We measured our HRTFs to try PERFECT headphone sound

We finally did it—we packed up the team and headed to the UK. On the surface, it was a chaotic travel vlog full of rain, banter, and questionable snack decisions. In reality, the whole trip revolved around one very specific goal: getting our individual HRTFs measured so we can take our headphone reviews to the next level.

The main reason we crossed the Atlantic was to visit the HRTF measurement dome at Imperial College London—specifically inside the Dyson School of Design Engineering. The lab houses a rotating measurement system that captures a full 360° head-related transfer function (HRTF). In simple terms, it measures how our individual heads and ears shape incoming sound.

For years, we’ve relied on industry-standard rigs like the 5128 and GRAS systems using Diffuse Field HRTF calibration. Those systems are incredibly useful—but they’re not our ears.

So we’ve decided to take the next logical step and move to measuring headphones on our own heads as well.

By capturing our own Diffuse Field HRTFs, we can now:

  • Measure headphones directly on our own heads using in-ear microphones
  • Better understand why we hear things differently from one another
  • Test assumptions about “technicalities” and frequency response at the eardrum
  • Put tools like AutoEQ to the test in a way that’s actually personalized

To be clear, this isn’t about winning points for a narrative or proving people wrong. This is scientific inquiry, so it’s about answering the myriad open questions currently still out there when it comes to headphone sound. We genuinely don’t know what we’ll find, and that’s the fun part!

Walking into the dome at Imperial felt surreal. From the outside, it’s just a very British-looking tower. Inside, it’s a tightly packed spherical measurement system with a rotating chair and a 5° speaker increment arc. You sit there for roughly 30 minutes while sweeps are played from every direction, then you get 3D scanned and photographed for morphological comparison.

We also got to learn about the broader research happening there—like large-scale HRTF databases, work on estimating individualized HRTFs from scans or photos, and even testing the sonic adaptation of ferrets! Fascinating stuff, but for our purposes, the key takeaway was simple: we now have calibration files for our own heads and ears which we can use to measure headphones to the standard we’ve already set.

We explored London in proper gray, rainy fashion—walking along the Thames, seeing Big Ben, Westminster Abbey, Buckingham Palace, and spending far too long at the British Museum making jokes about “borrowed” artifacts. We visited Winchester Cathedral (casually about a thousand years old), ate an unreasonable amount of Nando’s, and subjected ourselves to British snack taste tests.

We also spent time in Cameron’s studio, listening to his meticulously tuned speaker setup—dual subs, independent correction, proper crossovers—the works. The imaging and depth on that system were genuinely standout. It was one of those listening sessions that resets your expectations and reminds you that the “tonal/technical dichotomy” is just a symptom of audiophiles not understanding frequency response or the playback condition of headphones. This system had both in spades.

We even explored some experimental ideas involving prosthetics, 3D replicas, and future testing possibilities. That’s all we’ll say for now.

More than anything, the trip reminded us why we do this. A week earlier, we were scattered across different rooms and countries mulling over impressions and frequency response graphs online. Suddenly we were all around the same table, measuring our own ears in a world-class lab, debating things in person, and pushing our process forward together.

This was easily one of the nerdiest things we’ve ever done in this industry—and but also the most fun.

There’s a lot more coming from what we measured and tested, so stay tuned!

10 Likes

Sounds like a very productive and interesting visit. And I look forward to hearing more about it.

“Perfect” to me would be the in-ear response of neutral speakers in a semi-reflection room btw. Not a flat diffuse field.

Video didn’t come through on the post but here’s the video of the trip with some of the findings.

Would recommend giving it a watch for the intro alone!

3 Likes

We look forward to the unqualified positive reviews following the successful neural programming.

2 Likes

I’m realizing that now the Headphones.com review crew can publish more accurate FR measurements than anyone else. And not just accurate to yourselves, but very likely more accurate to anyone. Because neither the GRAS 43AG nor the B&K 5128 has turned out to simulate a human very well. Just look at the measurements of the Sony MDR-MV1. None of you hear it the way the rigs do, and I don’t think I do either.

2 Likes

Unfortunately I do hear the MV1 the way it measures on a B&K 5128 - at least for the big problem area. That treble peak is rather murderous on my head. But yes, the general push to have ‘objective’ measurements and ‘subjective’ measurements is precisely because those two things regularly disagree, and we can now show precisely why.

1 Like

If you have your own HRTF, assuming the measurement is reliable, does that mean that in theory you could compute an EQ profile to match any headphone to any target? (Asking sincerely.) Or are there interaction factors that mean you’ll always need an in-ear mic measurement to make it work?

Every headphone has emergent interactions on different heads that need to be factored in. Getting our HRTFs allows us to understand what those interactions are for us individually, and how the headphones are actually performing in-situ.

As a lay person interested in EQing their headphones, but not in possession of any measurement equipment, would I gain anything from also getting my HRTF measured like you did?