oluvsgadgets
patreon
Earfun Free Pro 2 - the new "newtral"
🕑 Added 2022-02-14 00:56:52 +0000 UTCBased on my findings from here with an additional dip trying to recreate some measured effects of my outer pinna I spent the last couple of days optimising this new setting and hopefully achieve an even better result, if possible at all. I mean there must be some point where there cannot be any improvement done, but so far I had the impression that every iteration I did led to some significant improvement approaching the final "optimum" better and better, although I can only talk for myself, as it is still just my own ear which captures the sound and will differ quite a lot from someone else's ear, I have no idea how your outer ear influences the final spectrum compared to mine, but I think that recreating both effects of the outer ear and inner earcanal trying to model them into a final response which will hit your eardrum is definitely necessary to make the sound of an IEM appear as natural and simply "good". What is "sound quality" all about, how much bass or how much treble, etc? For me it mostly comes down to how natural and smooth the recreated spectrum will sound especially compared to a spectrum created by a frontally placed speaker which is just smooth with no obvious peaks or other anomalies. You can distinguish the HHHH, FFFFF, SHHHH, SSSS frequencies integrated into the full spectrum but smoothly blending over one into the other with no part really taking over. A pure pink noise can sound just extremtely pleasant, but at the same time it's extremely hard to recreate in the same way with headphones or inears due to the complex acoustic interaction of all factors. The spectrum can be smooth to some degree, but it will mostly contain errors which the brain will try to filter out anyhow and take it as granted, and it can also have some slight "color" which the brain will also accept to particular degrees as long as it's rather a broad band deviation...
But I think that if pink noise will be perceived as sounding "natural", "smooth" and even exciting or "beautiful", the sound will just be "good" regardless how it was created, what driver type or technicallity was used, and I first need a real proof to be convinced about the opposite being true!
Thus understand my problem, that I was sitting there again comparing speaker, headphone, IEM with my brain trying to adjust everytime and make all of them appear somehow "similar". I am not an expert in this and it needs some time to really detect flaws, thus I took several approaches with more bass, less bass, somehow more mids, etc and always had the impression as if some particular version sounded better than the other one, finding another one which was still better while having to compare all sources back and forth to always have some kind of "reference" in mind and I think I approached one variant where I am not sure I can manage to improve it much further at least or myself and maybe not with these tools. It's fascinating that switching back to any other of my previous versions makes this one stand out as sounding simply best. My previous" neutral" setting sounds really bad to me now, but it just proves that the brain will try to adopt to any kind of sound and without having the constant reference in mind, it will accept any kind of sound, which of course is not the point here.
But when listening to the spectrum over my Free Pro 2 reference unit created with these latest settings it sounds much smoother to me than any of my previous ones. I always heard something was off with my pink noise spectra I created, but I couldn't quite track down the erros precisely and right now there is still some slight SSSS dominance which I tried to correct with getting maybe another spare filter for the 11k range, but all my tests so far failed. I came to a point where adjusting any of the filters just a bit would yield to a worse result for myself. They all interact with each other quite strictly, adjusting just the bass a tad, can change the overall balance quite a lot and although it may still sound decent, it may be less "correct". I mostly concentrated on the upper mids and treble because this is really the hardest to get right, while bass amount will usually vary with speaker placement, room geometry etc, upper mids and treble are not affected that much by room acoustics, but rather by the incoming angle of the perceived sound which will enhance particular frequency ranges, while toning other ones down, with peaks shifting all over.
The following example shows some HRTFs of my friend Max covering sound coming from 0° - 270° limited to the horizontal plane only. I also overlaid an average of all angles:
You can notice that depending on which angle the sound is coming from there can be differences more than 20dB within the treble range, besides he seems to have a constant dip at exactly 10k, while I perceive it closer to 9300Hz, thus everyone will obiviously hear something different and even the "correct" bass amount can be just a matter of interpretation.
