FAQ: Subwoofer Placement

[leonski]

In listening further, he goes RIGHT ALONG with the Harmon White Paper, which is nearly as I can ascertain, the FIRST big study of multiple subs and bass distribution.
He presents this stuff like he did ALL the research, which I somewhat doubt.

One other point, A room size is NOT square feet or those dimensions. A room is CUBIC and so has Volume. A room with 8 foot ceilings will have different proprties than the same 'square footage' with 10 foot ceilings.

Here is a link to a room mode calculator for standing waves and 'peaks' / 'nulls'.

www.mcsquared.com/modecalc.htm

A decent calculator among the several online. Understanding THIS information will help in understanding the 'why' of multiple subs.

[Deleted]

yeah he definitely has some headscratcher moments in there...but his fundamental idea is sound

[KeithL]

There are two other factors that usually seem to be overlooked - and which can be quite important.....

First, all crossovers have finite slopes, and in fact aren't all that sharp. If you have a sub with a crossover at 80 Hz, and a 24 dB/octave slope (which would be pretty typical for a sub), then that sub is only down about 27 dB at 160 Hz. You must also consider that our ears are much more sensitive to higher frequencies than very low ones, so that 160 Hz tone may be more audible (relative to the 80 Hz tone) than you might expect. You also need to consider location... Assuming that you have a 160 Hz tone which is being produced in equal amounts by the left front and right front, having a little of that 160 Hz tone being produced by a sub located directly front and center won't make much difference; but having that 160 Hz tone being produced, even in small quantities, by a sub located halfway back along the left wall is going to "pull" the imaged location of the source for that sound to the left and back - towards that sub.

And, second, nothing is perfect. All speakers, and especially subs - because of the major excursions required - produce some distortion. This may be from any combination of nonlinearity of cone motion (basic THD), mechanical movement of the sides of the box (which also isn't likely to be perfectly linear), various air movement and port chuffing noises, vibration in the floor and walls of the room, and even vibration of furniture or objects near the sub. Collectively, when you have a powerful sub playing very loudly, there are usually "extraneous cues" that you notice that tell you where the sub really is.

In other words, unless you can ensure that there will be no sound of any sort associated with the physical location of the sub that extends above 100 Hz, you're still better off positioning your sub or subs at least symmetrically between the left and right to avoid an obviously lopsided sound stage that pulls to the left or right when loud low notes are played.

(Now, of course, another argument in favor of multiple subs is that several subs, each of which is working less hard, are each likely to produce fewer of these extraneous location cues, and in smaller quantities.)
 

Oct 19, 2015 6:35:24 GMT -5 @chuckienut said:

Oct 18, 2015 18:39:02 GMT -5 highfihoney said:

I read that set up guide & overall its spot on , where i disagree is with the reviewers stance on bass below 80hz being omni directional & the subs placement does not effect stereo imaging .

Could perhaps you be mistaking the fundamental tone (first harmonic) of an instrument while you are also hearing the second harmonic (first overtone)?  For example, if you have a 5.1 system and you hear a 50 Hz tone from an instrument played thru the sub then that should be omni-directional and usually you would not be able to locate it regardless of the location of the sub.  Let's say the instrument was on the left side of the stage in the recording.  However, there is also the second harmonic at 100Hz which will be played thru the left speaker.  The third hamonic (second overtone) at 150Hz would also come thru the left speaker.  So your mind says oh that instrument sound is coming from the left speaker and therefore not omni-directional.  But you are really hearing from the left the 100Hz and 150Hz tones while the fundamental 50Hz tone is coming from the sub (for this example on the left).  Let's say you then move the sub to the right you will still hear exactly the same effect that the instrument is on the left.  In other words you are thinking the sub is not omni-directional because of the second and third hamonic, etc. which give you the directional clues as to the origin of the tone.   But if you were able to eliminate the harmonics and only play back the 50Hz fundamental tone like from a test CD then you could move the sub around and then without the higher harmonics you would not be able to locate the sub's/tone's location in the soundstage.  

