You're not only "measuring fundamentals". Remember that a recorded signal (let's say a recording of an instrument) encodes the fundamental and all the instrument's harmonics (and any noise etc.).
If a speaker has a flat FR, then every harmonic in that recorded signal will be reproduced with the correct amplitude relative to the fundamental. If the speaker has a non-flat FR, then the relationship between the amplitude of the recorded fundamental and its harmonics will change.
Yes, and there are thresholds at which these changes become audible. Decades of scientific investigation have determined to a high level of accuracy where these thresholds lie in the midrange and treble, although there remains further work to be done (or at least documented) with regard to these thresholds in the bass.
I don't have a fundamentally different attitude towards science.
Firstly, we know from psychoacoustics that our ability to hear differences in FR is as fine as 0.1dB for wideband deviations, rising to a few dB for narrowband deviations. See the studies
here and
here. The kinds of differences in FR you're claiming the current science says shouldn't be audible are in fact grossly beyond scientifically established audibility thresholds (see below, where I discuss this specifically with reference to measurements of the JBL and KRK speakers you've mentioned).
So the fact that you can hear differences between the JBLs and the KRKs is not a scientific anomaly, it's exactly what you'd expect.
Secondly, I most certainly don't believe anomalies (not that we have one here) should be disregarded. The scientific method demands that all anomalies should be invetigated. The first basic requirement, however, when an anomaly
appears to exist, must be to establish controls and then ensure that the apparent anomaly can be reproduced under controlled conditions. (This is no doubt how a signifcant chunk of our understanding of psychoacoustics emerged.)
Frequently, though, apparent anomalies (e.g. perceived differences between the sound of devices that psychoacoustics tells us should not sound different like, for example, cables) cannot be reproduced under controlled conditions. Wherever perceived differences can be reproduced under controlled conditions, I'm very much in favour of their further investigation.
Thanks for that. The VXT8 measures extremely differently from the JBL. Let me explain how.
Firstly, on-axis FR. While both are quite flat in the upper-midrange and treble, the JBL is smoother overall, and (more importantly) the KRK has a 4dB, wideband dip in the upper bass. The KRK also has lower bass extension. Here are the FRs side by side, with the scales approximately matched:
View attachment 106203
But the really important difference is not in on-axis FR, but off-axis FR. Unfortunately we don't have identical measurements of the two speakers to compare, but plenty can be inferred from what we have. Here's the KRK's horizontal off-axis response. Again, side-by-side and stretched to a similar scale to the JBL's spinorama. I've circled the most important region of the KRK's graph:
View attachment 106204
At the top of the KRK woofer's passband, it beams significantly. At the point where the tweeter then takes over (around 2kHz), the directivity jumps back out to being very wide. This is a classic cone/dome two-way dispersion pattern: beamy in the midrange, ultra-wide in the upper-midrange/low-treble.
The JBL, OTOH, is a constant-directivity design, with a waveguide ensuring the directivity of the woofer and tweeter match at the crossover point. The speaker's radiation in the midrange is slightly wider than the KRK's, and (more importantly) its radiation in the upper-midrange/low-treble is significantly narrower.
Based on what we know about human hearing (see above), one would
expect these speakers to sound very different. The approx. 4dB wideband dip in the KRK's upper bass relative to the JBL is unquestionably going to be very audible. Moreover, our perception of a speakers tonality and (especially) its spatial qualities are largely determined by its off-axis radiation, as this is what determines the spectrum of first-reflected sounds arriving at our ears a few milliseconds after the direct sound. In this case, the spectrum of the KRK's reflections are going to contain far more upper-midrange/low-treble energy, which one would expect (depending on the acoustics of the room) to result in its having a more spacious sound, and a very different tonality.