Milestones speaker met Thiel units

[Contour]

And now we need someone to LISTEN...

[Paul Vancluysen]

You can measure the driverunits also separate, each with its own filter and display the curves in the same graph. It gives a better understanding of the total response you measure now. I don't know how familiar you are with your measurement equipment...maybe some hints. In the case you use a mls measurement, you should place a window over the reflection free part of the impulse response. The start of the window should be set just before the start of the measured impulse, this is important too. Interpreting your measurement  looks like a FFT result without windowing. Measuring in a room can give a too short reflection free impulse response, which also can cause faulty dips in the frequency response at a few hundred Hz applying the FFT. Also possible is to measure the frequency response with a FFT measurement applying noise to the speaker and do some averaging in time. In this way the reflections are averaged too and cancel each other. Sometimes it delivers better measurement results in a room, in the case the reflection free part of an impulse response is too short. Hope this will help you a little.

Paul             

[torquemada]

My first room dampening test made very little difference.

I'll test reversing the phase of the mid driver, as soon as I get hold of a decent soldering station. Mine got left behind 200km from its new home when I moved.

[torquemada]

Inverting the phase of the mid did the trick. Thanks for pointing it out.

Now I'll try to address the in-room dips.