This also why the frequencies below 80Hz do not affect the stereo imaging, but the harmonic tones above 80Hz do affect the stereo image and allow you to locate the source of the sound.  If you change the location of the sub the stereo image should not change if you have your bass management properly set up, surprisingly many folks do not.  Some folks run their sub high like above 80Hz (especially if they have small bookshelf speakers) or they don't filter/low pass at 80Hz or below,  Now in this case the sub might be itself reproducing some of those tones above 80Hz and then yes you can locate the position of the sub.  So the rule about 80Hz and below being non-locatable is dependent on having quality towers or bookshelf speakers that have response well below 80Hz and a good sub that goes above 80Hz and then you have the sub and mains crossed over at 80Hz or slightly slower.  

Generally it is accepted that frequencies below about 80Hz can be heard but not located.  Frequencies below about 20Hz cannot be heard but only felt. The 80Hz effect is why we usually recommend feeding the sub the frequencies from 80Hz and down to the sub which thus can be placed in a number of locations for the best low bass performance.  

[vneal]

Oct 21, 2015 13:44:21 GMT -5 KeithL said:

There are two other factors that usually seem to be overlooked - and which can be quite important.....

First, all crossovers have finite slopes, and in fact aren't all that sharp. If you have a sub with a crossover at 80 Hz, and a 24 dB/octave slope (which would be pretty typical for a sub), then that sub is only down about 27 dB at 160 Hz. You must also consider that our ears are much more sensitive to higher frequencies than very low ones, so that 160 Hz tone may be more audible (relative to the 80 Hz tone) than you might expect. You also need to consider location... Assuming that you have a 160 Hz tone which is being produced in equal amounts by the left front and right front, having a little of that 160 Hz tone being produced by a sub located directly front and center won't make much difference; but having that 160 Hz tone being produced, even in small quantities, by a sub located halfway back along the left wall is going to "pull" the imaged location of the source for that sound to the left and back - towards that sub.

And, second, nothing is perfect. All speakers, and especially subs - because of the major excursions required - produce some distortion. This may be from any combination of nonlinearity of cone motion (basic THD), mechanical movement of the sides of the box (which also isn't likely to be perfectly linear), various air movement and port chuffing noises, vibration in the floor and walls of the room, and even vibration of furniture or objects near the sub. Collectively, when you have a powerful sub playing very loudly, there are usually "extraneous cues" that you notice that tell you where the sub really is.

In other words, unless you can ensure that there will be no sound of any sort associated with the physical location of the sub that extends above 100 Hz, you're still better off positioning your sub or subs at least symmetrically between the left and right to avoid an obviously lopsided sound stage that pulls to the left or right when loud low notes are played.

(Now, of course, another argument in favor of multiple subs is that several subs, each of which is working less hard, are each likely to produce fewer of these extraneous location cues, and in smaller quantities.)
 

Oct 19, 2015 6:35:24 GMT -5 @chuckienut said:

Could perhaps you be mistaking the fundamental tone (first harmonic) of an instrument while you are also hearing the second harmonic (first overtone)?  For example, if you have a 5.1 system and you hear a 50 Hz tone from an instrument played thru the sub then that should be omni-directional and usually you would not be able to locate it regardless of the location of the sub.  Let's say the instrument was on the left side of the stage in the recording.  However, there is also the second harmonic at 100Hz which will be played thru the left speaker.  The third hamonic (second overtone) at 150Hz would also come thru the left speaker.  So your mind says oh that instrument sound is coming from the left speaker and therefore not omni-directional.  But you are really hearing from the left the 100Hz and 150Hz tones while the fundamental 50Hz tone is coming from the sub (for this example on the left).  Let's say you then move the sub to the right you will still hear exactly the same effect that the instrument is on the left.  In other words you are thinking the sub is not omni-directional because of the second and third hamonic, etc. which give you the directional clues as to the origin of the tone.   But if you were able to eliminate the harmonics and only play back the 50Hz fundamental tone like from a test CD then you could move the sub around and then without the higher harmonics you would not be able to locate the sub's/tone's location in the soundstage.  

This also why the frequencies below 80Hz do not affect the stereo imaging, but the harmonic tones above 80Hz do affect the stereo image and allow you to locate the source of the sound.  If you change the location of the sub the stereo image should not change if you have your bass management properly set up, surprisingly many folks do not.  Some folks run their sub high like above 80Hz (especially if they have small bookshelf speakers) or they don't filter/low pass at 80Hz or below,  Now in this case the sub might be itself reproducing some of those tones above 80Hz and then yes you can locate the position of the sub.  So the rule about 80Hz and below being non-locatable is dependent on having quality towers or bookshelf speakers that have response well below 80Hz and a good sub that goes above 80Hz and then you have the sub and mains crossed over at 80Hz or slightly slower.  

