Don't trust your ear! (Patreon)

Hey guys, this is so important, furstrating and fascinating at the same time that I need to share this with you immediately, as it allows me to explain better how sound is perceived by individuals and that you shouldn't ever trust your own ear.

I am not sure where to start, I already wanted to deal with this topic long before, but was not aware how important it really is and I can tell you that all my previous tunings I did before  sound really off to me right now. Not because my "taste" changed, or mood or something, in fact the acoustic properties of my ear canal obviously had changed during the last weeks. Therefore I want to stress again how important it is to determine your own resonance frequency for the Earfun Free Pro, in case you want to hear the "perfect" tuning like I originally had planned.

For IEMs the main sound defining factor is the lenght together with diameter and shape of the earcanal and at an approximate length of 2.6 cm you will get a main resonance at 6-7khz which is half the wavelenght of sound travelling with a speed of around 340m/s. You will get additional harmonics at 12khz and 18khz, although the latter shouldn't matter for most of us anyhow, except you are still in the 20s, I could easily hear above 20khz when I was younger, now my limit is slightly above 16khz. The response of an IEM should "fit" into these peaks and dips to be able to create a smooth response, which if it manages should offer you a pretty smooth response with no additional peaks unlike with many over ear headphones that interact way more with your outer ear and tend to generate additional peaks, not even talking about the position dependent changes in sound. That's why I think that IEMs should generally be easier to tune if tuned correctly towards a particular resonance.

Usually I hear the main peak at 6.5khz with the Earfun Free Pro, this is with my "clean" ear when coming from the Otologist. Over time of course some ear wax and dirt etc will accumulate inside the ear, being pushed closer towards the ear drum that you cannot reach with any Q-tips etc. My Otologist told me that audio engineers visit him at least every 2nd month as they are all well aware of this effect because this additional covering or film or how you want to call it can have drastic effects on the acoustic propierties of an ear canal, which means my resonance simply shifted down from 6.5khz to right now 6k and even slightly below, yesterday I could determine the peak at around 5950, but it shouldn't matter a lot if 5900 or 6100, it starts to become more of an issue if the peaks shift more than 200-300Hz as that will also compromise the peaks of the harmonics, which means treble will also start sounding off, the more you move away from the main resonance.

I was not aware of the fact that my ear canal properties had changed lately (of course I was aware of that problem already before and had to deal with it even when I started to tune it for the fist time ). But now I just realised it some days ago, because when trying to optimise my tunings I noticed that too much suddenly sounded off. You usually won't notice this immediately as this shift happens over a longer period, more and more earwax gets accumulated and the brain will adopt to any change in sound within some minute anyhow. This is also the reason why so many crappy headphones can exist at all, because without having a real reference how "correct" should sound, your brain can get used to any kind of sound signature. I know this from my own experience, when listening to a bright headphone for longer will make a neutral headphone suddenly appear dark. It is so important to stay within this "circle of confusion" so that the headphone should resemble neutral speakers in front of you and the IEM either the headphone or again the speakers. This would be my final goal, having found a response curve for both headphones and IEMs, which perfectly imitates a smooth frequency spectrum in front of you. This is what Harman and many others tried with their targets, although they are all far too rough and too generalised and can be only be used as a rough guide for any tuning.

Because of this I started more or less from scratch and created a tuning which would sound perfect to me in my current state, and I am simply blown away how much the sound has improved again by simply listening, seeking for all the mini peaks here and there and then trying to get the overall balance right comparing to a referened HD600 again and again not too drift too far away from the overall timbre... this is a tedious process, I cannot stress how much effort it takes to really get it perfect, so that nothing in the spectrum sounds somehow off or bothers me. 

What does this mean? I have prepared a tuning for the Earfun Free Pro right now which I am not sure I enjoyed ever so much like my original tuning back then. The effect you get when suddenly the frequency spectrum is smooth and also with a more or less correct balance is just stunning. I cannot stop listening, I don't even want to listen with anything else now, but of course this tuning will again sound off after I get my ears cleaned...

