algorithm a-b seem not work, phase is near same

[user44455555]

Hi

I measure with microphone on side near ear when i sit on chair and when chair was empty(without move microphone or speaker). this cause upto 3 db diffrences and on alot frequencies. maybe it is because i use a 32 inch computer display which is near my seat so the audio bounce between my body and computer display often and change FR. When i do a-b the diffrence show is much more as 3 db which is maximum diffrence. I like the diffrence from chair empty to chair sit as a room curve so i can measure when not sit on chair, and EQ is correct as i sit on chairn and test if that sound better. I think the on chair measure can sound better , because i always reduce 4 khz and 2 khz without see a reason in measure when not sit on chair. How can i do that best in REW to get a room curve ?. 1/3 smoothing i think is ok

 

[John Mulcahy]

You should be doing A/B, or more likely |A|/|B|.

 

[user44455555]

You should be doing A/B, or more likely |A|/|B|.

i have done this and it give at 402 hz where on seat is 52.3 db , and seat empty is 50.3 0 . I do the meaures with 4 times repeat setting.
Wy A-B does not work as i think ?. when A is at 402 hz 52.3 db and B is 50.3 then A-B is 2 db. the phase can see is very simular
I also notice a problem when want change the color of the messure. when i set in 1. measure to blue in the color selector box, save it and load it in, it is not blue. the mdat i upload too

 

[John Mulcahy]

Wy A-B does not work as i think ?

Because the graph is showing dB, a logarithmic unit. Addition of dB values corresponds to multiplication, subtraction of dB values corresponds to division.

I also notice a problem when want change the color of the messure