Originally Posted by kiwi2
I can make the frequency response look better again with EQ, but I find it degrades sound quality from doing so. I'm not sure if it's the distortion it introduces or the fact you are altering the speaker's natural response. I would rather live with a peak or dip somewhere in the FR and no EQ. But of course I would still prefer a smoother response if I could achieve it with better room placement or treatments.
A simple FR measure is going to show the summation of direct sound from speaker and contribution from all room reflections at the mic location. It is quite possible (and likely) the direct sound from speaker is flatter than measured, with the room "contaminating" the response. Our ears discriminate direction and timing cues as well as frequency and amplitude, thus you may hear the flatter on axis response as, well, flatter than your measurement, and hear the room contribution as any of a number of things that are situation dependent... soundstage widening, or image smearing... spaciousness, or a masking echo... and yes even sometimes timbre shifting, etc.
A simple FR measure with typical omni mic doesn't resolve the additional cues our ears do. Not that this can't all be measured, it just takes the proper tools, planning, time and effort.
All that said, improving the speakers' direct frequency response (anechoic) via EQ can further improve the sound in room. Improving the measured in room response at the listening position via (normal, IIR filter) EQ but worsening direct speaker/anechoic response in the process may well look prettier and sound worse.