I believe those are separate issues and what you see in the FR plots can't have anything to do with room reflections.
We still don't know what are the actual implications of the impedance wiggles (probably caused by room reflections and a high acoustic impedance of the device), but I doubt this would be visible in a frequency response. For that the effect would need to be really strong, which it probably isn't.
Maybe here's a trace, but this is hard to tell -
Without and with those features included in the time window:
Without and with those features included in the time window:
I assume that is the A460G2? Seems to be what I expected. Reflections but 2.6 ms are 0.9 m that is 0.45 m for both ways. Is that even possible in your measurement situation?
Because there is no window applied in Impedance measurements these reflections have an impact on the impedance.
Because there is no window applied in Impedance measurements these reflections have an impact on the impedance.
When windowed as shown, there are clearly two peaks of energy corresponding to the FR deviations.
This must come from the back of the waveguide, IMO. Obviously these will be just the right wavelengths that have the chance to reflect and get back towards the microphone around the mouth. That's the only explanation that seems plausible to me. The bigger waveguide seems to be better functioning as an obstacle.
This must come from the back of the waveguide, IMO. Obviously these will be just the right wavelengths that have the chance to reflect and get back towards the microphone around the mouth. That's the only explanation that seems plausible to me. The bigger waveguide seems to be better functioning as an obstacle.
Put couple of pillows behind the waveguide to test it. At least in ABEC sims sound difracts all around the device, although here on your measurement band width seems narrow the delay time matches I think.
I've never actually tried it but this could help: https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-7229606
I have 4.5 - 5 ms. With the mic closer to the horn, the timing is still the same - it must be the something around the horn itself.
In any case, if you had a maximum of 5 ms for the previous plots, you have included the first reflections, floor or ceiling bounce, and cannot separate them from the presumed wrap around effect of the back wave. Were they done at 1m? Can you expand on what you mean by 'timing is still the same', and at what distance the closer mic'd measurements where done, 0.5m?
No, the first strong reflection comes a little later.
Here's the same in HOLM (saved measurement attached):
The current small room is all I have at the moment. Of course it would be better to have at least 10 ms.
Here's the same in HOLM (saved measurement attached):
The current small room is all I have at the moment. Of course it would be better to have at least 10 ms.
You could try measuring at different distances and combine the measurements. It's like using a directional microphone. That way the room reflections are suppressed. Reflections/diffractions from the waveguide will decrease a little too as they are not completely on axis. I guess this will depend on the microphone distance(s).
There is a paper and a way how to do it (in REW I think). I'll try to find them.
There is a paper and a way how to do it (in REW I think). I'll try to find them.
I recently finished some A520G2 Waveguides and the T520-4554-ext. Here are three impedance meansurements made with the Dayton Audio DATS v3.
The first is with the waveguide pointed at the ceiling and the driver resting on the floor (rug).
The second is with the waveguide laying on its side, rim on floor (rug) and side of driver on the floor (rug).
The last one is with the waveguide facing up outdoors and the driver resting on the grass, away from any buildings.
I conclude that the room reflections cause the tiny wiggles in the impedance plots. Not sure what the other wiggles are. Perhaps imperfection in the assembly.
The first is with the waveguide pointed at the ceiling and the driver resting on the floor (rug).
The second is with the waveguide laying on its side, rim on floor (rug) and side of driver on the floor (rug).
The last one is with the waveguide facing up outdoors and the driver resting on the grass, away from any buildings.
I conclude that the room reflections cause the tiny wiggles in the impedance plots. Not sure what the other wiggles are. Perhaps imperfection in the assembly.
I mean to remember from a set of measurements a similar result, but the waveguide was baffle-mounted in an enclosure. I remember I was getting frustrated about some very early reflections, which were happening too early for the measurement situation, but left a markable imprint. I had to use a very short window to exclude them, which didn't really make sense to me. Maybe it was sound going around the enclosure.
As a side note, this commercial waveguide coincidentally required a tubular adaptor.
Or just a reflection from the back of the enclosure. All of this eventually reach the microphone. The cleaner the "direct" sound, the more noticeable this is. In case of these waveguides it simply stands out, even it's still on the order of 1 dB (may be more with a longer gate)...
I hope a more sophisticated mouth edge will help to further reduce this effect. It's strange I can't reproduce this in a sim, no matter how I try, though.
Last edited: