My understanding is that beam width plays no part in room compensation? Either will do and then choose wide or narrow and save as a preset? Unless someone knows different?

Kinda forgotten how I did it on the BL90s? The room compensation is set up in a particular mode on the 90s. (i.e Narrow, Wide, Omni). You instruct the 90s what the mode is (1,2,3) being formed. The room calibration (one man on his own in a chair) centres around an imaginary cube around said mans head. Furthest corner, nearest corner, centre of head etc) Different room calibrations can be added together to "widen the area' (two people on a sofa) etc. This can actually be done in Narrow mode!   My ultimate take on the N/W/O thing is in a way that it is a gimmick - of sorts. There is no-doubt the narrow mode calibration is effective in reducing reflections and offers pin-sharp imaging and soundstage but it comes at a cost of high latency (computation time). Wide mode is just a "blurring" or dumbing down of narrow mode (in my opinion)  which reduces the detailed computations required to sufficiently reduce latency and make the BL90/50 usable as AV speakers. Omni is just "let it rip" mode.   So yes, beam width and room compensation are intimately linked. There may be slight variances between the BL90 and BL50 calibration...as one has a single acoustic lens and the other has 14 separate drivers but I imagine the subroutine will be practically identical?  

Hello Beofile and 10%, What I understand in the explanations of the BL50 Technical Sound Guide is that narrow mode minimizes the influence of wall-reflections by narrowing the angle formed by the 2 moving wings placed under the acoustical lens. On wide mode, the wings are aligned so room-reflections are less reduced. So, as 10%, I tend to think that Room compensation and beam width are linked. I'll stick to my plan. Thanks for your help and kind regards, Yann.  

Yes they are...as stated above. The process is the same for narrow mode and wide mode. The only difference is a) the lens adjusts its profile and b) there is (I would guess) some sort of "clipping"/"de-resolution" of the data/algorithm which reduces the latency. I do not think that "wide" mode is caused by the lens position alone.