IEM Impulse Response. What does it prove? How does it translate to music?

[600_OHM]

Well ... you can hear it more easily when the impulse response is purposely manipulated. Amirm did a great review of the AKG K240, which by all accounts is a loser to most of us in this realm.

However, what I think a lot of people miss is that with the K240, the leading edge of impulses are chopped off (by what means - group delay - dunno), so that that long-term desensitization and ear fatigue are lessened in long-term use, especially metallic high-energy things.

I mentioned this in the actual K240 review, but owners may want to look out for these tell-tale signs: you never hear the initial strike of a cymbal or triangle for instance, and only hear the "ring". That ring is also dithered so everything is splashy and less of a solid clank. It's even great fun to watch if a video and audio are in time, and you see the initial strike but never hear it!

That's why I tell fellow audiophiles to pick up a pair if they are bored! Not for sonic accuracy, but to go on the hunt for little details your ears might miss with tracks you know by heart. With things like cymbal strikes and drums toned down, you can pick up maybe an extra note your ears missed - it's in the track, but your ears are slightly desensitized at that point to miss it with the more accurate cans.

Some in our crowd might even think of the K240's weird impulse response - at least to my ears - a reason to own one just for entertainment at times.

 

[Tarnith]

So accurate representation makes sense, but how could the impulse possibly cut before time?
Also, is that a fact that a shorter decay is more accurate? I ask as it doesn't seem toalways relate to high quality IEM\Headphones models.

And are you claiming that it's undeniably the definition of "speed"? I ask as I see many people ridicule that and relate most to frequency response and THD.

Overly damped driver could result in a lower level signal being under-represented. This could be seen in an impulse that decays more rapidly than expected. I'd also expect to see this in FR, but it could be more obvious what's going on in the time domain. I'd really like to see wavelet spectrum plots of IEMs but I don't own a coupler.

I feel like it could be quite revealing as to why some people find certain driver types or different IEMs with similar FR Plots to sound so different. (Aside from the major difference, which is they seat differently in different ears and since everyone has different ear geometry everyone gets a different FR switching between two IEMs anyway)

Personally I'm a fan of more data, even though a lot of people seem to think it doesn't matter. You can tell a lot about a driver/enclosure combo by how it shapes a square wave or handles a step/impulse.