz, but that’s going to be -10dB down so this is really just a woofer. Port tuning for this is 36Hz.

Closed box for this driver will have it rolling off below 80Hz (-3dB). For this driver to be flat to 20Hz will require 16dB of parametric boost. That will mean the driver will require 40x more power at 20Hz compared to 100Hz.

Xmax is only 15mm so trying to get useful output at 20Hz will not be possible with this driver.

Where are you getting your numbers from?

Eminence calls it a sub, but, for HiFi, this is really just a woofer.

 

[SHB]

So all you've really done is make a 3 way speaker, without taking into account any of the issues associated with subwoofers not being located directly under the speakers.

Kinda need take your blanket assumptions about corner placement, power increases, soundstage, eq, timing, etc with a grain of salt it seems.

Yeah, a 3-way speaker that’s flat 20-20k that plays much louder than I can stand In A LEDE room that sounds incredible by any standard.

I’ve dealt with subs located across corners in the past where the rear sub was delayed 11ms to smooth out a room mode.

Bet you couldn’t have figured that out.

Thanks for the snark. Looks like the haters have shown up.

 

[levimax]

It's easy. If you have a distortion problem and can actually hear it then just stack up more subs. But even a single sub with reach to 20hz will do better than any passive tower that aspires to reach to 20hz. Or 30hz for that matter.

I have heard distorted subwoofer bass and I have heard clean subwoofer bass. I prefer clean and I think most people would if they got a chance to hear it. Why not try to get it right even if it is hard.

I really doubt that on my practical experience. If you are referring to "envelopment" that is really long shot.

It is trivially easy for me to localize a clean playing sub down to 50 Hz.... 40 Hz is where it starts getting hard. Same conclusion a new study has shown https://www.audiosciencereview.com/forum/index.php?threads/bass-direction-is-audible.55079/ Nothing to do with "evelopment".

This is just limitation of how you perceive bass management. It is actually solved problem, but you are probably not aware of it. It's always good to read the news when you are invested in the topic.

There are fundamental differences in co-locating subs and mains (or full range speakers) and splitting subs and mains regardless of the integration solution and there are trade offs to be aware of. Modern automatic systems can work very well in some situations and I am quite aware of what is out there and own several including DLBC and I will try ART when they release if for PC but we are very far from having bass issues in small rooms being a "solved problem".... but progress is not doubt being made.

 

[levimax]

That Eminence 18” woofer has an Fs of 36Hz.
It specs usable response to 27Hz, but that’s going to be -10dB down so this is really just a woofer. Port tuning for this is 36Hz.

Closed box for this driver will have it rolling off below 80Hz (-3dB). For this driver to be flat to 20Hz will require 16dB of parametric boost. That will mean the driver will require 40x more power at 20Hz compared to 100Hz.

Xmax is only 15mm so trying to get useful output at 20Hz will not be possible with this driver.

Where are you getting your numbers from?

Eminence calls it a sub, but, for HiFi, this is really just a woofer.

Again you can't look at driver specs and make all these assumptions, you need to look at the enclosure and driver together. While not designed as a Hi-Fi subs they are used that way to good effect and is how it is done with most serious HT subs. Here is some information from the GSG web site:

 

[Keith_W]

Why in the world would I want to go up to 200Hz? That is well into woofer territory (which is definitely not needed from a sub in my system).

You're right. I saw that in the first post and thought "what a strange recommendation". But then I thought about it a bit more. We don't normally want to be able to localize our subs, but that is because the usual recommendation is to place subs strategically in the room so that they reduce room modes. Which may or may not be symmetrical, which may or may not involve front/rear placement. But what if your subs are on L/R of the listening position? Would a high XO point increase the "acoustic envelopment" effect? I suppose the only way to find out is to redo my DSP with a high subwoofer XO point and go and listen.

