any thoughts on switching your curve to maybe something like soundguy's:
https://www.audiosciencereview.com/...ng-society-2023-conference.48497/post-1747115
or a harman compensated one? Though I think I remember in your 5128 switching video, you mentioned the upper treble is too smoothed out so it loses out on the higher "precision" of the 5128
Nope. I think there's a general misunderstanding when it comes to what the 10dB tilt is. To be clear, DF + tilt is
not a preference curve the way Harman is. If you want the best indication of how a headphone performs relative to a preference curve, Harman OE 2018 is still the best that exists.
DF + 10dB tilt primarily consists of two components. DF is used for the ear transfer function of the measurement rig, while the 10dB tilt is the bass to treble delta that's been found to be preferred in both headphones and in speakers across a fairly wide body of preference research (including Harman). This doesn't mean we should expect everyone to prefer a flat result based on
that particular ear transfer function + 10dB. Individual preferences are bound to fall somewhere within that range, but I doubt it would be perfect for any single individual. Some may prefer a bass shelf instead of a slope for example, others may prefer less ear gain, hence why preference research is key. Things get into trouble when people start treating it as a "one true curve", or some sort of guarantee of neutrality, which its not. Rather, it is an indication of how something performs relative to
the bass to treble delta that most people prefer.
Regardless, we're moving to showing the flat DF compensated FR against the preference bounds from the Harman research to better express this point (which is still 10dB for bass to treble delta but it'll make more sense when you see it). I'm working on the video for that as we speak.