This little "error" has a huge benefit though: I was able to create a tuning which should sound really perfect to all those who usually hear the peak for the Earfun Free Pro close to 6khz. The difference between my original 6.5khz tuning and my new 6khz tuning is insane. The entire coloration is gone, and although the overall balance is maybe not 100% perfect, it is still a little bit too bassheavy now, and maybe there is some inconsitency in the mids, I simply have no further filters left because I used all the others for the fine details, I highly doubt I can manage to make it even "more perfect" for someone with a 6K resonance. I will try to deal with the bass tonight and will post my final 6K settings tonight.

Because of these issues I had originally planned to visit my Otologist tomorrow in order to let my ears clean and hear "normally" again, but I decided to still stay in this state for some further days in order to do further optimisations, further measurements and maybe also take a more detailed look at the Air Pro 2 tuning, because I could at least create an "ear verified" tuning for the given resonance. On the Air Pro 2 I perceive the peak a bit lower probably because the tips are shorter and the insertion not that deep. To simulate that kind of shift from 6Khz to 6.5Khz the insertion depth is less than 1mm for the couplers, thus you must understand how crucial it is for me to exactly measure at a correct frequency, otherwise all measurements will be off and not represent at all what I am actually hearing. I am still not sure how well I can manage to "extrapolate" further tunings out of measurements I have done so far, as usually couplers shouldn't be judged at all above 7-8khz due to lack of consistency and lack of precision, but that is exactly the area which varies the most. I just think that the final resulting "curve" should still look more or less the same for all different insertion depths, and having experienced 2 different depths on my own now, I simply want to experiment a bit more as long as I have these "wrong" ears. I still assume that if the final response curve is correct for one particular resonance, it should also be correct for the other resonances, as long as it is measured with the given insertion delta. But getting to the correct curve is the problem here, because all targets so far I tried were just off, the only one which somehow worked and was a big help for me was the response of the Etymotic series, especially the dip around 10khz is the most crucial factor, how deep and how wide affects the final result a lot. But most IEMs on the market do not seem to consider these effects at all, the Grell TWS/1 even has a peak in this area, most measurements of IEMs just appear like some random painted curves towards some other curves it would seem. Also all published measurements are just "wrong" because the reference plane will usually give you a measurement with an 8khz peak but nobody will ever percieve the Earfun like that, I asked my son to verify his frequency and it was at 7.3khz.

All this is not easy and I think I am trying something quite unique. To tailor a kind of "correct" sound with no obvious peaks towards different ears. I am not sure something like that has ever been done, but to me it is another proof that getting a smooth response, not necessarily the overall balance, which is not that important, the sound will still sound good with more bass, less bass, more treble, less treble, but being free from any obvious peaks makes the sound as realistic as I have never experienced from any other headphone I have heard up to date.

All this is not my job to be honest, I can do this only because of all of you, "wasting my time" with listening to music or pink noise and getting chills out of it, and the best thing is: I have lots of fun doing it and just wanted to thank all of you who somehow believe in what I am doing, but I really need this support in order to maybe offer you a sound which you can simply dig in and get lost without having to spend 5000$ for an amp and 10000$ for a headphone, not even talking about the DAC and cables and all the other important bling bling.

All this "tuning" will of course remain a work in progress, and the more I deal with it, the more I am learing and can hopefully improve for future. I need to take another look and try to adopt my original 6.5k tuning towards my new "curve" and then check if this will sound the same as now. I will also try to achieve similar results for the Air Pro 2 and maybe other upcoming IEMs where I can get access to the internal DSP and PEQs. I hope I can directly transfer my current results and curves towards other products, but I think most of the success is still in careful listening and optimising more and more trying to find further issues, instead of trying to tune towards some curve...

Thanks to everyone!

Yours, O.