I've been struggling to understand some measurements I took a couple of weekends ago. This is the on-axis of the mid at 1m:
The hole at around 1700 Hz appears in all the off-axis measurements too. I was however unable to match this notch with the theory and baffle dimensions. Then earlier this evening I modeled the baffle in the Edge. Here is the result:
The scales are different, but it's remarkable how well they match up - the constructive interference peak from around 400 to 1400 Hz, the notch from 1500 to 2100, and another small peak at around 2800 Hz. So while I'm still not understanding what is actually causing the notch, the simulator agrees with the measurement.
Fortunately this baffle was a cheap throwaway prototype. For the next prototype baffle, I put a little more thought into it and the simulation yields (at angles of 0, 15, 30, 45, and 60 degrees):
Whew, that looks better! It will be interesting to see how well the measurements match after I've built it.
So, the lesson I've learned is that even for a cheap throwaway prototype, it's worth spending a little time to simulate in Edge or something similar first.
PS. thank you to Rudolf for the tip on how to do off-axis response simulations in the Edge
