This article might help, its from Tony Gee who has designed and built seriously good speakers for over 20 years and this article compares naked drivers (21 inch Precision Devices sub woofer) with U Frame, sealed box and hybrid ... Very interesting stuff and his whole website is a goldmine of info. https://www.humblehomemadehifi.com/download/Humble Homemade Hifi_Serious Sub.pdf
A nearfield measurement is not able to show dipole peak or anything else related to dipole behaviour. Indoor measurements of midbass are very difficult, there will always show room modes and reflections. Speaker almost in middle of the room ointing to corner, mic at 60-100cm distance is best. Mic on the floor would be even better, but it's difficult to have the driver pointing down...
For analysis try various gating times 20-100ms and use smoothing 1/2 or 1/3 oct. Don't worry about small wiggles, they are artefacts! Edge of VCAD2 simulations are for sanity check.
For analysis try various gating times 20-100ms and use smoothing 1/2 or 1/3 oct. Don't worry about small wiggles, they are artefacts! Edge of VCAD2 simulations are for sanity check.
Last edited:
Interesting. Seems Kreskovsky's method suggested earlier isn't as accepted as I had interpreted.
I went ahead with naked 18" midbasses raised 8" from the floor, ran the Acourate correction routine, reduced the gain from the subs, and sounds a lot better. Happy camper!
I thought I would see a large drop in the peak at 250Hz and subsequent drop by 280Hz, but it wasn't the case: measurements at the listening position, before correction, with psychoacoustic smoothing. Green is with U baffle, blue is nude.
the peak is still there. Something else must be causing it.
But it does sound significantly better!
In the coming days I will measure as recommended.
I went ahead with naked 18" midbasses raised 8" from the floor, ran the Acourate correction routine, reduced the gain from the subs, and sounds a lot better. Happy camper!
I thought I would see a large drop in the peak at 250Hz and subsequent drop by 280Hz, but it wasn't the case: measurements at the listening position, before correction, with psychoacoustic smoothing. Green is with U baffle, blue is nude.
the peak is still there. Something else must be causing it.
But it does sound significantly better!
In the coming days I will measure as recommended.