Five feet from the front wall

The distance from the front wall is going to effect the mid and upper bass/lower midrange. If the face of the woofer is three feet from the front wall, the distance to the wall and back is 6 feet, that is the wavelength of about 185 Hz. Thus 185 Hz and the frequencies around it are going to be reinforced. As you move the speaker closer to the front wall the reinforced frequency increases, as you move away the reinforced frequency decreases. At five feet it is about 100 Hz. Other frequencies are attenuated. None of this is good. Sound absorption on the front and side walls is very important but it is only effective above 200 Hz at best. If you could set up the speakers 10 feet from the front wall you could reinforce 55 Hx. 28 feet would reinforce 20 Hz. All this is impossible in residential situations. 

All the above is the reason that digital room control is so important particularly in the bass.

Any other changes in sounds quality with speaker positioning depend on the timing of reflections which depends entirely on the room. These reflections effect fine detail and imaging. If reflected sound can travel 100 feet before reaching your ear you hear an echo. Thus the more sound absorption you use anywhere in a room the more you are going to decrease the likelihood you will hear an echo, which IMHO is a good thing as the echo was not in the recording. Some people actually like echoes as they add a false sense of spaciousness. All the echoes of the venue where the recording occurred are on the record. Studio recordings frequently have echo added to them to create dimension. Live recordings generally do not need it.    

Sorry for preaching.

@pryso 

I have been using dipoles exclusively since 1979, I currently use 8 foot Sound Labs ESL and they are between 2 and 2.5 feet(towed in) from the front wall. The wall behind a dipole should always be heavily deadened. I use 4" thick acoustic foam tile, but that is all you have to do with them other than use subwoofers below 100 Hz. Dipoles have a lot of trouble with the longer wavelengths. 

I hate to disagree with Bell Labs, but that is a purely subjective evaluation. Any sound within 100 ms of the direct sound is sound that was not in the recording and by definition is distortion. After 100 ms it is an echo or a different sound altogether. Dipoles are 180 degrees different than other types of loudspeakers because the sound that comes off the back is 180 degrees out of phase with the front wave which means that at three feet from the wall 185 Hz is attenuated. At 5 feet 100 Hz is attenuated which is not good for bass. It is a lot more complicated than this as again other frequencies are reinforced. I keep my speakers closer to the wall because higher frequencies are easier to manage with acoustic absorption. Below 200 Hz there is not much you can do other than digital room control.

The issue that causes such varied opinions is that some distortions sound good to some people. They can add warmth and ambience to a recording. Being the stubborn purist that I am I only want to hear what is on the recording. It seems to me that like @rauliruegas I find that systems that are devoid of most distortions sound better.

@markalarsen 

That would be a mistake. When seated at your listening position the front wall is in front of you and the rear or back wall is the one in back of you.

@ditusa 

There is much more to imaging than lateral position, there is image size, location in depth in the sound field and the third dimension which is the sense that the instrument is a 3 dimensional object in space. Early reflections will disturb all of them even if lateral location is not affected. As far as stability of the image is concerned, especially with point source speakers, The image is extremally fragile when out of the listening position even by a few inches. 30 degrees or more off axis you can even lose the far speaker completely. Because lose of volume with distance is much less severe with line sources they create a more stable image and you never lose the far speaker of axis. The image is still perfect only at the listening position. This is why Line Arrays are used exclusively at large concerts. We were at the Arctic Monkeys concert at Red Rocks last night and even the subwoofers were linear arrays just outside of the main Linear Arrays. 

@ditusa 

Yes, a good horn can have almost the same frequency response way off axis then they will fall off dramatically which is an advantage as it limits room interaction. But, that is not the problem. Horns are point source drivers and like any point source driver the volume level drops off rapidly with distance. The center image shifts like any other point source driver and you can still lose the far speaker far off axis. 

With a line source array I can stand 3 feet away and right in front of one speaker and still hear the other.

@invalid ,

I had Apogee Divas for 6 years. It does not matter how far off the wall you are. If it is a hard wall behind the speakers sound deadening will always improve image specificity. The farther away from the wall you are the more of the wall you have to cover. Closer to the wall, like 3 feet and you only have to do right behind the speaker. This only control frequencies over 250 Hz. At three feet you will attenuate 185 Hz which will take some of the snap away from the sound. If you move it farther away you will attenuate lower frequencies. At 8 feet you are looking a 16 foot wavelength 180 degrees out of phase and that would attenuate  70 Hz. Unless you are using subwoofers (which I highly suggest with those speakers) you are missing a lot of bass. It is a compromise any way you look at it unless you want to move them to the front lawn.

@waytoomuchstuff 

No. With a point source the volume drops at the cube of the distance. With a line source it drops at the square of the distance. Volume drop off is an order of magnitude less. As for height that depends on the frequency. For a driver to act as a line source it has to be taller than the longest wavelength it is to reproduce with one very cool exception. Low C has a wavelength of 56 feet. A little hard to get a 56 foot tall speaker into most houses. BUT, if the ends of the line source butte up against barriers like the floor and the ceiling, the line source becomes infinitely long as if the floor and ceiling were mirrors.  This is the rational behind 8 and 9 foot tall speakers. In the case of ESLs it turns a modest gentleman into Axel Rose. Add a line source subwoofer array and bring on the concert! Line Source systems have a very different presentation. Everything is larger, front row vs rear of the hall. The image is more stable and the overall sound more comfortable as in less prone to sibilance and distortion at high volumes. Many people prefer Point source systems but usually not because of the sound, more likely the size of the speakers. I had a new friend over who is a Bricasti dealer. He listens to Franco Serblin speakers, very fine Italian point source speaker. His first comment was, "I'm not use to this! There is so much bass!" The system runs flat down to 100 hz then rises at 3 dB/oct down to 20 Hz. The treble falls off from 2000 Hz at 1 dB/oct. This is very standard target curve for most rooms. Dirac live usually starts with this curve. Most systems do not project much under 60 Hz. The frequency response specs are taken at one meter. 

@paulrandall 

I stated, I thought clearing, that the reference is the face of the driver, not anywhere else. 

@invalid 

It makes not difference whether or not you are direct driving ribbon speakers although you have to be careful. The tweeters have VERY low impedance, < 1 ohm and can burn up some amps. All planer dipoles benefit from subwoofers. It is not a mater of speed. Speed has nothing to do with it. Lack of spurious resonance's, time and phase coherence along with the right crossover point have a lot to do with it. 

@lanx0003 

That is the wrong way to look at it lanx. Sounds that reach your ear after 10 ms are perceived as an echo. Less than 10 ms and they are perceived as one with the direct sound. That does not mean these sounds are not distorting what you are hearing, they are in a big way. They are sound that was not present on the recording and as such are distortion. They change frequency response (amplitude) and screw up phase affecting the image.  

@bdp24 

If is extremely hard to over dampen a room. Most system/rooms that I have auditioned are under dampened. What people think is dull and lifeless is generally much more accurate, they are just use to sizzling hot with echoes and have a hard time reorienting themselves. Turn the volume up, way up to 95 dB. If the sound wants to cut your throat you are underdamped.