Thousand thanks detective work in take all the tests Amir and still be here in what looks relative good mood ...
... will admit from outside it looks that garage and Germany really had added some work load in 33+29+10 pages of posts and 2 times realignment of system after dismount/mount af microphone cage..
View attachment 73346
View attachment 73364
New sample 3 curves looks really nice but bit surprised it scored so high and passed Revel F208, below animation is new sample 3 analyze and the blinking overlay is a traced copy of the sample Neumann had asked Klippel Germany to measure, for info Neumann sample is offset +0,7dB but nice coherence there..
View attachment 73354
This one is bit funny or bit irritating, its sample 3 added microphone cage fix from Genelec thread plus MK-255 microphone calibration corrections in reverse or say mirror, and for folks following KH 80 it should look clear that say 500Hz and up it looks exactly as sample 1 did and okay GuyLayfield temperature analyze of KH 80 add some tiny bit of boost above 10kHz and same do calculate system on tweeter axis as was done for sample 1 verse 52mm lower as for sample 3, but still in my eyes it could rather look like MK-255 micrphone calibration in Klippel software has in past mayby been accidently set in reverse/mirror, should that happen its not a disaster if one take a lookup of that correction curve some graphs below this one..
View attachment 73355
Thanks to GuyLayfield do that temperature test and share data, have traced the 5º cold curve as seen below to get a correction curve to add to sample 1..
View attachment 73357
This beautiful curve is sample 1 added 4 times corrections, one is microphone cage fix from Genelec thread, one is MK-255 microphone calibration, one is in CAD software lower microphone height position in space relative to spindata or say tweeter axis -52mm, and fourth corrrection is above temperature fix..
View attachment 73358
Below is sample 1 animation including the 4 times correction fixes plus sample 3 on axis curve as reference, and even its modeled prediction it looks quite close, thanks to everyone share knowledge hard work and data, the four filters used below is attached to post.
View attachment 73361