How To Empirically Determine Best Speaker Position?

[watchnerd]

Pretty much everything I've ever read over decades about speaker positioning for hi fi stereo amounts to some variation of:

--Place the speakers such that the listening seat is at the apex of a triangle

--Toe-in according to taste to improve imaging

--Don't put too close to the corners, side walls, or rear walls. Space helps imaging, proximity helps bass. But don't go too far in either extreme.

--Play around with it until you find the goldilocks compromise.

And that's about it.

So, if one isn't using room correction, subwoofers, or EQ, what's a measurement-centric approach to stereo speaker positioning?

 

[abdo123]

There are none. Stereo image is an effect that is manufactured in our brain based on our already established spatial cues.

It’s a good start to try to even out the frequency response between the two speakers (incase there were some serious dips or spikes) but that’s all.

 

[MakeMineVinyl]

I would think any good RTA program (like REW) would be a good start, although I've measured my setup more times than I care to remember and always go back to just doing it by ear, which is how I listen.

 

[watchnerd]

I would think any good RTA program (like REW) would be a good start, although I've measured my setup more times than I care to remember and always go back to just doing it by ear, which is how I listen.

How does REW or any RTA help assess imaging?

 

[AudGuy]

I believe in the triangle. It has always been the best starting point for me. I want the speakers as far away from the rear and side walls as is reasonably possible. Then the L/R speakers need to be exactly positioned to form the base of an triangle. Tape measure in hand, find the middle of your listening area then ensure that speakers are an equal distance to each side and that they are the same distance from the rear wall.
Place your seat on the center line, it can be close in or far away and results will vary slightly depending where the seat ends up. I use a laser alignment tool positioned at the front of each speaker to aim the speakers at the center of the chairs back. The chair can be quite easily move forward and backwards to determine if there is much variation in your room and speaker setup. As a rule, if you cannot determine exactly where the speakers are from listening then you have achieved your goal. Of course, some music will be better, or worse, at demonstrating this.
I find that starting with this configuration then utilizing any Room EQ tools you may wish to use will arrive at results that can be hard to better.

 

[MediumRare]

Empirically: Measure bass at MLP and move around main speakers to avoid any nulls.
Not empirically: pull speakers away from front wall, avoids boominess and provides depth to soundstage. Someone else can advise about how far is too far (something to do with 1x a wavelength to avoid being anti-phase). Also, isn’t being on the long wall of a rectangular room better than the short wall, ceteris paribus? Degree of toe-in depends on the evenness of dispersion and also any harsh peaks - some speakers reward less toe-in to get more even response and more soundstage. The FR could be empirically measured with REW, the soundstage perhaps not.

 

[audio2design]

AMROC is nice online tool for calculating room modes: https://amcoustics.com/tools/amroc?l=40&w=25&h=18&ft=true&r60=0.6

It is not just about getting the speakers away from walls to reduce boundary reinforcement and high magnitude first reflections, but also pick, if you have the space, where the room modes won't be awful.

The 60 degrees triangle is actually part of the AES recommended listening spec, though not so accurate in other recommended specs. You won't want the speakers too close together or you lose the advantage of head shielding (the other ear).

You can use side reflections and diffuse to give a sense of a bigger space, but then imaging gets fuzzy.

It's all trade-offs. Where's Whiner when you need him?

 

[Doodski]

Pretty much everything I've ever read over decades about speaker positioning for hi fi stereo amounts to some variation of:

--Place the speakers such that the listening seat is at the apex of a triangle

--Toe-in according to taste to improve imaging

--Don't put too close to the corners, side walls, or rear walls. Space helps imaging, proximity helps bass. But don't go too far in either extreme.

--Play around with it until you find the goldilocks compromise.

And that's about it.

So, if one isn't using room correction, subwoofers, or EQ, what's a measurement-centric approach to stereo speaker positioning?

Go to the local cannabis retailer, get some indica, smoke a measured amount, pour a finger or two of rye and adjust for best sound.

 

[watchnerd]

Go to the local cannabis retailer, get some indica, smoke a measured amount, pour a finger or two of rye and adjust for best sound.

That doesn't sound very empircal.

 

[Mnyb]

I tried the same stuff as you trying it bye ear , also curios to if there is a better method. any chump can handle ?

Observations:

One are more constrained in placement than one might think when starting out, even if I'm blessed with a wife that says "speakers is your hobby get the ones you need" Entrances and windows makes up where you have to put furniture (it is our living room) . In a not super large living room (US citizen's may have more space) The sofa places itself in the only location where it fits and then the speakers are at the opposite wall to that. If your lucky it's the long wall. You are then faced with the next bad decision windows behind speakers or behind sofa ? Rooms have windows , I prefer to look out when seated.

