Forum Replies Created
24 July 2020 at 3:30 am #238
I looked over your system and project files.
The export format was set for “Binary file (32 bit, int)”, not “Binary file (32 bit, float)”. This will likely cause miniDSP to see NaN when importing.
Regarding filtering, there was an 2nd order all-pass IIR at 38 Hz and a 2nd order maximum-phase all-pass on the “FIR Phase Adjust” tab but the tab’s “Enable” was unchecked. Try removing the IIR filter and checking the “Enable” on the “FIR Phase Adjust” tab. This should give you a FIR that unwraps the LF phase for the measurement in the project.
EA Support14 July 2020 at 10:38 pm #223
>> Is it correct to assume that this loudspeaker is displaying non-minimum phase behavior
>> around the crossover region when one looks at the 2-nd order all-pass response of its
>> phase response?
>> Can the phase adjustments being done around the crossover region, to flatten the phase
>> response, be deemed to be “maximum phase” corrections seeing that we are inverting an
>> all-pass response?
Yes. But remember this is just one loudspeaker. Each loudspeaker is different.
>> Is the flattening of both the magnitude (EQ boosting) and phase response around the
>> crossover region not going to introduce audible artifacts like pre-ringing?
Maybe. Linear-phase brick-wall crossovers also have pre-ringing but it’s rare that people say they think they can hear it.
>> Can the same EQ concepts be readily applied in correcting the room transfer function
>> displaying non-minimum phase characteristics at the listening/measurement position?
Yes but I’d suggest doing a literature search on the topics of room impulse response correction and inverting room impulse responses; maybe in the Audio Engineering Society resource library.