Serious' cheap OB

Mic is flush with the ear canal opening, completely blocking the ear canal. In the living room I got a 10db upper midrange gain (like so*) and this was measured in exactly the same way. Part of it is the speakers and off-axis response and part of it is my room.

The mic won't roll off so much. Maybe -3db at 20kHz. It's a 1/4" omni mic. The angle will also be more like 60° (well, except for the 0° measurement where it was 90°). The roll-off is basically from my head and ear, but I think it's accurate to how we hear. I think it's that with an omni mic measurement we don't see this stuff, so we don't know if the B&K target will sound too warm, or if it may be too bright. In the living room I had none of this roll off compared to the omni mic measurements, in fact the treble was boosted with the in-ear method*.

*That graph is already compensated for the B&K target, so if the speaker measured flat at the listening position it would make for a rising FR.

 

Tried another H-frame experiment and saved my measurements this time. I used 4 55cm wide, 70cm high thin pieces of cardboard and used double-sided tape to stick them to the sides of the woofer baffles.

Here's the FR at the listening position (or slightly off-axis):

Honestly I'm surprised that the change is that big. Would be interesting to take an impedance measurement to see how much Fs moved. Looks like about 3.5db more below 65Hz, which is quite a lot. It's possible that the measurement looks better than it really is. I also attached what the stereo sum looks like at the listening position**.


Subjectively it's not that great. The bass seems to be more spongy, less distinct. Slower. I also feel like I can hear a panel resonance in the upper bass - sounds like cardboard*. Yes, you get more bass overall which is a welcome change, but the bass quality does suffer. I think I'll rip the panels off again.

It's probably a better idea to use 2 woofers in parallel and tweak the crossover in order to get more bass. Maybe a slightly smaller U or H-frame with more rigid panels, too. I just wonder how well the Ragnarok would handle a 3 Ohm load in the bass (from the 3 drivers in parallel).

*The two huge peaks between 100 and 200Hz in the measurement are from the room. The insane top octave rolloff is because the measurement was off-axis.
** I don't know what's causing the insanely sharp drop-off below 24Hz. It's like it's at -5db at 25Hz and then there's no output at 20Hz. I also added a 15 cycle long window to better show the early decay in the midrange.