Hi all,

I've been experimenting some with speaker placement and I found what seem to be a couple of good guides online for optimal speaker placement. I know that my current setup definitely isn't optimal (or even very close to it), though I'm limited by the fact that I'm in a dorm room, so I don't have too much liberty to adjust it right now. But I have been a bit bothered by the fact that these guides claim the speakers must be *exactly* equidistant from the speaker - in particular, I'm referring to this page: http://www.gcaudio.com/resources/howtos/speakerplacement.html. It says that differences of less than 1/4" can prevent proper imaging. I can see how one might be able to guarantee this kind of precision to the center of the listening chair, but doesn't the listener's head move more than a quarter of an inch during the listening process?? I have a hard time even stomaching the notion that hearing proper imaging requires the listener to remain motionless throughout a track. :-P Can anyone elaborate on how true this is and if there's anyway to mitigate that problem? Thanks so much!

- Daniel

When I sit down to listen to music (in 2 channel), I tend to adjust my position slightly until the imaging sounds optimal. I am only occasionally concious of doing this, it is usually automatic.

But I have been a bit bothered by the fact that these guides claim the speakers must be *exactly* equidistant from the speaker - in particular, I'm referring to this page: http://www.gcaudio.com/resources/howtos/speakerplacement.html. It says that differences of less than 1/4" can prevent proper imaging. I can see how one might be able to guarantee this kind of precision to the center of the listening chair, but doesn't the listener's head move more than a quarter of an inch during the listening process??
- DanielOf course. And how do they measure the distance? To which ear? Or do they compromise by measuring to the nose? :D

1/4" tolerances for loudspeaker placement? Time to get some more "forgiving" speakers. Sitting still within a quarter of an inch would require a seat with restraints like an electrocutioner's chair.

- Steve O.

Acoustic treatments will do far greater things for precise imaging than your precise physical relationship to the speakers ever could in an untreated room.

I skimmed that link expecting to see an article written by a crackpot who then went on to explain how sunspots would affect signal transfer unless you had the proper cables. It's actually not a bad article, and the guy covers a lot of the basics. Unfortunately it's marred by the ludicrous 1/4" recommendation and a few more things like that.

The 1/4" statement MAY well be true, I don't know. But, unless you're going to lock your head in position using one of those brain surgery fixtures, you'll just have to live with the realization that "perfect" isn't possible.

The 1/4" statement MAY well be true, I don't know.Well, think about it: how can it be true when our ears are significantly more than 1/4" apart? Even if the listener's head was locked in a vice, as you mentioned, which ear would we be time-aligning the speakers to?

Sanjay

My guess is this:

Sounds travels 1/4" in roughly 19 microseconds. Apparently, humans can use interaural time differences on the order of 10's of microseconds to compute the location of a source. Thus one might argue that if the center of your head is equally spaced from two speakers and then you move a speaker, a change in location as small as 1/4" should be perceptible.

Assuming, for the sake of argument, that this is true under laboratory conditions, I would still think it's not relevant to real life.

My guess is this:

Sounds travels 1/4" in roughly 19 microseconds. Apparently, humans can use interaural time differences on the order of 10's of microseconds to compute the location of a source. .

Yes, and one papers shows 1.5 us with special signals, but isn't that time difference between the two ears internally? I seriously doubt you will hear that 19us difference by moving the speakers a 1/4" to the front or back which gives the same 19us, no?
And easy to find out:D Have someone tilt the speaker back of forward a 1/4 or enough for the driver displacement of that amount and see if you can hear any difference.

Acoustic treatments will do far greater things for precise imaging than your precise physical relationship to the speakers ever could in an untreated room.

Exactly.

Sitting still within a quarter of an inch would require a seat with restraints like an electrocutioner's chair.

