Originally Posted by donktard
Interesting. Did you maybe compare measurements with and without blanket thrown over a seat?
Don't you think it might also be possible that fluffy blanket takes away enough energy from highs so Audyssey decides to boost them and once you remove blanket you actually encounter even more brightness/sharpness?
Those are good questions, and if you wished to, you could certainly test the procedure for yourself, to measure, using alternative approaches to microphone placement near a hard chair back. But, FWIW, this is a well-established procedure, which is listed in the FAQ, endorsed by Chris K, and used by many of the thread participants over the years.
The issue of potential comb filtering at high frequencies, from mic proximity to a hard surface, has always been a knotty one. Everyone understands the concept of a hard surface reflecting very short wavelengths into a nearby microphone. But there has never been any real consensus on what a too close proximity really is. In the early days, Ask Audyssey recommended keeping the mic about 10" away from a chair back, and even further away from a wall (I think it was about 15" to 18", but I'm not certain). That distinction never really made a lot of sense to me, except as a compromise. Very high frequencies would be equally reflected by any hard surface, so there wouldn't be a material (no pun intended) difference between a wall and a smooth leather chair or sofa back.
So, most people on the thread, IIRC, found themselves using about 12" or so, just in case. But, that presented a problem too, because keeping the mic 10" or 12", or so, forward of the chair back meant that the mic would be at least 6" to 8" forward of the actual ear position. For most of us, the average distance from the center of the ear canal to the back of the head is about 4". So, if you typically watch/listen with your head back, the Audyssey mic would be measuring well forward of your actual ear position (and even further from the pinnae which direct sound into the ear canal). And we know that in calibrations, inches matter.
Someone, and I don't remember who, came up with the bright idea of putting an absorbent towel or blanket over the back of a leather chair or sofa, and a number of people tried it with good success. Chris K was consulted, and the procedure was incorporated into the FAQ. It works well because it enables us to keep the Audyssey mic much closer to the actual location of our ear canals, while still preventing comb filtering. And since our heads would be occupying the position during listening, that the blanket occupied during calibration, nothing is lost or gained during the procedure. The very high frequency wavelengths still bounce wildly around the mic (from all directions) during calibration, but none of them are spuriously directed into the mic from short-range (which would tend to over-emphasize them with respect to all the direct, and other indirect, high frequency waves).
But as noted in my original post, Audyssey can be finicky, so too much of a good thing can be a bad thing. It is still necessary, in my experience, to exercise some care with the blanket, vis-a-vis mic placement. For instance, a couple of years ago, I got the bright idea that if one thick absorbent blanket were good, two thick blankets would be even better.
Not so much! In my case, some upper mid-range, lower high frequency waves were absorbed, as well (similar to your speculation about the very high frequencies which we are actually trying to absorb) and the result was some inappropriate boosting in the upper mid-range, and some audible harshness. I went back to a single blanket and the problem was solved.
More recently I screwed up a calibration late at night after adding some more room treatments. I was tired, and in a hurry to finish, and I inadvertently allowed my Audyssey mic to be positioned too close to the blanket I always use. Again, the upper mid-range, and lower high-range, had a harsh quality, which was clearly audible. I repeated my calibration with my more normal diligence, this time making sure that my mic was no closer than 5" or 6" from the blanket, and the problem was solved. I do recommend staying about 5" or 6" away from a blanket, until you have a chance to listen or measure, or both, to determine how the procedure works best in your specific room. With a thinner covering, for instance, I suspect that I could get closer to it than 5" and still not affect the slightly lower than very high-range frequencies that the procedure is intended to address.
I hope that this long explanation is helpful to someone.
Edit: I decided to edit this after double-checking a couple of things. First, I can't find the blanket procedure in the FAQ, but I know that putting it in was discussed, and I thought that Keith had incorporated it. I do know that he had endorsed the technique, and I believe that he also used it himself.
Second, I had already double-checked Ask Audyssey on the 10" from chair back recommendation, but with respect to the lack of consensus on the issue, the following is interesting. The Audyssey 101 Guide recommends staying a minimum of 12" away from chair or sofa backs during calibration, and the FAQ recommends 12" to 15". As noted earlier, what the optimum distance really is may be somewhat situation-specific, but using a towel or blanket makes the question somewhat moot, as it is now possible to get quite close to the actual ear position, without encountering spurious reflections into the Audyssey mic.