Originally Posted by John Schuermann
Short answer - distortion is audible no matter WHAT room you put a speaker in. For example, KEF speakers do well with the Spinorama, but they all suffer from audible intermodulation distortion (thanks to the Uni-Q design, also one of their strengths). That doesn't show up in the Spins, but can easily be heard with the right program material.
Andrew Jones responded to this argument several years ago on the DYI forum:
"Back in the days when I was at KEF and we were developing the Uni-Q drivers, we also questioned the audibility of any fm or am distortion due to the moving cone.
AM comes from the change in loading of the waveguide upon the tweeter as the cone moves forward or backward. If you measure the tweeter frequency response for different positions of the cone you do see changes in the frequency response of the tweeter, both in shape and sensitivity. This contributes to the AM.
The fact that the cone is moving due to the bass signals does give rise to some doppler, but there is also a similar effect occuring in all speakers, as part of the baffle of the speaker that is close to the tweeter is in fact the bass cone!
Although not as severe because the coupling is less, there is a measurable effect.
We measured other types of coax units, such as the TAnnoy dual concentric, and of course observed similar levels of distortion products.
We also measure the Quad ESL63 as a reference point. Being a full range diaphragm, this should have the highest levels of doppler.
The level of sidebands were lower but not greatly so. From this we made a simple conclusion. If the ESL sounds as good as it obviously does, with these levels of doppler, then it is a distortion mechanism we do not have to worry too much about!!
Of course this still leaves the issues of AM. The clear solution to this is to minimize cone displacement.
A two way based coax speaker system is the most difficult to engineer. Not only does the cone move a long way, but the half-roll surround that is neccessary for such displacement wreaks havock on the diffraction of the tweeter wavefront on axis. That is one reason why co-axials should always be listenened to around 10-15 deg off-axis. As Pallas said, toe the speaker inwards so that the axis crosses in front of the listener. This stabilizes the center image and also minimizes sidwall reflections.
Limiting the bandwidth of the driver to around 200 or 300Hz, in a three way system, greatly diminishes all these negative effects, and allows one to design a surround that does not mess up the diffraction, so that even on axis the response is good.
This is the approach I have used ever since."