Head-Fi.org › Forums › Equipment Forums › Sound Science › Would lack of amp power causes degradation to sound quality for high impedance headphones?
New Posts  All Forums:Forum Nav:

Would lack of amp power causes degradation to sound quality for high impedance headphones? - Page 4

post #46 of 81
10/9/12 at 12:17pm

Even if (and I'm not convinced yet by the article you posted) people can reliably sense infrasonic frequencies around 5hz, a half decibel rolloff is a negligable, barely audible change even with tones we can hear very well and reliably, never mind those which are "represented by a chugging, or puff or motion on the tympanic membrane". 

 

You are worried about the wrong things. I don't think a properly controlled test to see how ridiculous this is is needed, especially given how few recordings even include data in that range (and is not produced by any known musical instrument), but we can probably work something out to see if a .5-1 dB change in that frequencies in that range are perceptible via headphone. 


Edited by liamstrain - 10/9/12 at 12:27pm
Reply

Head-Fi's Best Sellers

post #47 of 81
10/9/12 at 2:55pm
Thread Starter 
Quote:
Originally Posted by Steve Eddy View Post

What on earth are people listening to that has content down to 5Hz?

 

se

 

Try movies?

 

Quote:
Originally Posted by stv014 View Post

It would be best to try some experiments on music that you think is audibly affected by the AC coupling in the O2 (can you upload/post links to any samples ?). These include:

- simulating the filter in the O2, and comparing the result against the unfiltered audio using the foobar ABX plugin

- filtering out everything above 20 Hz (also other frequencies like 10 Hz and 5 Hz), and listening to what is left

 

Note that while your headphones may have been specified by the manufacturer to have a frequency response of something like 5-35000 Hz, these specs are usually very optimistic, and the headphones themselves roll off (and also distort) infrasonic audio much more than the O2.

 

I do understand that.

 

 

How the O2 has "less bass", I think, has already been discussed to be a factor of output impedance plus other things. The O2 has lower output impedance than most of the amps I mentioned (Fiio E7, E11, etc...) so its bass would seem more "damped", but the fact is that it's just because other frequencies are no longer attenuated. Also there's the additional tidbit about the E7 and E11 being "colored" with some boosted bass (more gain for low frequencies). That's why I said it was my "subjective experience" and not a "fact".

 

However, I still would like to discuss this because I'm not sure we can completely rule out the sub-20Hz roll-off that the O2 has. Perhaps its effect is more pronounced given the right cases.

 

Also, the Audeze LCD-2 doesn't roll-off like other headphones... at least down to 10 Hz.

 

 

 

 

InnerFidelity's measurements also match with HeadRoom's:

http://www.innerfidelity.com/images/AudezeLCD2Rev2.pdf

 

You are right, though, that some other headphones, notably the DT880 600, do roll off:

 

 

And again, InnerFidelity's graph kinda matches with that:

http://www.innerfidelity.com/images/BeyerdynamicDT880600ohm.pdf

 

Quote:
Originally Posted by xnor View Post

This is a joke, right? What headphone do you propose? You do realize that the hearing threshold for such low frequencies is far beyond any headphones' reproduction capabilities?

 

We're talking ~110 dB SPL here to reach the hearing threshold. You know what's about 0 dB SPL at 1 kHz.

 

Also, what noises are you listening to? This can't be music created for humans.

 

I've heard them producing a nasty treble peak which can be cut with a bell EQ filter, not with high output impedance because the peak would still stand out.

 

It also can help a lot with non-ideal headphones. I take a small increase (boost) or decrease (cut) in distortion over nasty treble peaks, bass peaks or dips any time.

 

 

So does high output impedance attenuate high frequencies more, or doesn't it?

 

And try the Audeze LCD-2 for low-bass listening. It's not even me that is claiming legit on their ability to reproduce sub-20Hz frequencies. It's the manufacturer that claims that, and plus frequency response measurements from HeadRoom and InnerFidelity also show that. Unless you want to claim BS on their measurements as well.

 

As for recordings with 5Hz information. There is not a lot of them, I admit. But it's not like they don't exist.

