Below are the Roon EQ settings for my HEK V2. I used technical data as a start point (e.g. any FR graphs I could find), but final adjustment was always done subjectively, primarily (but not solely) using female vocals as a guide. My primary (but not sole) objective was to reduce the effect of hotly recorded vocals (including sibilance) without over-smoothing the overall sound.
The first change is the 3db dip centred around 8.3khz to remove a perceived thinness to the sound. From the FR graphs I've seen, the HEK V1 would benefit even more than the V2 does in this lower treble area.
The next band I've changed is a small 0.5db boost centred around 1.9khz. This didn't subjectively affect tonal balance much, but did add "body" to female vocals and just about everything else. I've only added this band quite recently, so still experimenting with this one, as well as its affect on the other bands. Curious that the fundamental male/femail vocal range is below 1Khz, so I guess something else is going on (e.g. harmonics) that is affecting the perceived "body" of the vocals.
The 3rd band reduction at low frequencies is purely to counteract the bass boost that is a by-product of DAVE's cross-feed function (this example is for the most extreme setting CF=3). I wouldn't bother with this change if I wasn't using CF. Curiously, the subjective bass response didn't seem to be affected as much as mid-range transparency did when I changed this band. I don't know if the end result exactly balances out the CF boost - probably not, I just went for what sounded best.
The above graph scale has a range of only 24db, so it magnifies any changes. The 60db scale below more realstically represents the volumes we listen at (I listen typically around 70db), and shows that these adjustments are really quite minor in practice.
I'm not claiming that these are the only EQ changes that are worthwhile, but my motto is to avoid any DSP unless I can clearly hear the benefit - I don't go looking for changes if I'm happy with the current result. YMMV, particularly as I've found that tiny adjustments matter. E.g. changing the mid boost from say 1.9Khz to 2.0Khz was detrimental. Also, each band's frequency/amplitude/Q combination can affect the settings for another band's combination.
All this implies that the HEK is extremely sensitive to tiny upstream changes. Or it could just mean that I have an extremely vivid imagination