 

[Satanica]

Why in the world would I want to go up to 200Hz? That is well into woofer territory (which is definitely not needed from a sub in my system

Not 200Hz, but Dirac ART can work up to 150Hz.

 

[Jaxx1138]

Even that hideous, idiotic, $788k, alleged greatest speaker ever (not) by those douche bags at Wilson Audio is essentially a sub/sat system. They were just stupid enough to bolt the mains to the subs.

Yes of course, but let's not beat around the bush, tell us how you truly feel.

 

[Satanica]

Many super-expensive high-end top-of-the-line systems are actually sub-satelite systems...

I actually can't think of any that fit that criteria.

...I'm always baffled why it's not done more.

Floor space, extra cables, ugliness, reliance on dedicated amplification (sometimes).

 

[Jaxx1138]

Subwoofers in any HiFi system will enhance the performance if they are implemented correctly. The problem is; they rarely are implemented correctly. Any well designed subwoofer will have low distortion, low group delay and good linear output(with little to no compression) to at least 110dB. With proper placement (application dependent variable) , time alignment, phase integration, and DSP correction, you can achieve extended bottom end and better system dynamic capability. In room measurements are required. Crossover slopes need to be independent to the sub or mains (slopes may be different on one vs the other). Delay on the mains (in most cases) will be required. If you're just dropping a sub in , feeding it a signal and hoping for the best, it is highly unlikely to get good integration in your system. Personally, I think DSP/EQ should be applied to just about all systems. Every part of the electronics and passive networks are EQ. All recordings have EQ. This antiquated "audiophile" belief that EQ is bad is just a myth and a lack of understanding of how sound reproduction works. Hell, the room EQ's your system acoustically for God's sake! As far as where you want it crossed over at is up to you. As long as you take first principles into account concerning wavelength of your X-over frequencies and distance to your mains. Reducing group delay and reducing ringing/RT60 below 100Hz goes a long way in getting the tight impactful bass, even into subsonic frequencies. Trinnov's bass array tech or the cardiod woofer applications can really improve bass articulation in most rooms. Follow the science and you can learn a lot and get great sound.

 

[thcdru2k]

Subwoofers in any HiFi system will enhance the performance if they are implemented correctly. The problem is; they rarely are implemented correctly. Any well designed subwoofer will have low distortion, low group delay and good linear output(with little to no compression) to at least 110dB. With proper placement (application dependent variable) , time alignment, phase integration, and DSP correction, you can achieve extended bottom end and better system dynamic capability. In room measurements are required. Crossover slopes need to be independent to the sub or mains (slopes may be different on one vs the other). Delay on the mains (in most cases) will be required. If you're just dropping a sub in , feeding it a signal and hoping for the best, it is highly unlikely to get good integration in your system. Personally, I think DSP/EQ should be applied to just about all systems. Every part of the electronics and passive networks are EQ. All recordings have EQ. This antiquated "audiophile" belief that EQ is bad is just a myth and a lack of understanding of how sound reproduction works. Hell, the room EQ's your system acoustically for God's sake! As far as where you want it crossed over at is up to you. As long as you take first principles into account concerning wavelength of your X-over frequencies and distance to your mains. Reducing group delay and reducing ringing/RT60 below 100Hz goes a long way in getting the tight impactful bass, even into subsonic frequencies. Trinnov's bass array tech or the cardiod woofer applications can really improve bass articulation in most rooms. Follow the science and you can learn a lot and get great sound.

That’s why a simple 2.0 can make a lot of sense. With good placement and some boundary gain, you get solid in room bass with everything coming from one place, no crossover, no extra variables. It just sounds natural.
Subs are great for movies and full range impact, no question. But for music, especially in smaller or secondary rooms, the simplicity matters. Less setup, less tweaking, more time actually enjoying it.
I’m not biased either way. I run dual subs in my main system, a single sub in another, and a 2.0 setup elsewhere. Each has its place.
True integration takes effort, and sometimes a clean 2.0 gets you closer to the goal with less work.