Dennis,
You are incorrect. Those are not chamber measurements, DM does not have an anechoic chamber -- very few loudspeaker companies / designers actually do. It is BECAUSE the measurements stop at 200Hz why they were not taken in a chamber
If they were in a true anechoic chamber, they would extend down accurately to 40Hz or below (depending on what the chamber is actually tuned to)
You do not need an anechoic chamber to take accurate anechoic measurements. Curtis is correct, the measurements are pseudo-anechoic or sometimes called quasi-anechoic. With the proper equipment and knowledge, reflections can be gated out of the response by calculating a Fast Fourier Transform (FFT) on a limited time domain window (impulse response), typically 5-7 milliseconds in length. Anything outside of this window size is meaningless. In addition, the smaller the gate time the less resolution in the actual measurement so while we are removing reflections from the measurement, we are also limiting resolution of the entire measurement. It is because of this why we still prefer to use MLSSA systems for our measurement gear. MLSSA offers an "Adaptive Window" mode which allows it to maintain excellent resolution in gated measurements. However, MLSSA is brutally expensive and requires years of experience as it is absolutely NOT user friendly. It is, however, still used extensively and considered an industry standard.
There are as many techniques to obtaining accurate gated measurements as there are measurements themselves. It is because of confusion/misunderstanding like this why I mentioned comparing measurements from one manufacturer to another is often meaningless. I can make a loudspeaker look fantastic in one method or poorly in another.
Compare manufacturer claimed anechoic measurements to actual NRC measurements (taken in a true chamber.) You are going to see many surprising differences.
Hope this makes sense.
Please take any further discussion of the matter off-line. This thread is about the Sierra Towers.