Hi all, this is my first post here on this forum, after long-enjoying reading (and donating for
) ASR’s / Amir’s great work. I am a DIY loudspeaker builder and so can add a few points. I don’t know whether these have been covered in this thread already. Apologies if I repeat a few points.
First: on the bass hump:
As far as I can tell from the pictures, the enclosure contains 3 bass reflex ports. Is this correct? Amir’s measurement mention that he kept the backside reflex port open. If this backside port is a reflex port of equal length to the ones in front, then increasing combined reflex port-size leads to a higher frequency port resonance of the three ports combined than of the two ports on the baffle-only. Hence a hump in upper bass. I recommend a measuring again, with that back-port closed: the speaker should go deeper and will have less of a ‘pseudo bass’ ‘hump as seen here (and, as applied with many commercial designs).
I presume that the woofer used is (close to) the ($220) Scan-speak 15w-4530k00. That woofer is known to be one of the better small bass-woofers, e.g. it is one of Troels Gravesen’s favourites. It deserves a larger enclosure of 12+ liters, so that it will truly play well, down to 50Hz.
Secondly, it's great to read Amir’s comments on off-axis responses compensating on-axis measured ‘deficiencies’: we listen to the sum of on- and off axis response, so we shouldn't read too much into on-axis measurements (only). This is a known dependency in (DIY) speaker design, particularly with wide-dispersion units like the ($220) Scan-speak D3004-660000 (check its datasheet, and its very wide horizontal dispersion at 3 kHz in Amir’s measurements). IMO it is wise of Wilson Audio to take some energy out of the on-axis tweeter response at 3 kHz by creating a filter-x-over gap, albeit that I think they also created a bit of a ‘BBC-dip’.
Third, Amir mentioned baffle-step impact. At this size of enclosure, that applies to the 700 Hz -> cross over point. Amir didn’t often mention baffle-step impact before. Baffle-step impact combined with any internal resonance / port resonances can create nasty humps.
Lastly, I agree with many comments that low distortion really matters; it matters more than linear frequency response accuracy; see also Amir’s ‘Understanding Audio Frequency Response & Psychoacoustics’ Youtube video (recommended).