Most headphone reviews today lean on the same piece of research, whether they say so or not: the Harman target curve, a frequency-response preference model built from years of listening tests at Harman International. It’s become the closest thing headphone reviewing has to an objective standard — a curve a pair of headphones can be measured against and scored on how closely it matches. The scientist who built that standard has spent the last couple of years publicly telling reviewers they’re using it wrong.
The scientist versus the scoreboard
Dr. Sean Olive, Harman’s head of acoustics research, has criticized the way his own team’s work gets turned into single-number rankings online. His complaint isn’t that the curve is being used — it’s that reviewers treat small differences in a match-to-curve score as meaningful when they aren’t. "Anything within seven [points] of a rating is statistically tied," Olive has said of the scoring system built from his research — meaning a ranked list with headphones a few points apart is, statistically, presenting a tie as a clear winner. Olive has also pointed out a more basic problem with how some of the underlying listening tests get run: they’re not always as blind as they claim to be. A listener can often tell which headphones they’re wearing just from weight and feel — Olive’s own example is the Audeze LCD-2, a 600-gram planar-magnetic headphone that’s hard to mistake for something lighter — which means a "blind" preference test can end up measuring familiarity as much as sound quality.
A challenger to the standard itself
In 2025, the Harman curve got a direct methodological challenge, not just a misuse complaint. PEQdB, a Stanford-linked project, released research built on data from 266 headphone models arguing that Harman’s own measurement approach has structural flaws — specifically criticizing Harman’s reliance on the GRAS 45CA test rig, which PEQdB describes as "a flat-plate headphone measuring device with no human-like anatomical features beyond its anthropometric pinna." PEQdB’s alternative uses a diffuse-field head-related transfer function built from the average eardrum response of 47 individuals, following the ISO 11904-1 standard, and proposes its own "Diamond" target curve as a replacement.
That challenge is still being argued over, not settled. PEQdB’s claims drew pushback in enthusiast and measurement-focused communities almost immediately, and the back-and-forth was still active in mid-2026, with rebuttal videos and forum threads picking apart each side’s methodology. Whichever curve turns out to be more accurate, the dispute itself makes Olive’s original point more directly: even the "objective" side of headphone reviewing — the measurement-driven alternative to pure listening impressions — rests on research choices that credentialed acousticians are actively disagreeing about, not on a single settled ruler everyone’s using the same way.
What that means for reading a review
None of this means measurements are worthless or that "trust your ears" is the better alternative — Olive’s own career is built on the idea that preference can be measured rigorously. It means a review that reduces a headphone to one number, ranked against a list of other single numbers, is compressing a genuinely unsettled scientific debate into something that looks more certain than it is. The more honest version of a measurement-based review shows the actual frequency-response graph, says which target curve it’s being compared to and why, and treats anything within a few points of another entry as what Olive says it statistically is: a tie, not a ranking.