 

Quote:

Originally Posted by liamstrain View Post

Even if (and I'm not convinced yet by the article you posted) people can reliably sense infrasonic frequencies around 5hz, a half decibel rolloff is a negligable, barely audible change even with tones we can hear very well and reliably, never mind those which are "represented by a chugging, or puff or motion on the tympanic membrane". 

 

You are worried about the wrong things. I don't think a properly controlled test to see how ridiculous this is is needed, especially given how few recordings even include data in that range (and is not produced by any known musical instrument), but we can probably work something out to see if a .5-1 dB change in that frequencies in that range are perceptible via headphone. 

 

According to that same article, people perceive the difference in level differently at different SPLs (sorry, really clumsy sentence. Words escaped me). So given a high enough SPL, I believe a difference of 0.5dB would be noticeable.

 

Maybe not particularly with the low frequency. I do agree, to some extent, that the roll-off of the O2 is non-trivial in most cases, however, given the correct circumstance, I think it may show after all.

 

Just because it doesn't happen 99.9999% of the time doesn't mean it won't happen at all, right?


Edited by Bill-P - 10/9/12 at 3:03pm
Reply
post #48 of 81
10/9/12 at 3:53pm
Quote:
Originally Posted by Bill-P View Post

How the O2 has "less bass", I think, has already been discussed to be a factor of output impedance plus other things. The O2 has lower output impedance than most of the amps I mentioned (Fiio E7, E11, etc...) so its bass would seem more "damped", but the fact is that it's just because other frequencies are no longer attenuated. Also there's the additional tidbit about the E7 and E11 being "colored" with some boosted bass (more gain for low frequencies). That's why I said it was my "subjective experience" and not a "fact".

 

E7 has lower output impedance (0.13 ohms or so), and E11 is around the O2 with about 0.5 ohms.  Both also have a pretty flat FR across the audio band, so that does not seem to match any of your impressions.  For that matter, they actually both roll off the bass more than the O2—unless you turn the hardware bass boost EQ on—though still by a negligible amount.

 

 


Quote:
Originally Posted by Bill-P View Post

So does high output impedance attenuate high frequencies more, or doesn't it?

 

Has anybody implied something that was wrong?  It depends on the impedance of the headphones.  For many models, the impedance rises near the very top of the range, so higher output impedance would mean that other frequencies get attenuated more, for an effective very slight treble boost compared to some other frequencies.

Reply
post #49 of 81
10/9/12 at 4:09pm
Thread Starter 
Quote:
Originally Posted by mikeaj View Post

E7 has lower output impedance (0.13 ohms or so), and E11 is around the O2 with about 0.5 ohms.  Both also have a pretty flat FR across the audio band, so that does not seem to match any of your impressions.  For that matter, they actually both roll off the bass more than the O2—unless you turn the hardware bass boost EQ on—though still by a negligible amount.

 

Then how will you explain the difference in sound?

 

Or should I assume that since the performance of the E7 and E11 are similar to O2 (where voltage is not the limiting factor), that I should consider them equally good?

 

Quote:
Has anybody implied something that was wrong?  It depends on the impedance of the headphones.  For many models, the impedance rises near the very top of the range, so higher output impedance would mean that other frequencies get attenuated more, for an effective very slight treble boost compared to some other frequencies.

 

No, no one implied anything that was wrong. I'm simply asking.

Reply
post #50 of 81
10/9/12 at 4:32pm
Quote:
Originally Posted by Bill-P View Post

Then how will you explain the difference in sound?

 

Or should I assume that since the performance of the E7 and E11 are similar to O2 (where voltage is not the limiting factor), that I should consider them equally good?

 

I would think that the difference is often overstated, but it's not like they're the same in all cases or equally good.  O2 is capable of higher current (sometimes current is the limiting factor for power output, not voltage), lower noise, lower distortion, some other improvements.  Other amps can do even better in some regards, particularly in power output.

