(Thread title inspired by @Boden from @bwaslo's thread "Xsim-3D development... I could use some math help")
A little background first: I've done many loudspeaker measurements with the tool set that many DIYers have, a calibrated mic and some free software, but I'm looking for a better system. I've been looking more closely at Earl Geddes measurement method, and Gabriel Weinreich's paper "Method for measuring acoustic radiation fields" and what I can find on the function of Klippel's Near Field Scanner. Bwaslo's recent thread "Xsim-3D development... I could use some math help" took a turn that made me want to start getting some help on understanding all this. And like Bill's thread, I could use some math help...
In addition, a number of very interesting questions were rasied to Earl Geddes about more advanced measurement methods that I think could use their own thread.
So so far, I have two big questions I'd like some help with, please:
1. what is needed to turn a measurement set into a view of radiation modes (for lack better way to ask the question)? Geddes uses a FORTRAN routine, but would something like open source mathematical software work, or is it best to have purpose built software?
2. Weinreich's paper and the Near Field Scanner both are capable of "sound field separation", in other words, they can pull the room reflections out of the measurement and give you just the direct sound from the speaker (no need for an anechoic chamber!). Klippel only appears to do this up to 1kHz (plus or minus an octave) and uses measurement windowing above that, Weinreich didn't appear to us it very high in frequency either. It appears to work by looking at the timing of sound waves passing through a pair of measurement shells? But what I'd really like to know is if Weinreich could do that with technology from the late 1970's, can intrepid DIYers do the same with available software and hardware today?
And please, if anyone else has questions about the function of more advanced speaker measurement systems, please post them here.
A little background first: I've done many loudspeaker measurements with the tool set that many DIYers have, a calibrated mic and some free software, but I'm looking for a better system. I've been looking more closely at Earl Geddes measurement method, and Gabriel Weinreich's paper "Method for measuring acoustic radiation fields" and what I can find on the function of Klippel's Near Field Scanner. Bwaslo's recent thread "Xsim-3D development... I could use some math help" took a turn that made me want to start getting some help on understanding all this. And like Bill's thread, I could use some math help...
In addition, a number of very interesting questions were rasied to Earl Geddes about more advanced measurement methods that I think could use their own thread.
So so far, I have two big questions I'd like some help with, please:
1. what is needed to turn a measurement set into a view of radiation modes (for lack better way to ask the question)? Geddes uses a FORTRAN routine, but would something like open source mathematical software work, or is it best to have purpose built software?
2. Weinreich's paper and the Near Field Scanner both are capable of "sound field separation", in other words, they can pull the room reflections out of the measurement and give you just the direct sound from the speaker (no need for an anechoic chamber!). Klippel only appears to do this up to 1kHz (plus or minus an octave) and uses measurement windowing above that, Weinreich didn't appear to us it very high in frequency either. It appears to work by looking at the timing of sound waves passing through a pair of measurement shells? But what I'd really like to know is if Weinreich could do that with technology from the late 1970's, can intrepid DIYers do the same with available software and hardware today?
And please, if anyone else has questions about the function of more advanced speaker measurement systems, please post them here.