Do I need to invest in new audio hardware for "neutral" head Options

[nesf]

I'm wondering about this and unsure with the amount of misinformation in audiophilia whether how big a problem this is. So I'd like to get some neutral sounding headphones but I've been recommended that I'd need to look into getting a neutral amp and DAC to go with them otherwise I'm defeating the purpose of going neutral in the first place. How big of a problem is this? Do amps and DACs colour to a significant extent in the audio chain or are we talking about very minor differences here? I'm very new to all this so I'm getting out of my depth quickly! From my understanding amps especially can colour a sound in a particular direction (thinking SS here) and this is part of the whole "X amp sounds great to me with Y headphone" debate.

Mostly I'm wondering whether I can get a neutral set up (to satisfy my curiosity) without spending an awful lot. Current DAC/Amps are Fiio E10, E7 and Xonar DG soundcard.

[pdq]

The thing to watch out for is a headphone output on a single-supply device, such as a DAP, that is capacitively coupled. If the capacitor is too small then you lose some bass response with low impedance headphones.

Othe than that, any modern DAC and amplifier should have flat response and will not "color" the sound.

[DVDdoug]

From my understanding amps especially can colour a sound in a particular direction (thinking SS here) and this is part of the whole "X amp sounds great to me with Y headphone" debate.

There are many ways to build a good amp. With solid state electronics, it's cheap & easy to build an amp with low noise, low distortion and flat frequency response. With tubes, its more expensive and more difficult. It's especially difficult to build a good tube power amp, because the audio has to pass through a transformer.

If changing tubes (or "matching" or "selecting" tubes) makes an improvement, it's a bad design, IMO. Tubes age, and a good design should tolerate normal variations without affecting performance!

If you go by measured performance, I believe the best solid state amps will test better than the best tube amps. But since both designs should be better than human hearing, hopefully they will sound identical.

"X amp sounds great to me with Y headphone" debate.

If the amp has (relatively) high source impedence, the frequency response will vary with the headphone, since the headphone impedance varies over the frequency band. A solid state headphone amp should have a source impedance in the ballpark of 1 ohm and should be immune to headphone impedance variations. (Regular soundcards may have higher impedance.) The source impedance (or damping factor) spec is often not published... They usually just publish the recommend load impedance.

And as pdq mentioned, the output capacitor can have an effect too.

[nesf]

Thanks guys, so basically so long as my Amp/DAC measures decently in tests I don't need to upgrade them unless there's an impedance issue, in which case I upgrade the amp based again on measurement reviews? NWAvGuy has a review of the E10 coming out next I think so I'm looking forward to his results. Are there any other people I should be reading for measurements of low/mid end gear?

[DVDdoug]

...so basically so long as my Amp/DAC measures decently in tests I don't need to upgrade them unless there's an impedance issue

Probably... But it's against the rules of HydrogenAudio (TOS#8) to talk about sound quality based solely on measurements. We can easily measure things that we can't hear... I've got an oscilloscope in front of me that goes to 100MHz, but I can't hear that high!

[MichaelW]

But it's against the rules of HydrogenAudio (TOS#8) to talk about sound quality based solely on measurements.

Could we clarify that? I know that it's stupid to worry about measured frequency response at 40kHz and so on, but I had thought that some facts about hearing were empirically well established, so that it was entirely legitimate to say things like "This amp measures flat to 1 dB in the range 10Hz-25kHz, so you will hear no problems in frequency response (unless you're a dog)." That would be a statement not directly the result of a DB listening test, but very soundly based on a great deal of testing of the appropriate kind. Or have I got it wrong?

[nesf]

Yeah that's what I was thinking when I was asking the question. If the curve's not flat outside the range of my hearing will that affect things? I mean I've read arguments about ultrasonic positioning cues and such but I'm really new to this so beyond a "the human ear can't hear beyond range X-Y" I'm pretty clueless as to the importance of very high and very low frequencies with respect to sound quality.

With respect to rule #8, is a theoretical discussion about whether such frequencies can affect perception ok or is it banned? New here etc.

Edit: Again, interested in any double blind or blind studies that have been done with this rather than subjective opinion. Would test it myself but have no idea if I have equipment of substantially different frequency response outside of the audible range so not in a position to do any testing myself unfortunately.

This post has been edited by db1989: Jan 5 2012, 12:34

Reason for edit: Please stop pointlessly full-quoting the previous post.

[MichaelW]

With respect to rule #8, is a theoretical discussion about whether such frequencies can affect perception ok or is it banned? New here etc.

Many tests have been made; none have succeeded. As I recollect, there is one recent paper that claims some success (search on Ookashi, I think), but at best it's not replicable, and there seem to be other problems. Without robust and replicable double blind empirical results, theoretical discussion would be regarded as a waste of time--properly so. But you might want to grovel around through the forums to see what has been said. But also, if you could produce real results, you would become famous.

[nesf]

Thanks. So the evidence so far leans towards it not making an audible difference or at least it has not been shown in any decently constructed study that it does? (I lean towards the camp that says if you want to claim X is true but all studies about X haven't found it to be true then X should be treated as false until proven otherwise)

This post has been edited by db1989: Jan 5 2012, 12:34

Reason for edit: See post #7.

[MichaelW]

Thanks. So the evidence so far leans towards it not making an audible difference

Leans so far as to be pretty well horizontal, it seems.

[xnor]

Take a HD555/558 and drive it from an amp with 120 ohms output impedance. An up to 7 dB peaky bass boost resulting from the output impedance is quite audible.

@MichaelW: ""This amp measures flat to 1 dB in the range 10Hz-25kHz" is useless if you don't know what the load was, if the amp was loaded at all during the measurement. You can find many RMAA graphs on the net that are useless, because they were made from headphone out to line in (10 k ohm or more).

This post has been edited by xnor: Jan 5 2012, 13:10

[nesf]

Thanks, yes I'm aware of some rules of thumb for matching output impedance and the impedance of the headphones. Saved me from a nasty surprise if I'd bought a Fiio E9.

[xnor]

The worst thing imho is that most manufacturers still don't seem to be aware of the problem, or just don't care.

[nesf]

Yeah. Though to be fair to the E9 it was always supposed to be (as far as I can make out) for headphones that the E7 couldn't drive on its own which means the 10 Ohm impedance on the 1/4 headphone out wouldn't be a big issue (the 40 odd Ohms on the 1/8 though is another story). Still, it really should be marketed as such and things like the O2 amp show you can drive most headphones without the need for high impedance on the headphone out anyway.