Podcast #15 - diffuse field (Patreon)

In this podcast I'll be talking about headphone sound and diffuse field in general and why IMO diffuse field is a completely wrong approach for tuning headphones.

Aside of Harman target, diffuse field is often used as compensation for many headphone measurements (including headphones.com) and many others. Headphones are claimed to be "diffuse listening devices", nevertheless in reality you never experience music as "diffuse field" as this would mean the sound will hit you from all angles equally. Not even in a concert hall you will get this, you will still mostly perceive the orchestra playing in front of you with just some diffuse reflections contributing from the other angles. But modern music is usually recorded in studios and mixed over studio monitors which are placed in front of the engineer and for correct playback you preferably should also listen to music over speakers.

My SpX target tries to translate the sound you would usually get from speakers to that of headphones achieving a similar spectrum or soundfield. As speakers will change the sound a lot depending on the given room and placement, I captured the spectrum of neutral speakers in various rooms and with different placements always with exactly the same listening distance. Mixing all these different room spectrums together results in a smoother and more generalized representation of how neutral speakers usually sound within a closed surrounding, in future I may also add a spectrum with speakers playing in an outside environment for an even more generalized target. By doing it that way if you listen to music over speakers and then immediately switch to a headphone tuned towards my SpX target, you should still get more or less a comparable sound without the need of further adjustment or the need getting used to the given sound first.

The problem is that you cannot even simulate "diffuse field" properly, and to make a headphone sound like "diffuse field" you have to rely on measurements only and finally adjust the measured response of the headphone towards the measured graph you captured as "diffuse field" from all angles. This is normally done with head and torso simulators which on their own are already heavily generalized approximations of a human ear. The resulting headphone sound will be just something which you can get used to after a while, but it will never sound like speakers do.

For my SpX target I tried to approach the final sound by real listening and just using graphs as a rough guide which I would like to show you below. Without listening I could never achieve a realistic and convincing sounding result.

Here you can see the spectrum of the KH80 captured in 6 different rooms like demonstrated in the podcast, in white the response of the SpX-R with studio mode and averaged over several repositionings:

Below the response of the KH80 playing in our living room compared to the averaged response over 6 different rooms in white:

And finally the response of the frontally listened KH80 in our living room compared to playing allaround captured from various angles (in red) which would roughly represent "diffuse field" and you can see that both will sound completely different:

This was just a quick experiment which I would like to work on more in future and try capturing a kind of "diffuse field" soundfield in different rooms to achieve a better average which could be also used for upcoming headphone tunings, although I still think this sounds less natural than my SpX target funnily most headphones are rather tuned closer towards diffuse field and my idea would be to include as many different tuning approaches as possible featuring also more natural spectrums I captured in nature for and include them all for the listener to choose from within a given headphone.