Generally it is accepted that frequencies below about 80Hz can be heard but not located.  Frequencies below about 20Hz cannot be heard but only felt. The 80Hz effect is why we usually recommend feeding the sub the frequencies from 80Hz and down to the sub which thus can be placed in a number of locations for the best low bass performance.  

On my JL Audio subs the slope can be 0-12 or 24 db. Using my ears as the only test I prefer the 12 db slope it seems tighter and less boomy sounding.

[leonski]

Keith, I addressed the crossover slope by indicating that if you want an 80hz crossover you must put the subs crossover LOWER and the main speakers HIGHER so that they 'cross' in output at 80hz.
With a 24db/octave slope, the sub may be already 6db down while the main speakers with a lesser slope may only be 3db down. The GOAL is flat amplitude response THRU the crossover. Phase will, of necessity, suffer.

I set my panels / sub with such a scheme and get good response thru the crossover region and seamless transition. My panels are 'supposed' to go to 36hz IN ROOM while the sub should be good an octave+ (easily) above that and
down to below 20hz by test.

[highfihoney]

Oct 21, 2015 2:06:36 GMT -5 leonski said:

It's LATE so i'll just respond to your point #2. You do NOT set the sub crossover to 80hz and the SAME setting for the mains. I point to THX because their must be a REASON they went with that.
So, in short I set my sub to say 45hz and the mains higher. This means that the mains are 6db or MORE down at the crossover as is the sub. Sub is 24db/oct. and mains are 12db/oct. This is itended to produce a flat response thru the crossover region.
ALL crossovers murder phase. You should really also be looking at mid/tweet crossover and the Frequency as well as low/mid. If you have a 3-way w/Sub which is in effect a 4-way system, that's ANOTHER crossover to worry about.
I don't remember the critical frequencies. It would appear the intelligable frequencies are between maybe 300hz and 6khz or some such. This is the ATT standard for Telephony. And it matters.

Oh well!
Point #4: Low bass is low bass. However, the instrument does produce harmonics. Those are addressed by the main speakers which gives the 'spatial' cues. Pure bass is not localizable by the human ear. This is at least in part because of the physics of most listening areas being very SMALL in comparison to the wavelenght of the lower frequencies. 28hz, a low note on the Piano is nearly 40 foot wavelength at 1100 ft/sec.

Point #3: What do Bose 901s have to do with Anything? Fact is that the ear is poor at distinguishing arrival times of sound. My panels SHOULD be over 5 feet from the wall so the back wave is 10miliseconds behind the first, main wave and does NOT smear it as a result. Sound perception of humans is mainly based on being hunter/gatherers so we are best, maybe, in a semi-anachoicic environment. That's up for grabs, maybe. DSP systens offer correction for time delay.

You sure are good at making stuff up. WHERE was it said that 80hz was the be-all / end all of sub crossover frequencies? It is used by THX for a REASON which is as much practical equipment oriented as physics as physiology. Since I have ZERO interest in HT, I must as what the effect is of setting the main speakers to 'small'? Does this attenuate the bass TO those speakers? Does the setting to 'large' leave them run full-range?
Who said a sat/sub was 'superior' to other or full-range speakers? Finding a musical sub is easy if you've got plenty of $$$, but more difficult as budget diminishes. The very lowest tier are non-musical boomers best for movie effects.

Please read the Harman White Paper on sub placement. Good stuff which will reinforce some things you already know and maybe challenge some things you THINK you know. Always a nugget or two in a article like that.

And I too have been at this around 40+ years. And have heard more than my fair share.

Look here , you do realize all this s#!t is getting way beyond childish , it was childish when it started now its chest bumpimg nonsense, especially when it all started by me agreeing with a set up guide with exception to 2 things , i didnt agree with the 80hz crossover point ( apparently you dont agree as well )& i did not agree that omni directional bass from a single sub placed where the sub sounds best rectifies the variables introduced, as a matter of fact here are my exact words .

I read that set up guide & it is spot on , where i disagree is with the reviewrs stance on bass below 80hz is omni directional & the subs placement does not effect stereo imaging .

