- Apr 1, 2019
It's hard to be tough on something so inexpensive, but that THD plot is a bit much. IEM's shouldn't be distorting this bad in the modern day. Distorting highs and lows? Come on..
I would think unit to unit variation would be the key variable if you were gonna make that point. If there is high unit to unit variation then EQ's you find on the interwebs won't really get you close to where you want to be.....and in that context frequency response is on balance more important than distortion. It is hard to guage unit to unit variation.....you can go off brand reputation in that area, as well as look at how closely the two channels are matched as an indication, and of course if a reviewer or people have measured more than one unit then you can start to see the spread.I wish Amir started early in measuring IEMs. It would have saved me a lot of time/money and making my iem collection way smaller than it is now.
From the few iem reviews that Amir have reviewed/mesaure so far, it tells me that distortion is more important measurement than FR. As long as your player/source has a decent PEQ an iem with low distortion can be easily be tuned to anyone's liking.
I guess the holy grail is finding the cheapest iem with low enough distortion and tracks the harman curve as close as possible.
This.I think spending 20 usd for the Moondrop Chu makes more sense.
Below measured against the Harman curve.
I think that would have improved the treble, perhaps even with improvements up to "20kHz". It's harder to EQ the treble, so I find sometimes it's useful to align the measurement so the treble is right, but it depends from frequency response to frequency response which is the best approach for me. It wouldn't help distortion though as you "always end up at the same place" no matter how you align the target on the frequency response....assuming you EQ all the same areas to comply with the target.....(just that in this case with this IEM aligning the Target lower on the frequency response would have meant possibly a better experience in the treble and up to "20kHz", because he didn't EQ the higher parts of that area).@amirm This is a perfect example where the misalignment of the frequency response curve in comparison to the Harmon curve has pulled you in a bad direction. Could I suggest if you slid the frequency response curve up about five Db you would have a very different target for what the EQ would be. In fact I think all you would need is a boost at 7KHz and then you can pull down the broad bloat in upper bass. Would you mind giving that a try and showing how the graphs look? I suspect it would also help the distortion quite a lot to EQ it that way.
How do you compare HD381 to HD387 review here?Superlux' products used to be bargain gems, mostly available only to people living in Europe (and I guess Asia) back in the day, since Thomann sold them for chips. I bought a bunch of their stuff many years ago (about a decade ago); HD668b, HD669, HD662, HD651 (these were €5!) and the HD381 IEMs. I also picked up some of their accessories, like a DSLR microphone mount and god knows what else. The market moved on a bit, but cool to see they're still around.
Given that the 94dB is loud enough to permanently damage your ears with extended listening, and 114dB is loud enough to permanently damage your hearing in minutes, I'd say it's totally acceptable.For the price, I put Great!
The Frequency response is actually quite good, as Amir has the divisions at 2db, not the typical 5db/.
The distortion....hmmm.....for this price, still quite good.
The relevance of those distortion graphs is mainly in bass EQ. I've recorded 0dBFS sine sweeps of my headphones on my miniDSP EARS after setting amp levels towards the upper end of my normal playback levels and it's around 84dB at 1kHz, but then with nearly +10dB bass boost EQ that would pull the relevance of Amir's distortion graphs up to around 94dB in the bass - so for me I just make sure that the 94dB distortion line would be clean in the bass and hopefully no glaring problems higher up in the frequency range at that same 94dB - but you won't really be listening to music at 94dB except in the bass so you can probably assume your actual experienced distortion level is lower than the 94dB line everywhere on the frequency response except in the bass. That would be valid for me at my listening levels. I suppose if you listen ultra loud at times then you'd want to push that up to the 104dB line in the bass and start giving the 94dB measured level a bit more relevance in the areas above the bass. But you won't be damaging your ears by listening to comparitively high levels in the bass - that's why H&S Guidance is based around measuring noise that has been A-weighted - ie reducing the contribution of the bass to the measured dB level. So it's not as simple as just saying that 94dB or 104dB is enough to make you damage you ears, because it depends on where in the frequency range you're referencing that measured dB level.....and besides the 94 & 104dB levels you're quoting are Amir's distortion graphs which is referenced to those values at 400Hz, and the rest of the frequency range of that headphone that he's measured sits relative to that.....so you'd have to work it back from that where your EQ boosts would place you for various parts of the frequency range for that headphone.....it's a bit hard to explain, but it's not as simple as saying "94 & 104dB will make you go deaf". I think it would be useful if Amir additionally included an 84dB measured level of distortion in his headphone reviews - that would be more relevant for me (& probably others) for distortion levels I'd see in areas above the bass....due to the mechanisms I've tried to explain in this post.Given that the 94dB is loud enough to permanently damage your ears with extended listening, and 114dB is loud enough to permanently damage your hearing in minutes, I'd say it's totally acceptable.