Quote:
Originally Posted by
unqlenol
➡️
Guys, thanks again for all.
Time is short and before I know it my demo will expire. So, please help me out a little more here, and hopefully I can produce something worth sniffing at:
Using Reflex has beenuncomfortable compared to QR Dude. It feels totally 'freeform' with no guidelines towards 'optimising' a diffusor. Feels random. Like you can just fiddle forever and hope for the best.
I did skim through the manual and it did not seem to give practical application advice to help one arrive at an optimised design.
So, my questions for now are:
1. Where do I start? Should I just decide on a depth, well width and period width and just fiddle randomly with well depths?
2.I am not sure I know how to interpret the polar response graph. What am I looking for?
2a. Do I want my selected freqs projecting at as many angles as possible (ie: many 'lobes' or 'spikes')?
2b: Do I want the projections to all be of as high an amplitude as possible?
What say you all?
Thanks
Nolan
Hi Nolan,
In answer to your questions:
1) SETTING UP THE GEOMETRY IN REFLEX
1a) Start with the simplest depth sequence I've provided in the blueprints, which is [0 cm 4cm 5cm 3cm 5cm 4cm 0cm], and use a well width of 6 cm. This is your basic module. (make sure reflex is entering the well depths in cm... or, if reflex is in millimetre mode, remember to multiply all these numbers by 10)
1b) You want 4 of these modules, so make 3 copies of your basic module. You should now have 4 modules side by side.
1c) Now, you want to test how to mount the 4 modules at different depths. You can start by testing the modules, mounted at depths [0 cm, 5 cm, 6 cm, 0 cm]. To do this, you would add +5 cm to the height of each well in the 2nd module, and +6 cm to the height of each well in the 3rd module. Modules 1 and 4 would stay the same, as they are in part 1b.
So, the resulting shape in reflex would consist of 28 wells (4 modules x 7 modules/well), with the following well depths:
[0, 4, 5, 3, 5, 4, 0, 5, 9, 10, 8, 10, 9, 5, 6, 10, 11, 9, 11, 10, 6 0, 4, 5, 3, 5, 4, 0] cm
2) POLAR RESPONSE PLOTS
You can basically ignore the polar response graph, as it only gives information at specified frequencies. But if you're curious, an ideal polar response would look like a semicircle. The more uniform the polar response looks, the better the spatial diffusion is for that particular frequency. By uniform, I mean that a semicircle would be perfectly uniform. On the other hand, if there are several large spikes with a high dB value, with big gaps in between that have a very low dB value, then the polar response it is very non-uniform (poor diffusion at that particular frequency).
But these polar plots can lead to much confusion. So to assess the performance properly,
you want to focus on the graph of the normalized diffusion coefficient.
Does this make sense to you? I hope this helps!
Tim