Now what i got in return for those 2 factual statements ( and god only knows why & for what end )was a multi paragraph explanation of what i might be hearing to explain away why my 2 factual comments were wrong , which they were not then ,& still are not wrong now .

Pretty simple statement & apparently you partially agree that 80hz isnt the magic mark ,which is what i said to start with, then we've also heard from others including Keith that sub placement does effect imaging ,again point #2 is something i said to start with & backed up by others so whats the big issue here & what is the goal in continuing this nonsense ?

So we are all on the same page here i will reiterate .

#1 I contended a single sub placed where the bass is best is not whats best for stereo imaging.

#2 80hz is a thx standard & not the true point where bass becomes omni directional , its a thx standard derived for theater use & not the accepted cut off point by the Audio Engineer Society .

These were my original factual statements that were met with multiple paragraphs, which in no way disproved the 2 facts i stated & in following replys still havent been disproved, but have beem reinforced by yourself Keith & others .

But since ive now been accused of " being good at making things up" please explain exactly what you , or anybody else thinks i just randomly made up .

At alot of forums there are members who's daily goal is to find some ridiculous detail to argue fractional issues over , then run with it in order to prove somebody wrong so they can be right , apparently this forum is plagued by this as well ,this entire bit of nomsense completely derailed a faq for no other reason than to argue 2 points that werent wrong to start with .

Again anybody is more than welcome to tell me where my briefly mentioned points about low bass from a single sub effecting imaging when placed where bass is best , or to point out any instance where i imserted fictional material into any of my posts , i will gladly respond with apologies if im shown to be posting fictional info then split, if not then i have a right to be pissed because this entire derailment was caused over 2 nonsensical issues that were not false or made up info to start with .

This was meant for both members who made a mountian out of a mole hill .


Edit , i forgot to mention mono bass , as told to me in this thread multiple times that all low bass is mono this is not true , stereo bass material is recorded on many many quality recordings , i'd love this fact addressed since every fact i brought up about the recording determining weather low bass is mono or stereo was overlooked & has not been addressed in any way , again anybody showing where i just " made stuff up ", especially about how recordings determine mono or stereo bass & i'll apologize & split this forum forever , im tired as he!! of arguing minutia every last time i post anything not related to baseball .

[Deleted]

Keith makes good points in his post.  However, although I usually agree with the sub(s) along the front soundstage or only slightly down the side walls and I do go for a 24dB slope, I do slightly disagree that -27dB's will still pull the source of the sound off significantly.  I will try some testing with a 160Hz tone
sometime with my system and see what I find, he might be right on.

It is difficult in a several paragraph response to cover all the exceptions.  So it we are left to normally make general statements that we feel to be usually true. It is also difficult to get others to accurately interpret what we have intended to say unless we are very careful with our wording and not going on for pages.  I made a mistake in wording when I said that the lower bass signal was mono.  To clarify I was not talking about the recording process being in multi ch, stereo or mono but the fact that after the lowest bass (for example below 80Hz) in the 5 channels of signals are re-directed and combined with the LFE signal and sent to the sub we now have a mono signal sent to the sub.  Yes, if we aren't very careful in our bass management, like simply allowing Audyssey to make all the setting we can end up with a mess.  I believe if we usually set it up carefully we will end up with the sub(s) and the mains passing each other ideally at -3dB's at about 80Hz and a 24dB slope.  With the sub preferably along the front wall that 160Hz signal that Keith mentioned will be at about 27dB level in the sub below the level of the signal thru the left, center or right speaker depending on the recording.  Obviously there is no brick wall cutoff like there is with a correctly recorded LFE channel at 120Hz.

