Cool topic!
 
For a variety of reasons that I'll spare you all, I have been working in a truly horrible environment for quite a while now, but I was really focused on composing, arranging, and just generally developing a new work-flow, so I tolerated it.
 
Now I am still not in a position to build my next purpose built studio, but I did need to do something, so I did.
 
My room is not yet treated, and yet I was able to place loudspeakers and ears such that I have a great stereo image, and smooth frequency response to the limits that my monitors can produce. As I've learned one too many times, placement will get you closer to usable than almost any other factor. In fact, I think placement is the first step in room treatment.
 
So, guidelines that I've used many times to optimize placement.
 
Start with ergonomics. Unless you have software that can accurately model a small critical listening space you have to start somewhere, and this is as good a place as any. Place your keyboard, computer, and other gear someplace that makes sense, and arrange them so that you can use them.
 
It isn't quite that simple, there are some rules of thumb that you'll want to follow as you set things up:
 
1) symmetry is critical. It isn't always possible, so if you have to make concessions make them so that your ears are equidistant from the walls. In my case that meant moving everything back so that the bump out would not directly affect reflections from the side walls. Note that this violates the rule of thumb that suggests that your ears should not be half way between any two boundaries. While this is true with respect to low frequency problems, it is still better to keep this symmetry in tact - it preserves the stereo image, and at higher frequencies you won't have problems with room modes.
 
1) symmetry is critical. (no, that's not a typo) The loudspeakers must be equidistant from your ears.
 
2) distance behind your head is more important than distance in front of your head. Ideally you want at least 12 feet from the back of your head to the back wall. I was able to do that, but you may not. If you can't, the try to place your ears somewhere between 1/4 and 1/3 of the way from the front wall.
 
3) try to place computer monitors, or anything else that will be in the same vertical plane as the loudspeakers so that you don't create obstacles.
 
4) with the exception of left/right symmetry at your ears try to avoid placing the loudspeakers halfway between boundaries. Lateral should be easy, since you have two loudspeakers<G>, front to back can be a little trickier.
 
5) the whole bit about placing the loudspeakers at ear level is, I think, a myth. Sometimes you have to place them higher (or lower) to get rid of bad early reflections. I usually go for higher, and angle them down, but I have worked in rooms where they were lower, and angled up, and that works too.
 
6) then there is the question of orientation - tweeter above woofer vs tweeter aside woofer. They can both work, but in small rooms the aside configuration seems to work better - for me. Even tweeters on the inside vs outside will generate debate, I prefer them on the outside. Again all of these are starting points, and it takes only a few seconds to change orientation as a test.
 
7) now to your original question... try to place the loudspeakers somewhere between 8 and 12 times the diameter of your low frequency driver - for a 6-inch driver that'll be 4 ft to 6 ft. It isn't critical that you create an equilateral triangle, again it is a good place to start, but it is not critical, and in fact I've not found it to work any more often than not. As a rule, the distance between the speakers will have a large impact on the stereo image, and you'll just have to play with that distance till you get your optimal image.
 
Now listen to some of your favorite recordings. Pay attention to the sound stage and the low frequencies. One or both may exhibit problems.
 
Even if you LOVE the results of your initial placement you'll want to move things around a bit, to verify you got it right if nothing else.
 
Once you have optimized loudspeaker and ear placement you can start to think about treatments.
 
Hope this helps...

The problem with putting the speakers back further is that I already have a computer screen mounted on an arm swivel stand, floating above my mixing desk. I try to make sure the screen isn't protruding in front of the front bezel of my HS80m's. While I realize HS80m's may have not been the best choice being rear mounted, I am pretty used to their sound having used them for 3 years now, so I'm more concerned with where to put my computer screen in this whole equation since there's no way I could see anything on a computer screen sitting 5 feet away.
 
