Morello skrev:Jag är skeptisk till mätmetoden - hur trollar man bort efterklangen?
Om jag vore Amir skulle jag komplettera med mätning i ekofri miljö.
Här svarar Amir på en liknande fråga, som beskriver lite om Klippels metodik bakom mätningsförfarandet. Utan att kunna syna resonemanget i detalj bedömer jag att Klippel verkar veta vad de pysslar med och tillämpningen väcker min nyfikenhet.
"Hmmm. I think you are thinking of how a human could make such measurements. That is not how the klippel works. It has no idea of sound sources, baffles, etc.. It simply takes a snapshot and uses that to solve the wave propagation formula. Once there, the sound waves can be projected to far field for any distance. This is not based on any kind of cleverness. It is pure mathematics.
Now, there are heuristics used for measurement point optimization and prediction complexity of the wave equations (made up of different order Hankel functions). Fortunately the system makes redundant measurements and then compares those specific measurements against the predicted ones. if it fails to get accurate results (within 1 dB), it will indicate so. We lose it at the low end because these little speakers don't produce any output down to 20 Hz so room noise takes over. At the upper end, it is actually showing that I can go up to 10 kHz using this method. But I don't. Instead, gating function is used from 1 kHz.
The more complex the soundfield as measured in near-field, the higher order hankel functions are needed and more points to sample. For 2-way speakers, 500 points is plenty. Order is 5 to 10. Both can be increased and I will likely have to do that if I measure floor standing speakers. Reports are generated by the tool if either one of these parameters is not enough to compute the soundfield. Room reflections are taken out by making a second independent scan parallel to the first one. These two used differentially can help dial out the reflections. Per above, this technique is used for frequencies below 1 kHz.
I wish the math and method was easier to explain but it isn't. There is a reason the software costs much more than the hardware! There are some AES papers if you are interested but they are not at all light reading. We will be discovering some issues I am sure as we go by. But the simple stuff is not it."
/mats
Skivaffär, IC kretsar, mobiltelefoner, hifi-skribent, näringslivsanalyser i CV:t. Har snöat in på dipolära högtalare, Stig Carlsson och psykoakustik