The idea of 1:1.26: 1.6 ration is to distribute the room modes 1/3 octave apart - the 1.26 dimension will have a standing wave 1/3 octave lower than the 1 dimension, and the 1.6 would be 2/3 octave lower. This distributes the peaks and dips more evenly.
The same argument applies for spacing the speaker from the nearest walls. The floor bounce, side and back wall bounce dips will be more evenly distributed.
My more recent suggestion of 1:2:4 ratio aims at the boundary bounces to cancel. At the frequency where the shortest distance cause a dip, the bounce from the 2 will cause a peak at that same frequency, thereby nulling out (one cancels, the other reinforces). The dip from the 2 is nulled out by the bounce from the 4. Then by arranging the dip for the 4 to be below the crossover, we will have no dip whatsoever.
Originally Posted by rbpeirce
Years ago when I bought these speakers I followed your 1:1.26:1.60 placement recommendation, except I used the woofer height, 33.5", as the short dimension and set them 53.5" from the back wall and 44" from the side wall.
Now, I am re-arranging my room and got to wondering whether there might be some reason why you suggested the other setting. Is there something that makes it better?
On a related issue, when I designed my listening room (25 years ago!) I used a ratio of 1.62 all the way around, having read somewhere that this would greatly reduce room peaks and nulls. I turns out I have a sharp dip around 32hz but the room is otherwise pretty flat. Therefore, I wonder if you know how 1:1.62:2.62 would compare to 1:1.26:1.60, and in either case, whether speaker height would be best at any one of the three ratios?