High frequency roll off

[MatiasR]

I am new to REW and UMIK-1 and have been doing my first measurements. One thing that seems strange to me is that both my systems, in different rooms with completely different components and speakers, different PC running REW with the same UMIK-1 mic + calibration file, both have a strange 10 kHz roll off.

Am I doing something wrong or are my systems really rolling off the highs? Also EQ does not try to fix it, it just leaves it as it is, even if I specify to try to fix up to 20kHz.

Measurements: blue is system 1 in a largish living room, green is system 2 in a small office.

1. HTPC - Uptone ISO Regen - Calyx Femto - Wyred 4 Sound STP-SE stage 2 - Apollon NC800 SL PRO - Thiel CS3.7
2. Desktop PC - Burson Conductor V2+ - Calyx Femti - Monitor Audio PL100

 

[xSilmarilSx]

I have no problem going to 20k with Polk Signature series.

 

[MatiasR]

I will try to change some things on my main system and remeasure. After I posted this I though better and my measurement is probably OK since the second system does not have such high frequency roll off.

 

[jtalden]

If you measured both L and R together and the mic was not perfectly centered that can create this 10 kHz rolloff effect. If the mic was slightly more off center there would also be more obvious comb filtering dips in the response at other frequencies. It's best to measure L and R separately. We can measure them individually, align the impulses in REW and then use (A + B) / 2 REW arithmetic to see the L+R combination that would occur with the mic perfectly centered.