If I take Bitflipper's equation above, the distance from the center of the tweeter to the center of the woofer on the HS80m is only like 8-12''. So does that mean a 2'-3' distance from HS80m to ear is okay? What about the minimum distance, 2'? (This would let me get them out from the walls more if so.) In that scenario I'd have about 2-3' from me and 5' in between the speakers.

Again we're talking about what's supposed to be nearfield monitors here right? We're also talking about bedroom studios where people don't have half a house to spread their gear around. What about a 2.5x2.5x4.5' or so triangle? Trying to push from the wall more would simply put me in the dead center of the room and waste a lot of floor space. I'm in a 9'x15' or so room and I've been dead center short ways, I'm turning the entire setup and getting treatment to try and fire longways down the room. Pushing out from the wall more than 1.5' would defeat this purpose. Is 1-2' from the wall enough? The HS80m manual says 5', which is insane. I'm looking at getting a proper treatment setup from Gik Acoustics as soon as I can afford it with like 3 giant monster bass traps, 8 wall panels and a cloud or two. (Cloud is going to be a head bumper as I'm only at a 7' ceiling as is. Old house. Upstairs bedroom.) but the mixer and monitors is a real chore to keep moving around so I'd like to be a bit more sure on this equallateral triangle thing as well as wall distance.

I stll feel not enough is spoken about the correct distance between monitors. It is too easy to fill up the distance betwen the speakers with computer screens and such and not enough attention is put into distance between monitor speakers. In my opinion it is important.
 
When the distance between the monitors is excessive you get a poor centre phantom image. It will make you push things pannned centre up too high in order to correct for it. On other systems you will then hear the centere panned stuff too loud. Too close and you get an excess of the phantom image and it will be slightly loud making you pull centre panned stuff down a little to compensate.
 
Here is a simple experiment to get this right. Try it and you will hear what I am talking about. You cannot rely on any mathematical approach to determine this distance, it is an audbile use your ears thing really. I also feel that speakers or tweeters at ear level is also important. It sounds better to me compared to speakers being up high or down low as well.
 
A good thing to do is to set your speakers up on a long plank at the right height for this test. (that way once you find the distance if you are using stands you can put them back on stands.) You need two friends to do this. Feed a mono signal to both speakers (music pref something very nicely mixed eg Steely Dan) Start with them reasonably close togther eg 3 feet and centered in front of you. You close your eyes and listen for the phantom image in the centre. The two friends gently slide your monitors apart while you listen with your eyes closed. At some point the centre image will break up and fade away. That is the point when they are too far apart. The idea is to slide them back again so the phantom image re appears strongly. (and it magically does!) And maybe touch more in for extra measure. I have done this and this all happens at 1.1 meters for me and my room.
 
You should do this test first and then see if two computer screens will fit in between. (if you have two screens that is) They may not. If they don't and you have to widen the monitors to fit them in then this is a compromise. You have weakened your centre image. Everything sounds better and more robust when the monitors are the right distance apart. This is a very important aspect of monitor placement. I have got three screens actually but only one in between the speakers. The other two are on the right of the right speaker and it all works well.
 
 
 
 

I have two screens, but I chose to mount one above the other so as not to disturb the distance between monitors.

Jeff,
 
That makes good sense, except one problem I'm dealing with a Mackie 32*8 mixer between the speakers.
 
I am getting a feeling while listening the speakers need to sit a bit closer together, but they'd have to float above the mixer to do that. The mixer sits on a table 28'' from the ground. The table is the exact width of the mixer.
 
Ok, so let's forget computer screens for a second, suppose I abandon the idea of being able to use the computer screen/mouse/keyboard while looking in the general direction of the mixer. I'm ok with that for several reasons actually but even when just dealing with a mixer, it's on a table 28'' off the ground. Not sure how I'd physically get the monitors in closer to create say a 3' triangle... as the mixer is blocking the way. I mean, what else can I put a mixer on aside from a table to allow room for monitors to sit closer together? :p

I could lower the mixer by setting it on a sturdy keyboard stand, but still couldn't get the speakers in closer? Would it be a good idea to lower the mixer 5 or 6'' by getting it off the table, regardless?
 