Due to sizes of typical apartment's I've had MLP is always close to back wall , there is simply not room to drag the sofa or chair out a couple of meters. I rarely had such large living room that you can walk around the sofa coffe table comfy chair group you typically have in a living room .
In the one room where i had such space i was of course forced to an asymmetric speaker placement and had other problems.

Your placement is acoustically asymmetric anyway , I for example have a large opening to the next room at one side and a smaller door to a small room at the other side and different wall material. I see that in the measurement i get in the room DRC i use L & R does not have exactly the same modes.

Simple room mode calculators seems to favor odd fractions of the room distances for placements like 1/3 2/3 or 1/5 or 2/5 or similar.
So I do that. a good starting position for experiments for me is something like the speakers at the long wall placed at the 1/3 and 2/3 of the length.
And at 1/5 of the depth and then juggle it from there .

The other compromises includes getting speakers away from other furniture especially nothing close to where they radiate .

Wonder what i shall do this xmas there might be small improvement's to be had .

Q:

You always have some nulls. Can they be placed where bass instruments does not have fundamentals or harmonics ?
I was thinking that it does not matter if a have null if there never is any content played there or are drums and other percussive instruments to broad in their spectrum ? or piano, seems to be a big resonant box to me ?

 

[Eetu]

--Don't put too close to the corners, side walls, or rear walls. Space helps imaging, proximity helps bass. But don't go too far in either extreme.

Here's a handy cheat sheet on SBIR from Neumann:

This is also what Genelec recommends in terms of SBIR, placing the speakers either as close to the front wall as possible or over 2. 2 m away. Or over 1.2 m away if crossed over to subs.

The main thing is that the speakers are symmetrically placed.

I would take measurements to see how (non)destructive the SBIR is in your case.

 

[MakeMineVinyl]

How does REW or any RTA help assess imaging?

Only to the extent that the frequency responses of each speaker is the same or as close as possible

 

[watchnerd]

Here's a handy cheat sheet on SBIR from Neumann:
View attachment 99583

This is also what Genelec recommends in terms of SBIR, placing the speakers either as close to the front wall as possible or over 2. 2 m away. Or over 1.2 m away if crossed over to subs.

View attachment 99586

The main thing is that the speakers are symmetrically placed.

I would take measurements to see how (non)destructive the SBIR is in your case.

Are the distances in those charts measured from the front of the back of the speaker? Or the middle?

And how does it change if there is a rear-firing reflex port?

Obviously you can't put a rear-ported speaker flush against the wall.

 

[tw99]

Are the distances in those charts measured from the front of the back of the speaker? Or the middle?

And how does it change if there is a rear-firing reflex port?

Obviously you can't put a rear-ported speaker flush against the wall.

The distances would be measured from the front of the speaker.

 

[watchnerd]

The distances would be measured from the front of the speaker.

Why, if the bass is omni-directional and there is a port in the back?

 

[Mnyb]

Well some bass from the port and some from the driver ? With a low tuned speaker maybe port freq is below some of the room interaction ?

 

[watchnerd]

Well some bass from the port and some from the driver ? With a low tuned speaker maybe port freq is below some of the room interaction ?

Errr...in any normal size domestic room, a bass port will be tuned below the Schroeder frequency, i.e. in the room dominant zone.

In my living room, the usual calculation gives a Schroeder frequency of 126 Hz.

The port in my speaker is tuned to 32 Hz, so about two octaves below the Schroeder frequency.

 

[MediumRare]

The distances would be measured from the front of the speaker.

That applies only to the Gene;c speakers which happen to have a front port. Rear-firing is entirely different as the sound from the front has to travel the depth of the speaker twice but from the rear only once. This is relevant for phase alignment and nulls, not for directionality.

 

[watchnerd]

That applies only to the Gene;c speakers which happen to have a front port. Rear-firing is entirely different as the sound from the front has to travel the depth of the speaker twice but from the rear only once. This is relevant for phase alignment and nulls, not for directionality.

For a rear-ported speaker, without trying to get fancy and calculate different travel paths for different frequencies, maybe the lazy, but good enough, approach is to just measure from the middle of the speaker.

 

[Jimshoe]

I think I am missing something in the Genelec approach. The diagrams appear focused on nulls (as a result of comb filtering), but surely there is a reinforcement element to consider too? Thus the diagrams seem only to tell half the story.

Or have I misunderstood?