Previously, I attempted to solve the Lowther DX3’s nasty treble by lining a paper towel over the driver like this.
The results were mixed. While this approach smoothed out the frequency response at 11.25 degrees off-axis, it didn’t do much for the on-axis response. We can see this in the plot below. While paper towel strip did slightly even out the dip at 7kHz, it didn’t do jack to the upper mid and treble peaks at all. Given how the paper towel pushed down the treble (at 11.25 degrees off-axis), the polar response was even less uniform than before! (As an aside, any kind of EQ would be pointless unless the polar response can be made more uniform).
Green = Stock on-axis
Yellow = Paper towel strip on-axis
I knew that a better solution would be required. I searched the Internet and discovered possible solutions such as the DX4’s pepper pot phase plugs and bulb / mushroom type phase plugs. I ruled out the DX4 phase plug based on these measurements which seemed show no better on-axis performance. A few subjective impressions indicated that the pepper pot plug actually made the Lowther drivers even more directional, which is not what I wanted. Thus I opted down the path of bulb / mushroom type plugs. Unfortunately, I couldn’t find any on the Internet for sale; and I didn’t want to spend time on a wood lathe to construct one (another issue would be finding a mounting screw of the right size and thread pitch, which no one on the Internet seems to know).
So being the cheap bastard that I am, I opted for this: a folded paper dodecahedron from http://www.korthalsaltes.com/
I made the version with an extraneous trianglular fold that I could tape on to the existing phase plug. Here it is fit onto the driver in the horn. LOL! I do think it looks much better than the paper towel strips!
Let’s take a look at some on-axis measurements with the folded paper dodecahedron phase plug in place. Note that I took these measurements inside the house, so the baseline measurements with the stock phase plug might be slightly different from prior posts where I took the measurements outdoors.
Yellow = Dodecahedronon plug on-axis
Gray = Stock phase plug on-axis
Holy crap, this actually works! The dodecahedronon plug doesn’t do much to the 8.5kHz peak, but it eliminates that big honking peak at 4.5kHz (which seems to be a huge contributor to the horn coloration and treble nastiness). We end up with narrow dip at 5kHz, but I’ll take this dip over that peak any day. Subjectively, there is quite a bit of difference while sitting directly in front of the horn, placing and removing the dodecahedronon plug.
Now here’s the 15 degrees off-axis frequency response with the folded paper dodecahedron phase. (15 degrees happens to be the incident angle at the listening position):
Yellow = Dodecahedronon plug off-axis 15 degrees (approximate)
Gray = Stock phase plug off-axis 15 degrees (approximate)
Compared to the stock plug, the dodecahedronon plug measures with slightly more emphasis between 5-6kHz (not necessarily bad because this region is now more consistent with the on-axis response), but lowers the problematic 7kHz to 10kHz area.
Subjectively, the dodecahedronon plug offers a good improvement over the stock plug and paper towel strip approaches. The more uniform polar response is definitely noticeable – the sound is more consistent when I move my head forward, back, left, or right a few inches while sitting on the couch. There’s almost a mellow quality to the DX3s now, to such an extent that that I needed to toe out the horns a bit from 15 degrees off-axis to 10.