So I'm stuck with a super huge 6' triangle that takes up my entire room and leaves a dispersed stereo image, or, an isosceles triangle that also has a dispersed center image? I'm not the first person to use a mixer, on a table between 2 nearfields that need to sit on stands or magically float - is there an elegant solution to this? I'm much prefer to get them closer in, somehow, if possible.

Yes it all comes down the the question as to whether you need to have the mixer directly in front while facing the speakers. In the end I ended up putting the mixer to the right and getting the speakers and computer set up directly in front as the main priority.
 
Often there is not that many moves you are making on the mixer during a mix down. (or is there?) You tend to spend time working with the computer and using automation and things while the mixer could be regarded as more static and summing in its purpose.
 
I had a larger format console like that too but replaced it with a much smaller and smarter and way better digital mixer. It does way more and is around a third the size. I have some external midi synth hardware which I just send to a nice small footprint 16 channel mixr that sits on top of a rack. That then feeds into the digital mixer.
 
You have got to ask yourself what do you actually do with your main mixer while you are mixing down? If you think you need the mixer for its sound that is a myth as well. You can replicate the sound of very nice analog consoles and mixers way better than the Mackie as well such as Harrison Mix Bus sounding like a Harrison console. I always found Mackie mixers as being rather transparent (which is good) but not actually adding any important colours to your tracks or stems. You can do that these days much more so ITB.

those pesky computer screens!
 
Jeff presents an excellent technique to find the optimal space between the loudspeakers. Except it isn't quite that simple.
 
We are dealing with a geometry problem with several variables:
1) the distance between the loudspeakers
2) the distance from your ears to the loudspeakers (this can be measured from each ear to each loudspeaker, or from your nose to the line that connects the loudspeakers)
3) the distance from the loudspeakers to the room boundaries
4) any and all surfaces (console, screens, racks, etc) that can reflect sound
5) the absorption of the boundaries (and this will be frequency dependent)
6) the absorption of everything else in the room, including you, the listener, and of course any treatments.
7) the geometry of the room
8) the size of the room
9) the transmission loss across the room boundaries (another pesky frequency dependent factor)
 
(NB: note that we do not talk about the reverberation time of the room - small rooms are not actually reverberant.)
 
So, that's a mess of variables! You can typically change only a few of them, most likely those affected by the placement of the loudspeakers and the ears.
 
So Jeff's test is an excellent one, although I'd make a small change to the method. Find a recording that has a nice stereo image, and make a copy of that some by adding the two channels together (most audio programs will let you sum to mono, usually with a 3dB or 6dB attenuator inserted into each channel.)
 
Place the loudspeakers at some arbitrary spots that address as many problems as you can - your starting point - and then play the mono version of the recording. Pay attention to where the 'phantom' image sits, and whether it sits still or not. Now listen to the stereo version and compare. Now move the loudspeakers closer together by about a foot and repeat. Did the image change, and if so did it get better or worse? Now move them back to the original position and repeat. Now move them a foot further apart and repeat. And so on. One of these three positions will sound better than the others. Use that as a starting point and repeat the whole process again, and maybe even again.
 
IF there are limiting factors - furniture, monitors, a mixer, whatever, use that to your advantage. If you want to have two 24 inch monitors in front of you then you are going to have to start with the loudspeakers at least 2 feet (approx) apart. Sometimes limitations can make the job easier!
 
As you do this you also want to change the distance from your ears to the monitors. Start with an equilateral triangle for no other reason than it is easy. As you conduct each test move forward and backward, again in 1 foot increments. For each test the best combination is the one you want. And yes, there may be multiple combinations that work in your space. That leads us to...
 