 

 

Well, I can't exactly say for sure what you're hearing or what anybody else is hearing, but some explanations would include

 

  1. Difference in noise levels is audible and mistaken for something else — seems very unlikely for the kind of higher-impedance, full-size headphones mentioned
  2. Difference in distortion is audible and mistaken for something else — but those have low distortion already, especially for higher-impedance loads, so I'm not so convinced here
  3. Difference in bass phase for E11 — around 10 degrees off for 30-40 Hz, worse if you're talking about dinosaurs stepping on land mines or whatever is causing those 5 Hz tones mentioned before... though phase error like that at low frequencies is supposed to not make a difference
  4. Difference in channel matching for any of the amps, maybe being mistaken for something else — but matching is pretty good at higher volume settings, within a fraction of a dB, so probably not it
  5. Difference in overall loudness (unless every comparison ever made was volume-matched) causing different perceptions of the sound — this is likely, especially considering Fletcher-Munson / equal-loudness contours; increasing volume can increase perceived bass relative to other frequencies, for example
  6. Difference in what you were listening for causing different perceptions of the sound — happens whether or not anybody is trying to do it
  7. Expectations of differences in sounds causing different things to be perceived and remembered — also a large factor

Edited by mikeaj - 10/9/12 at 4:33pm
Reply
post #51 of 81
10/10/12 at 1:45am
Quote:

Originally Posted by Bill-P View Post

 

Also, the Audeze LCD-2 doesn't roll-off like other headphones... at least down to 10 Hz.

 

For comparison, I show the frequency response of the O2 (in green) overlaid on the graph at 0 dB. The resolution is not high enough to give an accurate picture (click the image to zoom), but the response is down by about 4/5 of a pixel at 10 Hz at the original size. Note that even a slightly imperfect seal will result in more roll-off, as shown by the grey traces at the bottom.

 

 

Quote:

Originally Posted by Bill-P View Post

 

Maybe not particularly with the low frequency. I do agree, to some extent, that the roll-off of the O2 is non-trivial in most cases, however, given the correct circumstance, I think it may show after all.

 

Can you provide a sample on which the roll-off is audible ?

Reply
post #52 of 81
10/10/12 at 3:06am
Quote:
Originally Posted by Bill-P View Post

Try movies?

 

As for recordings with 5Hz information. There is not a lot of them, I admit. But it's not like they don't exist.

I couldn't name even one, can you? Unless you can, there's no need to even talk about being able to "hear" +/- 10 dB at 5 Hz with headphones.

 

Quote:
So does high output impedance attenuate high frequencies more, or doesn't it?

Yeah, overall it usually does but peaks will still be there.

Reply
post #53 of 81
10/10/12 at 7:44am
Quote:
Originally Posted by Bill-P View Post

 

Also, the Audeze LCD-2 doesn't roll-off like other headphones... at least down to 10 Hz.

 

 

 

 

InnerFidelity's measurements also match with HeadRoom's:

http://www.innerfidelity.com/images/AudezeLCD2Rev2.pdf

 

 

The resolution on these charts are scaled such that it would be very difficult to see a 1/2 dB change on the graph. Any given sample of the headphone might roll off that much or more on its own. 

 

All the listening data you have provided for people perceiving infrasonic frequencies require huge boosts in spl at those frequencies (the threshold of audibility/perceivability increases massively as you go down the scale below about 80hz). I think this remains a non-issue. I'm not saying it cannot happen (your 99.99% comment), I'm saying there are other much more pressing concerns regarding overall sound quality to deal with before you worry about the theoretical .001% roll-off issue.

Reply
post #54 of 81
10/10/12 at 2:07pm
Thread Starter 

Okay, then let's disregard that for a second and say that I'm wrong. The O2 doesn't have any "roll off" that is significant.

 

This just occurred to me. Going back to the same thing about voltage drop earlier, assuming the following:

 

[V out] is the output voltage from the amp.

[Z out] is the output impedance of the amp.

[V load] is the load voltage delivered to the headphone.

[Z load] is the impedance of the headphone.

 

Then:

 

[V load] = [V out] * [Z load] / ([Z load] + [Z out])

 

Obviously, this means the higher [Z load] is, the more [V load] will be closer to [V out], thus less voltage drop, or less attenuation at this peak.

 

However, going by this equation:

 

[Power] = ([V load])^2 / [Z load]

 

It looks to me like the higher [Z load] is, the more [Power] will drop.

 

Let's put some numbers in.