Yep:

http://www.ethanwiner.com/head-vise.gif

Hi all,

Thanks for all the good feedback - it sounds like that's definitely a too-strict expectation, even if it's technically true about optimal imaging. My guess as to why 1/4" could matter is this:

I would think it's not the issue of your brain processing the time-delay in arrival of the sound in order to place the source (presumably, we can't possibly respond on an order of microseconds). I think probably it has to do with the fact that sine wave addition affects which sounds at which frequencies are enhanced or diminished - positive and negative interference affect what you hear. It's the same reason you need to adjust the phase on a subwoofer to match your loudspeakers, really - if the low point of one wave from a sub hits the listener when the high point of the same wave from the loudspeaker, the two will cancel out and the listener hears nothing (or very little). I bet speaker placement incorporates the same philosophy, because in order to hear what the recording engineer intended (and what the microphones in the studio or wherever heard), you need to recreate the same constructive and destructive interferences as they were recorded by the original microphones.

Does that make sense to everyone? To that end, I would also assume the distances measured are from each speaker to each ear - the time between your ear receiving the sound and your brain processing the signal is probably negligible, or even if not, unimportant, as it's the same for both ears.

So, my guess is optimal imaging occurs when waves constructively and destructively interfere like they did when they originally reached the recording microphones, and so getting the most precise speaker placement to guarantee that interference is an effective way to recreate the soundstage (though as someone else mentioned, there is plenty else to consider like room treatments, which aren't an option for me in the dorm room :-P). Comments? Rockin'?

- D

Yes, and one papers shows 1.5 us with special signals, but isn't that time difference between the two ears internally? I seriously doubt you will hear that 19us difference by moving the speakers a 1/4" to the front or back which gives the same 19us, no?
And easy to find out:D Have someone tilt the speaker back of forward a 1/4 or enough for the driver displacement of that amount and see if you can hear any difference.

The work I was thinking of was exploiting the fact that as a source moves around you, the arrival times of the signal vary. The head-related effects are also valid, just not what I was thinking of (if I understand you correctly). Anyway, my post was in the spirit of "here's one way a well-intentioned guy could misapply some laboratory research", and certainly I agree with you about testing; I bet it would only show up as a valid effect under pretty carefully controlled circumstances, or with headphones.

Shameless pimping of my thread I started a few weeks ago..... ;)

Don't let the pursuit of perfection be the enemy of excellence.
http://www.avsforum.com/avs-vb/showthread.php?p=12083005#post12083005

I would think it's not the issue of your brain processing the time-delay in arrival of the sound ... I think probably it has to do with the fact that sine wave addition affects which sounds at which frequencies are enhanced or diminished

You nailed it. This is exactly what happens, and the particular type of skewed response that results from early reflections is called comb filtering. More here:

http://www.realtraps.com/art_spaces.htm

There's also a comb filtering video here:

http://www.realtraps.com/videos.htm

I bet speaker placement incorporates the same philosophy, because in order to hear what the recording engineer intended (and what the microphones in the studio or wherever heard), you need to recreate the same constructive and destructive interferences as they were recorded by the original microphones.

It's even simpler than that. If a room has no absorption at the first reflection points not only do you get comb filtering, but the response is different for the left and right ears. Precise placement is not sufficient to fix the skewed imaging because moving your head even half an inch would throw off the precise balance needed. Versus absorption which stops the problem at the source.

--Ethan

I would think it's not the issue of your brain processing the time-delay in arrival of the sound in order to place the source (presumably, we can't possibly respond on an order of microseconds).


One surprising thing, to go further off-topic, is that our brain does seem to be able to use information at these time scales, even though they're well below the timescales at which neurons operate. Again, please don't take me as supporting the author... Just relaying a finding from neuroscience.


I think probably it has to do with the fact that sine wave addition affects which sounds at which frequencies are enhanced or diminished - positive and negative interference affect what you hear.


I think the audibility of comb filtering in-room, with two functioning human ears, isn't really clear -- it'll be very obvious on a recording made by a single microphone, but our brains do an awful lot of processing and seem to be pretty well able to cope with comb filtering effects. Otherwise you'd expect a radical change in perceived timbre if you shifted your head an inch or two, and that doesn't occur. Howver, I don't know what the latest thinking is on how we use or don't use comb-filtering effects to localize sound, but I'd be curious to hear if someone knows.

But anyway, your guess as to what the author meant is as good as mine. :)

I bet speaker placement incorporates the same philosophy, because in order to hear what the recording engineer intended (and what the microphones in the studio or wherever heard), you need to recreate the same constructive and destructive interferences as they were recorded by the original microphones.


