I don't have a problem with you, as a person, because I don't know you, as a person. I have a problem with your claim; you said something that I disagreed with and then failed (refused?) to provide any evidence to support your claim(s), instead only providing (as I said before) pithy truisms as "support." That seems counter-productive (to me) for a thread asking about advice, especially where the disagreement is relevant (e.g. interpretation/understanding of specifications).
That's a relatively new "rule" and like I said, it's mostly marketing run amok (IMHO, but I can point fingers at where it seems to come from, which are all marketing/advertising sources). There is absolutely no objective/factual basis for that (beyond the OTL example above, or similar/tangential examples of that where you're talking unsafe operation) - the whole "rule of eight" and "it must be zero" and on and on is not some sort of bona fide fact that must be followed (lest we unleash C'thulu and a thousand years of pestilence and death), and just repeating something again and again doesn't magically transmute it into a fact. Let me expand on that: The primary argument for "as low as possible" is that it results in "the least" FR shift, however it is based on two assumptions; that there's a "right" and "wrong" FR response for a given headphone/system (which is a flawed assumption) and that FR shift represents "error" (also a flawed assumption). The first one is flawed because we often don't know what the manufacturer "intended" and even if we do, there's so much variability in sound from listener to listener (if there's one thing recent research on headphones has told us...), it's hard to make any sort of broad summation (IOW its hard to say a given headphone will even sound the same to all listeners). The second one is flawed because it is predicated on the first one (you have to have an "ideal signal" in order to have error/deviation from it).
There may very well be a "manufacturer intended FR response" for a certain headphone, but as long as you're not putting the headphone/amp into an unsafe operating region, that can even be ignored if you like (and before anyone gets really worked up: I'm talking about everything from differences in Zout impacting FR to the judicious use of EQ and tone controls). And as I pointed out earlier, not all headphone designs target the same thing - not all headphone designers assume "zero ohm" outputs for their cans, have the same performance ideals, etc. This may seem like a trivial nitpick but it's important so I'll say it again: when [whoever] is designing a headphone, they aren't working from some across-the-board canonical cookbook - we currently have no universally accepted standard measurement of "flat" for a headphone, and no universally accepted "standard design" for a headphone. So every manufacturer largely does whatever they want (or whatever they think they can sell), and there's a wide variability in terms of what assumptions are made about how the headphone will sound, how it will be used, and so forth. Finally, the IEC's standard for headphone jacks specifies 120 ohm Zout on the amplifier connection, and there's a reason for that - it isn't some arbitrary value (IEC's rationale is for consistently good noise, distortion, and power performance across a very wide range of nominal load impedances and devices). So there's a published standard that disagrees with many audio guides (and it won't be the first, last, or only one to do so).
This goes back to the "sound is subjective" point - it doesn't matter how well you can measure or document something, because that doesn't tell you whether or not you will LIKE something (or more broadly, how you will feel about something at all). And that's the biggest problem I have with "the rule of eight" as an oft-repeated truism - it's trying to take an over-generalized inference ("all high Z results in big FR 'error'") and make a claim about how people will subjectively respond to it ("people will find this to sound bad"). It seems more productive, in my view, to just come out and say that Zout can mediate FR shifts, and like any other change in FR it may or may not appeal to an individual listener, and ultimately they will have to make up their own mind as to whether or not it sounds good ("sound is subjective").
On the "unsafe operation" bit: the caveat to the above is that an individual listener should be cognizant of the output specifications of whatever amplifier they are choosing (this also applies more broadly to the question of amplifier importance). If the amplifier manufacturer says, for example, no lower than 30 ohms nominal is acceptable, don't plug your 8 ohm IEMs into it. Or OTOH that the amplifier doesn't have sufficient output into 600 ohms, your high-end Beyers may not be a good choice. But as long as the amplifier (and cans) are used "within specs" I'd hope we can assume things will work swimmingly (there's always probably going to be some mfgr that "does it wrong" though), and then it really just comes down to a subjective interpretation of what is being heard.
This is exactly what we're talking about above, and exactly what I'm saying is "wrong." There's absolutely nothing wrong with plugging your 30 ohm headphones into a device with 10 ohm (or even 30 ohm, or even 300 ohm) output impedance, as long as the device is stable/rated into that load (a lot of SS devices should be). Depending on the specific headphone, there *may be* a shift in the frequency response, but that isn't explicitly a bad thing - it is just a difference. You may or may not notice it, and once you notice it, you may or may not have an opinion about it. PersonalAudio.ru has actually measured and modeled the 1A, and shows no more than 3.6 dB of averaged FR shift (and that's at 300 ohm Zout (over 10:1); at 10 ohm Zout its just less than 1 dB (that's pretty small/subtle (http://www.animations.physics.unsw.edu.au/jw/dB.htm try it for yourself), and that's also averaged and going into lower frequencies where hearing attenuates (that is, the "peak" is lower in frequency than where the "rise" begins)), and no "wildly variable impedance" that I see. Here you go:
http://personalaudio.ru/raa/otchety/naushniki/so/
I'm not at all trying to tell you what your opinion of this change will sound like, but that's more or less what will happen ("by the numbers"), its up to the conscious listener to place a value judgment on how it sounds. And yes, in theory, you could do the same thing with EQ (or you could EQ down the regions affected by the FR, or whatever else). You (or whoever else) may not notice the .9 dB shift (or may not notice it without a side-by-side comparison), may find it somewhat pleasing, may find it somewhat offensive, etc (and remember: which "way" is actually "right" is also up in the air - there isn't an explicit "right way" they're supposed to sound) - that's the subjective part, and measurements can't tell you how you're going to feel about something. The change in FR will happen though, the question is just "what will you do about it?"
