Tim in S. California:
The procedure I've utilized with several pairs of speakers is to measure bass response (from 200 Hz down) with various speaker placements to identify the smoothest bass response. A smoother overall bass response is more important to me than extreme bass extension. I use the Stereophile Test CD #2 and my calibrated Radio Shack analog SPL meter. I understand the RS meter is not the most accurate device but I'm not worried about exact dB readings as much as the relative readings. And for whatever it might be worth, before I retired I calibrated my RS meter by a test signal with the B&K meter we had at work. The SPL meter is placed on a tri-pod at my ear level in place of my sweet-spot chair. An 80 dB level is chosen as a balance between lower levels which may have fall off and louder levels which might emphasize bass tones.
My procedure is to position the speakers under test for separation and toe in based on experience, then begin with the speakers at a minimal distance from the front wall and record the SPL for each bass frequency, move the speakers 2" further into the room and measure SPLs again. I adjust the gain for 80 dB (at the listening position) with a 1K reference tone, then record the SPL reading for each bass frequency for each third octave from 200 Hz down to 31 Hz. I continue this as I move the speakers in 2" increments further into the room. Once I've covered a range from maybe two feet to nearly four feet I then go through the measurements for each distance and add up the variation (ignoring positive or negative numbers) from the 1K reference measurement. Every time I've done this set up evaluation I see the summed variation diminishes up to a point as I move the speakers away from the front wall, then it begins to increase. This tells me the distance that approximates the smoothest bass response curve (minimal variation total). With all my prior speakers I would then move them 1" closer and again 1" further back than the best 2" measurement position to fine tune bass response even further. As I'll cover, this last step was not necessary with the zA2.1s.
OK, that is the background. Now on to specifics.
David while you recommend an equilateral set up, that is not possible in my room, at least not within reason. The room is quite large but has an irregular shape. It is 20' long and averages something like 16-17' wide. The ceiling is flat and 11' high. There is only one practical area to locate speakers. Facing that area the wall on the right side is straight and unbroken front to back. The left side is irregular with a closet built into the room, an alcove (5' wide where components and records are placed), then two more protrusions to enclose a stairway and a column with a fireplace. The front wall (behind the speakers) has a wide sliding glass door and the main entry door. That width between the right wall and entry door area is 11' 6", the width where speakers must be placed.
Pushing the width of the zA2.1 to the maximum, considering distance from the right sidewall, I elected 7' 11" separation between centerlines. This resulted in the right speaker being 24.5" from wall to centerline. I know you recommend a minimum of 3' and that would be my preference but moving that speaker inward would sacrifice separation between channels. I began my testing with the front inside corner of the base of each speaker 24" from the front wall. I then moved the speakers forward in 2" increments to a distance of 40" out from the wall, playing the test tones and recording response levels for each one. I was strict to maintain the same distance for each speaker from a center line reference as well as the same toe-in angle (inside edge of each speaker barely visible), for each distance from the wall. The zA2.1 was more complicated than other speakers I've measured this way because to the woofer level control. I elected to test only for the 0 dB and +3 dB positions, ignoring the -3 dB position. So including setting the reference level tone for each distance I measured 10 tones by 2 woofer levels by 9 position distances, a total of 180 measurements!!!
But I believe it was worthwhile. I learned two very important facts about the Janszens. First of all, the totals for the variation numbers did not show as much difference with position change as other speakers I tested in my room*. This tells me the zA2.1 design may not be as critical for room placement as many other speakers. Certainly they can be fine-tuned but a more casual placement by an average consumer should provide good performance unless they do something really silly. The second lesson was seeing the totals for variation numbers over the measured distances were significantly lower than those for other speakers. This suggests to me the zA2.1 produces a flatter in-room response than other speakers I've measured. Now I'm not saying a perfectly flat response at the listening seat would necessarily be desirable, but at least I'm assured there are no major peaks or valleys in what I'm hearing. It is even more impressive when I admit I included measurements down to 31 Hz with the Janszens while I ended with 40 Hz for all the other speakers. With other speakers I threw out the 31 Hz measurement because it was always too signifiant a dip.
Now my specific finding is less important since every room and system will measure somewhat differently, but I ended up with the front inside of the base for each speaker 28" out from the front wall. Also, I get a smoother response with the woofer control set for 0 dB at that position.
I realize this is somewhat lengthy but I felt to give credence to my two observations of significance it was important for you to understand how I arrived at my findings. Now I'll enjoy my revised set up and possibly play with toe-in after I'm acclimated to this sonic result.
* Previously owned speakers measured in this fashion include Duntech Princess (3-way, 5 driver, acoustic suspension), Cain & Cain BEN (HE with rear loaded horn and external super tweeter), and Magnestand modified Magnepan 1.6 (hardwood frames, reversed panels, revised crossover with high quality parts).