Let us take another look at the perceptual problems involved in making engineering decisions regarding baffle (and cabinet) construction based upon objective measurements. In the case of the two-sided baffles (at 2m distance) it has been shown that baffle thickness and edge rounding are determining factors in the linearity of the frequency response - in addition to the baffle width and shape.
The new data on one-sided baffles provides an even better example of edge effects, in that we now have 3 meter data with which to compare instances. Three meters being a more likely listening distance than 2 meters.
Below are three comparisons of exactly the same data. The three meter response from the rectangular one-sided baffle just 12 inches (total) wide. In the first case (WHITE) the 'free' or 'bare' edge of the RD75 driver is untreated; in the second case (BROWN) a 1.5 inch radius half-round section has been built up on the mounting flange of the driver. In both cases, the baffle width is approximately the same, the half-round section extends no farther out from the driver than does the mounting flange.
The first comparison is without any smoothing and represents the actual response of the drivers (consistent with aquisition and processing parameters). The second and third comparisions are 1/12 octave and 1/3 octave smoothed respectively. In the unsmoothed comparison we note substantial WHITE behind the BROWN - this represents a reduction in the frequency response 'envelope' (in the higher frequencies), the same as we observed when applying the chifon fabric to the front of the RD75 driver.
We take it on principle that any reduction in the 'variance' of the frequency response envelope must be considered a positive step toward the search for the optimum RD75 baffle - hence full-width rounded-edges on the baffle must be used. That is, until something better comes along.
As always, the question is 'Does it make it sound better?' Unfortuately the answer is "I don't know". There is just not enough time for listening comparisons for such small, but important, details. We must look at the comparisons and judge for ourselves whether the effort in building up the half-round on the edge of the RD75 worth the effort. Aesthetically, the answer is most definitely yes! Would 'Golden Ears' hear the difference - well of course! How about you and I? Sadly, I will never know because of the lack of time for critical listening to such details. Objective measurements are the best method I have at present to process such large amounts of change.
Microphone at 3 meters distance and a height of 1 meter.
MLS: 32 averages, 432/4096 point FFT, Half-Bingham window.
No, 1/12 & 1/3 octave smoothing - Amplitude Axis @ 6dB per division.
Comparion: Old & New Test Setup.
Table of Contents - One-Sided Baffles.
Table of Contents - Main Menu.