Smoothing the spectrum even more which I tried, can make the final sound appear a bit dull. I perceive the current setting more transparent, maybe it's a tad too bright in the range above 16k which I cannot perceive that well anymore, but reducing this range more made the sound lose most of its "detailness" and made it appear simply more boring and dull to me. Keep in mind that I didn't boost any treble additionally, but getting a bit different balance between bass, mids and treble, could open the upper most register way more and make the sound more detailed at the same time, I also do not often listen at levels much higher than half volume, while at higher volume levels I would definitely prefer less treble.
In the measurement below you can see some settings overlaid with green being the original neutral tuning, red the previous setting with the added 7K dip and black being the latest setting I was working on. I even had to reduce bass more than before while I think "perceived" bass amount even increased with the sound becoming more open at the same time, if that makes any sense. It's also shocking to see how close all measurements look, they hardly differ more than 1-2dB in most areas but overall are still perceived as different. This also proves that measurements cannot really show you any real quality or let you interpret the sound in any detailed way, as small changes in bass can also affect the perceived amount of treble. I can only stress again that it's fucking hard to get this right and you can probably only get it right for yourself.
I would really recommend some torough testing by applying some external filter and switching between different settings precisely which unfortunately is not that easily achieveable from within the Airoha app, but using my values within EqualizerAPO on a PC with an empty EQ for the Free Pro 2, lets you switch between different settings easily and more comfortable, that is also how I usually approach and achieve my values by adjusting them through EqualizerAPO first until I really think the result is good enough or better than the other one. I need to be able to hear any nunace in change between one or the other setting . When I think it's worth being listened to or finished, I transfer those settings into the Airoha app and start my daily listening, just listening to various kinds of music and see if there is still something that would bother me or sound off, and while I still won't dare to claim it's perfect, it sounds pretty darn good to me right now and I was listening with this setting the last 2 days in order to see or "hear" if I can still find some parts that would bother me after a while, but so far I am really enjoying it a lot. I still need to try it on all my other units as well, and maybe do some further small adjustments before regarding it as "final", if there ever will be one. I rather see it as a kind of "novelty", like getting a new headphone every time, but instead of buying one again and again, you simply adjust some numbers and suddenly have something new hopefully better:
Therefore from now on I will post all my upcoming settings in EqualizerAPO "language" so that you can immediately include them as configuration files for EqualizerAPO without the need to edit them at all, just select include from within the control menu and load the txt.file and the EQ should be active systemwide on your PC and you should also hear the Free Pro 2 with this setting correctly. I verified that the settings from the Airoha app correspond 1:1 to those created within EqualizerAPO including all lower values than -20dB etc.
I would welcome some feedback from those with a resonance of 6000Hz or close to this. As soon as I get some verification from others that I am on the right path and the setting improves the final result for those who already perceived the previous setting as better, I can start my work on creating the settings for all other resonance frequencies and hope that I can improve the results for those affected as well. It's really not easy to translate my personal setting which may work perfectly for my own earcanal resonance on to other resonance frequencies hoping that those small deviations will still sound "the same" to you like to me with even small filter adjustments causing big changes... I just want to say, that what I am doing here can be just a base for yourself trying to optimise the setting even more to meet your own target better. So far it would seem to have worked quite well among other people as well, but I cannot guarantee that you will get exactly the same effect I am getting. You may have your peaks or dips at different positions than me, and I would assume that some generalisation among different HRTFs could be necessary, because in this case my current setting is rather optimised for my own ear, therefore you may try to shift some filters in frequency and listen if this improves the result for you, I readjusted some frequencies a bit compared to my previous settings, simply to achieve an even smoother balance between all frequencies. EqualizerAPO is a great tool for this, as it let's you adjust frequencies gains and Q-factor precisely and let you immediately hear the change without the need having to update any settings manually all the time.
I wonder if you hear more bass now, or less bass or too much bass. For those complaining about too little bass before, I think this setting somehow fixed the problem. And the others complaining about too much harshness should be satisfied as well, although most of the harshness was already fixed to the additional 7.4k filter, the rest was simply fine adjustment of the remaining frequencies.
Thanks for all your support so far.