Actually you'd want to eliminate all interference as much as possible during playback -- whatever constructive/destructive interference was picked up by the mic is already on the tape; the system should just try to reproduce what's there.

I bet it would only show up as a valid effect under pretty carefully controlled circumstances, or with headphones.

Yes, that is what the paper shows, special jittered test tones and headphones:D Thanks.

I don't think a quarter of an inch is really that small a difference. It's fairly small, but if it's rotationally for instance, that certainly could make a small audible difference. An inch would be a pretty sizable amount for instance, IMO.

Yes, that is what the paper shows, special jittered test tones and headphones:D Thanks.

That sounds interesting -- can you post a reference (or link if it's available)?

That sounds interesting -- can you post a reference (or link if it's available)?

No link but PM Chu and ask for a copy of the Nordmark paper:D

It is easy enough to place your speakers within 1/4" or so. Do you have to be that precise. I don't think so. But moving speakers an inch (or two) can make a difference - sometimes - depends on the speaker, the room, and the listener. The linked article has lots of truths to it:
"The trick is to place the speaker (and/or treat the room) in a location that will take of advantage of the desirable reflections while diminishing the unwanted reflections."

IMO treating the room (first reflection points on all surfaces - don't forget the ceiling) makes a huge difference.

Clamping your head in place is just silly. The human head has a neck for a reason. Part of what your mind uses to locate sounds is the movement of the head. Plus the head clamps would reflect the sound and give early reflections (see Ethan's discussion of putting heavy blankets over the back of listening seats)

I'm not sure I really believe:
"Difference of less than ¼" will be audible in better systems to careful listeners. "
But maybe my system is not good enough, or maybe I'm not a careful enough listener.
I'm may be good to an inch or two of back or forth or side to side, rotating the speaker seems to make more difference to me - I think I could tell if you left a front corner of the speaker in the same place, but rotated it so the rear corner was moved 1/2" - maybe. A small change in the rotation of the speaker can make a big difference when you sit 10' away. . .

In my room, with my speakers, and my ears, I can move my head many inches before I notice a change in sound. I can move the speakers less and hear a change - I think, but it is hard to tell for sure because moving the speakers requires I get up, move them, then sit down again.

In an untreated, normal listening room, I can't imagine that exactly (withing 1/4") equal distances would ever be the right set up. I've never been in an actual listening room that was exactly symmetrical. But the author addresses this:
"I will add that there are times you may want the speakers to be at slightly different distances from the listener. This is mainly true when trying to compensate for a level imbalance cause by room acoustics or placement limitations."

In summary, speaker and listener position can make a big difference. In my experience changing speaker position by an inch or two is about equal to changing listening position buy 6-12". I assume this is because the speakers are usually closer to walls than ears are. Experimenting with speaker and listener position is free, and it can have a big effect on sound. It costs nothing, and the benefits can be great. By fixating on the 1/4" mentioned you are potentially missing the point of the article - speaker position does matter. Treating the room can also be cheap and makes as much or more difference. Treatments don't have to be permanent - a few portable panels can make a difference in almost any room and are easy enough to move. I was shocked by how much $200 of compressed fiberglass and cheap fabric covers (to contain fibers) made to my room.

Oops, I've got to go. I'm going to hit post rather than delete . . .

never met a person who could not hear the difference in moving a speaker a 1/4". honestly. Your hearing is much better than what you give it credit for. Have a friend tip your speaker a 1/4" forward sometime. There will be a difference. Now that article is really only the start. You want to have that equilateral triangle as a start point. That honestly is not optimal placement unless your room is perfectly built. By this I mean that behind the right wall is the same thing as behind the left wall. If the outside is behind the right wall but another room is behind the left wall that actually does affect acoustics and placement. Obviously if you treat the room placement of the speakers is still very important but not quite so much. As for the sweet spot and not moving your head. Most speakers have a decent sized sweet spot where after you really get the speakers hooked into the right spot you can move your head and still enjoy great imaging. Sumiko (importer of vienna acoustics, REL and Sonus Faber) gives classes on this stuff to the people that sell there stuff. Taken a few classes with it. It really does work. Never had a person who could not tell the difference between bad placement and good placement. This article is honestly pretty darn close to what Sumiko trains on. Pretty cool. it works well.