We all can disagree on audio/HT on points that are subjective like which speaker, amp sounds better, and on and on.  However, there are basic facts that are not disputable like lower tones are bass and high tones are treble or that a +3dB volume increase requires twice the power and so forth.  But here we have a member who has 40 years in the hobby who posts that a Polk speaker that has published FR of 20Hz-20kHz therefore has 15dB's more output at 20Hz than a Klipsch speaker that has a published FR of 35Hz - 24kHz +/- 3dB.  I explained that the FR of the Polk speaker is meaningless the way it is published.  If he had looked at the Polk website he would see that Polk clarifies the FR by stating "-3dB Limit 35 Hz".  In other words both Polk and Klipsch indicate by their own specs imply that these two speakers have the same exact low bass extension.  In another post statement he shows that he does not understand the standard speaker sensitivity rating  of decibels at 2.83Volts (one watt) at one meter into 8 ohms.  These are both very basic specs of speakers that any person with just one year experience in our hobby (let alone 40 years) should be aware of.  This is not meant to attack this individual but to point out that as far as simple basic objective facts especially concerning accepted specs, one should verify before posting misinformation here. Unfortunately it has become more common for even well known speaker brands like Wharfedale, Polk, and many others to publish meaningless FR specs in ads or even website and owner's manuals.  Yes, lots of the correctly stated speaker specs are exaggerations but that is better than meaningless specs obviously directed to newbies.  Reminds me of the old days when many HT in a box systems at Walmart had 1000 watts of power.  In the 70's when friends would ask me how many watts of power in my new Marantz 2270 receiver had, I would respond 70 watts per ch into 8 ohms, etc.  They would say, Oh that's nothing my Emerson stereo puts out 500 watts!   

   

[vneal]

I got both my JL AUDIO F110s dial in and my system has never sounded better. Discriminating ears cannot tell where lower bass is coming from. I experimented with crossover and settled with 80HZ (the CM10s  measure a 6db point at 28hz) I like the sound of the Low Frequency Filter set at 12db the best for my room. The ARO sytem for this sub works great. I have heard few systems as satisfying. Cheers

[leonski]

Bass being localizable depends on Size of the space. Once you are in a very large space with dimensions greater than wavelength, you can find bass source. Until than? I'd say that overtones / harmonics predominate.
Bass wavelengths range from 14+ feet @80hz (@1150 ft/sec) to over 40 feet for 27.5hz, the lowest note on a standard 88 keyboard. The Imperial Grand of Bosendorf goes lower. Lower speed of sound results in shorter wavelength.

www.avsforum.com/forum/155-diy-speakers-subs/1268766-what-point-does-bass-become-omni-directional.html

From AVS Forum. 8th reply down sustains my point about room size. Some persons MAY be more sensitive to this than others.

Did I see Harmonics mentioned? and PHASE? I will repeat that crossovers murder phase. Graphic EQs are even worse, with phase distortions every octave or 1/3rd octave, depending. IF you are sensitive to phase / picket fencing / modulation, than I'd suggest keeping EQ, graphic and parametric, to a minimum. Or even experimenting with FIR filters, which do Not mess with phase. But also do not preserve group delay characteristics.

[leonski]

www.adamjhill.com/Hill%20+%20Hawksford%20-%20AES%20135.pdf


ONE last thing: The AES paper to which HFH ref's.

[leonski]

legacy.spa.aalto.fi/research/cat/psychoac/papers/kelloniemiaes118.pdf

And 'paper 6431' which has yet more information and probably should be read FIRST



cheers:

[leonski]

In re-reading this thread, one other very small point occurs to me. Thanks to Keith for pointing this out.
A sub which is ELECTRICALLY crossed at 80hz w/24db/octave will be Electriclly down 24db @160hz. In addition, the sub driver will be down an additional amount. 4db? 8db? 15db? I don't have any idea.
The question than is it ADDITIVE? Or do you somehow MULTIPLY the effect? The fact that a given sub MIGHT be down 24db PLUS the driver roll-off of an additional 12db at the 'octave' point helps in NOT being
able to localize it. The sound will simply be overwhelmed by the main speakers which are presumably 'stereo' and lower in distortion.

My speakers may or may not be considered full range with response to <40hz. The rating is 34hz or 36hz. I don't care. The Sub is probably good to 100hz or so. THAT'S why I set the sub low pass to maybe 45hz or so/ 24db/octave and the HIGH pass in the preamp to the main amps at 55 or 60 hz with a 12db/octave slope. I HOPE to get a flat amplitude response THRU the crossover with phase falling where it may.

If I had a Cost No Object solution? Yep, a pair of subs, not doubt. Lower distortion at lower levels ALONE would make this worthwhile. But WAit, there's MORE! ALSO I'd redo EVERYTHING with FIR Filters using a MiniDSP Product. FIR filters have NO phase problems thru the crossover. This does away or makes moot some of hifihoney's objections. Which apply to ALL conventional crossovers, not just sub crossings.