EqAPO or REW overemphasize the Gain.

[James-F]

Hi.

What was a peak now is a dip after remeasuring with REW.
There is something wrong with REW "auto EQ" (aka Filter Tasks) or EqAPO or me.

REW generates an auto EQ curve to make the frequency response flat (target) but when its exported to txt file and used in EqAPO then remeasured again with REW the reading is practically inverted.
Meaning the peaks now dips and vice-versa.

What could be the problem?
It appears that REW or EqAPO just cuts and boosts too much, thus practically flipping the spectrum.

Thanks

[Jonas Dahlinger]

When I implemented all the filter types that REW supports, I checked for each one that the result of processing using E-APO is exactly (at least very close to) the frequency response predicted by REW. So in theory, after setting the configuration for E-APO, the response when measuring again should be what REW shows as the "predicted" curve.

In practice, there might be non-linearities caused by room acoustics that can e.g. cause filters with positive gain to have no or only minimal effect. But as I read it, you are also using filters with negative gain, which normally don't have this problem. One thing you can try is to disable all filters with positive gain and see if the measured frequency response is close to the prediction then.

You could also attach your filter configuration to this thread. Best would be if you attach the .req file, so I can directly import it into Room EQ Wizard and see if I can reproduce the problem with the prediction.