Sweeps and RTA discrepancies help

[Andrew Slater]

Hi guys ,

I have CSL UMIK2, and a creative SBL sound card , I go aux into my system for noise

Calibration on sound card using loop back from line in from headphone out

Cal files all loaded ok and working

So when I measure in RTA doing left and right together there are some frequencies that show less energy when left and right play together, and when they play by themselves there no losses


When I use sweeps, and do 30 sweeps around my head (equivalent to 28 average rta) it shows two BIG dips per side

Left at 1k and 6500 and right at 1400-7000, on RTA dips not there , sweeps averages they are


Lastly , how do I get frequencies when summed to not fight

I use fir filters everywhere, all phase has been forced to 0 using measured phase , at the speaker and again at the listener position as a gloabal correction

No matter how I measure those sides , they should not cancel….. the phase is not near -180 on one side compared to the angle of the other side.


It sorta makes no sense

And why does live RTA sound so much better then sweeps outcome?

 

[John Mulcahy]

It is essentially impossible to get a perfect phase match across the whole frequency range between two sources in a room measurement, no amount of filtering will change that.

 

[Andrew Slater]

It is essentially impossible to get a perfect phase match across the whole frequency range between two sources in a room measurement, no amount of filtering will change that.

I’m going to ponder this for awhile

Thanks for helping….

My biggest issue is 130-180hz ,

When left and right are not summed , full responce , add them together and cancels and have to put a ton of amp power to get that to fill in

I’ve tried putting one side out of phase , nada. It’s so weird

I wish I knew why is all …. So then maybe I could think about a different location for these midbass drivers.

Super appreciate you sir :-)