Touched by an IEM guru, running the SBAF Dick Gauntlet, and why target curves for IEMs won't work

Today's strip seems appropriate here.

 

Y'all just jealous. ‍♂️

What can I say? I love it! That's what I call neutral!
But I do have to lift the volume above my comfort zone just a little bit. I think some loudness correction would suit the target very well.

 

Good illustration, thanks!

Yup, that's part of the problem... but the other part is significant variance among human HRTFs. I tend to think that a "one-size-fits-all" ear simulator is simply not doable.

On a related note, here's an interesting article on the development of an electronic travel aid for the blind. They started out using non-personal HRTFs from the CIPIC Database, but soon realized they needed personalized HRTFs to achieve satisfactory sound externalisazion for all participants.
https://www.eurasip.org/Proceedings/Eusipco/Eusipco2007/Papers/d4p-h05.pdf

 

Well, I'll just leave this here cuz I think it's relevant:



To connect that to the current narrative:

John is a neuro-scientist who has studied and memorized everything there is to know about the science of sound. He knows precisely what "neutral" means, and could precisely represent it in math. But he has never heard or experienced "neutral" speakers.

Inversely, there's Wayne, who is an ignorant doofus who listens to "neutral" speakers all the time every single day, but has never tried to study the speakers, and is completely oblivious to how they measure or what their physical properties are. But Wayne is a master of the art of writing, so he won't have any trouble describing his feelings and experience. Wayne also happens to have insane memory and can recall all events in his life down to the minute details.

When John and Wayne both listen to a random pair of speakers, which one of them will be able to better describe the way those speakers sound?

And for the objective folks who love blind tests: if you blindfold John and Wayne, does it make a difference?

 

I think a more appropriate question is: Who to hire to build a great pair of neutral speakers?
I'd say you want both, but who to choose if you can only hire one person?

 

Interesting choice. So basically Wayne will do a lot try and error, but will know if accidentally has built a great speaker. John will know what to do, but can't tell, if he really was successful...

 

See, this is what folks on here have been saying for almost the past decade. Compare Rin Choi's DF calibrated measurements with Marv's "this is the shit I hear curve" calibrated v2 IEM measurements from changstar from back in 2012. You'll see that the DBA-02 IEMs and the UERM were almost spot on - I've always used those as a reference to look at other IEM measurements.

Now what would be way more interesting is if you yourself came up with your own target and shared the results with us. I'll maybe try to see if I can work out something myself with the IEMs I have. Once something that sounds flat is a flat line on the graph it becomes much easier to see any deviation. A 2db deviation could be a dealbreaker to some.

 

It's not so easy. At my previous company we had project were we tried to find an improved compensation curve (or target if you will). The people participating in the exercise were all trained listeners (and yes Marvey, more than K-pop on IEMs). We were given EQ controls that shaped the a flattened headphones response. Flat here really was flat, as in a flat curve on Hats. That doesn't make any sense as a target, but it's a good starting point to shape an EQ on top.

Long story short: it wasn't so clear on how big the 3 kHz bump had to be (gain and Q) . It was certainly on of the areas of variation among listeners. Together with the low mid and 9-10k range. After all that study (took 2 years), they ended up continuing to use a curve which is very similar to Harmon, but not quite the same.

 

On that, while I'm not shy to use EQ, I'd still far prefer something that has a flat-ish response without having to resort to EQ. I have heard the same headphone (HD580) in both conditions: one where it's the stock headphone EQ'ed to a flatter downward-sloping curve, and one where it's the same thing modded slightly to shape the response to be about the same curve (obviously not the exact same since I don't have godly engineering chops).

And I greatly prefer the acoustically tuned one over EQ.

Not just because I was the one that did the work either (note: I also enjoy Marv's KISS mod on HD650, and Luis' Code-X).

Something about EQ just doesn't sound natural to me. It seems to inherently mess with something in the upper mid that I don't quite like, or at least that's how it is with the HD580. An acoustically tuned headphone sounds far more natural and more "at home", more like "this is how it should sound like" to my ears.

Even when I took the "inverse route" of trying to EQ the stock headphone to have the same response as the modded headphone, the modded headphone still sounded better to my ears.

Maybe it's because I'm not trained to do critical listening, but... I'd say relying on just EQ to get a "flat response" is not the best method, or at least it isn't to me.

 

There are arguments ad nauseam on talkbass.com (and tgp, gearslutz, etc) about tonewoods and their effect on sound. Often "science" will be trotted out to "prove" one side of the argument. The problem is that it is impossible to remove human bias from the equation since at the end of the day a human has to pick up the instrument and play it. So if the person believes there is a difference, there likely will be - for him. While there are certainly advantages to having quantifiable metrics, the last nanometer into the human brain is so subjective the science gets largely washed out. That doesn't mean we stop trying to find objective measures, but it does mean (to me) that they are not the be-all, end-all, and at some point if it sounds good, it is good. Reality is indeed perception.

btw, the guy at the Schiitr spoke highly of you today.

 

Unless your using linear phase eq which will induce some amount of delay, you introduce phase anomalies around each area you EQ. Besides EQing with linear phase EQ is a rotally different beast with much more subtle changes for the reason mentioned above.

That's probably why well implemented physical mods are superior to an EQed stock headphone.

 

You're absolutely right, the EQ-ing is not a straightforward problem. I can't go into details, but lest say the "flattening" EQ was very carefully considered as well as the "shaping" EQs. The flattening method probably took the longest time in the project.

 

While perhaps not a 1-to-1 correlation, when I'm looking for bass amps, I typically want something that sounds "good" to my ear without touching the eq. Then the eq becomes either an effect (with a particular sound/tone in mind) or as a fix for resonances in a particular room or to deal with a band instrumentation lineup.

None of that translates necessarily to "flat response." The amp eq set flat invariably is not flat at all, with different amps having different signature profiles. Likewise, what one player thinks sounds good another might hate. Lots of guys love Ampeg SVTs - I usually sigh when I have a supplied backline and have to play through it. Just isn't what I'm trying to get at sonically.

Then you have the other universal problem of keeping yourself on the farm after you've been to Paris. Once you start moving into better gear (bass or audio), the law of diminishing returns is in full force, but there are sublime things to be had if your open to it. Luckily my wife actually encouraged me to plunk down stupid money for my main bass, but I've also played other instruments of similar cost that didn't float my boat. Some combinations just work, other don't. But for someone else, could be 180 degree difference.

Humans are weird...