Now mute one loudspeaker at a time and listen to swept sine waves or pink noise or even music - what you are looking for here is frequency response problems, excessive bumps in the curve. The nice thing about using your ears instead of an SPL meter is you won't go chasing the needle - you only have to trust your ears. I find swept sine waves to be the easiet source in which to detect problems, but once I find problems I still listen to noise, and then music to make sure that the problem is real.
 
For the most part, in small bedroom type spaces, it has been my experience that it is pretty easy to pick the optimal location for the stereo image, but that this often leads to frequency response problems. And when you solve those you usually end up collapsing the stereo field a little. Your priorities will determine which problem to favor.
 
In larger spaces, such as a basement or garage it is usually much easier to find a configuration that minimizes response problems without sacrificing the image.

Don't forget that loudspeaker orientation will also play a role, so once you find your sweet spot don't be afraid to try different orientations.
 
As I mentioned earlier, I recently (finally?) went through this exercise in my own basement studio. The stereo image is excellent, and I have no annoying bumps in the low frequency range. In fact I think I got a little lucky, and the early roll-off that I hear at the bottom is in fact the roll off from the small monitors. I have a set of larger monitors that I will start experimenting with as soon as I replace the surrounds, which finally rotted.

This exercise can be done in an afternoon - maybe a long afternoon depending on your room. And tolerances are forgiving, if you get the loudspeakers within 4-6 inches of that target you'll have a valid test. You can then get out the tape measure and place things exactly. You may not even hear a difference between approximate and exact locations.

The other comment that was made earlier that I'd like to underline is levels. In a smaller space you can overdrive the space itself, definitely take that into consideration. And since you won't be using fancy test gear, perform the tests at a level similar to the loudest level you are likely to work at.

Oh yeah, and have fun!

Bill's points are good. I am sorry I did not mention it but my test does work better in a situation where the speakers are out from the walls and in a relative symmetrical position in relation to the room as well. I should have mentioned that. Also it is good to get a little closer in also do it a reasonably lowish volume too as Bill suggested.
 
The reason why I like the speakers close togther and roughly 3 feet apart and you sitting that sort of distance out and at a lower volume is that it is possible to remove the room acoustics to a certain extent and they don't have to be as involved as per wider and further out with the listening positon and louder volumes too.
 
 

I agree with Jeff on the 3' wide placement.  I only have 1 22" monitor between them and my listening field is really open.  I have sound foam behind the speakers and they are about 6" from the wall.  My keyboard is to the left.

With all this talk I have been doing some checking again and I have decided to move my monitors a few inches closer to me and also an inch or two closer together.  I noticed that by bringing them even very slightly closer together the phantom centre image is now a little more pronounced and robust.  Interesting I may have been just on the cusp of it still holding together but perhaps should have brought in them slightly to reinforce it a little more.
 
Also by moving the monitors an inch or two closer to me I have noticed the bottom end is a little stronger and they seem to be coupling a tiny bit better creating a slighty more powerful and robust bottom end too.  The sound is louder and slightly tighter inside the triangle now and there is slightly less of the room sound involved.
 
All very cool.  So my distance apart is now 102 cm or 40 inches.  I am sitting the same distance away from both of them.
 
This slight adjustment has really reinforced the idea in my mind that distance apart is almost critical.  Once you start spreading them too far apart they are just not in the same league compared to when they are a little closer.  Some are in way too much of a hurry to dismiss the importance of this. They are simply wrong.  It really works and sounds different bigtime when you get this distance right.   It is well worth experimenting with this stuff and getting this right. The priority should be on speaker distance apart first and foremost (after room symmetery and distance from wall of course, the stuff Bill has mentioned previously of course) and then fit other things in once you get this right.
 
I have also re-arranged my 27" screen so it is a little behind the speaker axis now. That seems to have changed the mids and highs ever so slightly too improving clarity. It means the screen is now probably less involved with secondary reflections from the front speakers.