For the purposes of comparing 2 or more FR graphs, isn't normalizing to 1Khz just an arbitrary convention? One could normalize to another frequency if he wants to, say, start with the bass volume the same on both, and see how the rest of the FR compares. Or one could argue it might be better to normalize by dB (line up the FR graphs along the horizontal axis so they have equal area above and below 0db line). But in this case, how do you compensate for the highest octave of human hearing contributing the least to the overall sound? Or is that already accomplished by the logarithmic frequency axis?
I think this really underscores why it's difficult to get much more than an overall impression by comparing FR graphs. Not to mention those sound characteristics that are not even captured by FR.
One recommendation when looking at normalized graphs is to consider the music you're listening to.
I like to normalize at 450Hz to get a picture of how the general perceived frequency would be if listening to "bassier" tracks or tracks with focus on, let's say, bass guitar and drums.
Normalizing at 1000 or 1500Hz would give you a clearer picture on the overall in general, from my experience, as this will show you the mid to upper range voicing. So vocals, (female vocals especially), electric guitars, strings, winds, etc. As you see, normalizing at 1kHz is kind of convention because it's more informative.
Take for example Singularity, a very consistent tuning with epic bass & Fei Wan, a very variable tuning with more focus on equally epic bass (sub+midbass) and treble.
At 450Hz aligned:
For a lot of music, Fei Wan is exciting and engaging and just really a "wow" factor because of the generous amount of sizzle and bite yet not going way too overboard, while vocal range remains clean and lows have a beautiful bass slope which provides satisfying bass. This is at
450Hz aligned & generally how I'd hear this IEM with Electronic music (bass focused), smooth jazz or modern genres Hip Hop/ Rap. I really enjoy
this Fei Wan.
At 1000 Hz aligned
But, and this is how I generally hear Fei Wan... Aligned at the convention
1000Hz you can see a lot of midbass fullness (or some would say generous bloat), and this is countered with the treble being ramped up even further to maintain clarity. The problem, for my personal preferences, is that you gain a substantial amount of upper midrange forwardness, in my case very shouty vocals and electric instruments. Fei Wan sounds like this 70% of the time for me, and with most genres. It's a quite aggressive W-Shape with a lot of musicality and dynamics, but a heavy coloration. Think decadent ice cream dessert with extra candy sprinkles and hot fudge, but you kind of lose sight of the ice cream's taste; you like the ice cream dessert for its toppings
Singularity barely changes its signature no matter the alignment, so you get a consistent FR across pretty much most genres which is to say, LOTS of sub bass thump/ kick, a lot of midbass weight, super clean midrange, and a notable amount of treble roll-off. It's mister good ol reliable for a generally neutral + bass boost signature. You will never get sibilance with this set, but you may also not get the same wow factor as other more resolving sets.