 

Assuming [Z out] = 50 Ohm, [Z load] = 250 Ohm, and [V out] is 2.7v, then:

 

[V load] = (2.7v) * (250 Ohm) / (250 Ohm + 50 Ohm) = 2.25v

 

[Power] = (2.25v)^2 / 250 = 0.02025W = 20.25mW

 

Now, assuming [Z out] = 50 Ohm, [Z load] = 350 Ohm, and [V out] is the same 2.7v, basically just [Z load] increased, then:

 

[V load] = (2.7v) * (350 Ohm) / (350 Ohm + 50 Ohm) = 2.3625v (obviously less voltage drop)

 

[Power] = (2.3625v)^2 / 350 = 0.015946875 = 15.95mW

 

Oops... so power output actually suffered. By almost 5mW.

 

Let's drop [Z out] to an ideal 0 Ohm now, so:

 

[V load] = (2.7v) * (250 Ohm) / (250 Ohm + 0 Ohm) = 2.7v (no drop in voltage)

 

[Power] = (2.7v)^2 / 250 = 0.02916W = 29.16mW

 

And do the same for 350 Ohm:

 

[V load] = (2.7v) * (350 Ohm) / (350 Ohm + 0 Ohm) = 2.7v (no drop in voltage)

 

[Power] = (2.7v)^2 / 350 = 0.02083W = 20.83mW

 

Oops... now it's a 8mW drop.

 

It looks to me like the impedance peaks actually cause power to suffer more for higher impedance headphones. So... what am I missing in this picture?


Edited by Bill-P - 10/10/12 at 2:08pm
Reply
post #55 of 81
10/10/12 at 2:38pm

As soon as the output impedance is negligible compared to the load impedance, the power will be close to inversely proportional to the latter, and thus a 2x impedance peak results in a ~3 dB "power drop" regardless of the nominal impedance. However, that is normal unless the headphone is specifically designed to be used from a high impedance source (in which case the "problem" can be worked around with a simple adapter).

Reply
post #56 of 81
10/10/12 at 2:41pm
Quote:
Originally Posted by Bill-P View Post

It looks to me like the impedance peaks actually cause power to suffer more for higher impedance headphones. So... what am I missing in this picture?

That we're talking about voltage sources, first mentioned by mikeaj in post #5.

Reply
post #57 of 81
10/10/12 at 2:46pm
Thread Starter 
Quote:
Originally Posted by stv014 View Post

As soon as the output impedance is negligible compared to the load impedance, the power will be close to inversely proportional to the latter, and thus a 2x impedance peak results in a ~3 dB "power drop" regardless of the nominal impedance. However, that is normal unless the headphone is specifically designed to be used from a high impedance source (in which case the "problem" can be worked around with a simple adapter).

 

I'm concerned about the difference in power. Because that translates into difference in SPL. At a higher output impedance, it shows that the volume difference between a "peak" and some other frequency at nominal impedance is actually lower than if the output impedance was "ideal" at 0 Ohm. But that can't be right. Unless you're trying to say lower output impedance also improves dynamic range.

 

Quote:
Originally Posted by xnor View Post

That we're talking about voltage sources, first mentioned by mikeaj in post #5.

 

And I'm trying to see how the voltage source and load impedance affects power output. Correct me if I'm wrong, but I don't think mikeaj mentioned this in post #5.

Reply
post #58 of 81
10/10/12 at 2:59pm
Quote:
Originally Posted by Bill-P View Post

 

I'm concerned about the difference in power. Because that translates into difference in SPL.

There seems to be a misunderstanding. The frequency response graphs show the dB SPL the headphone produces if you, for example, play 100 Hz, 1 kHz, 10 kHz sine waves with a fixed Vload of 1 Vrms. It does not show what happens if you vary Vload in a way to get 1 mW across the headphone drivers at each frequency. So unless you create those graphs you do not know how difference in power translates into difference in SPL.

 

That's way I mentioned the voltage source. It doesn't matter how the power changes, because as mentioned in an example before, the drivers are usually more efficient at the resonant frequency requiring less power to produce the same SPLs.


Edited by xnor - 10/10/12 at 3:19pm
Reply
post #59 of 81
10/10/12 at 4:02pm
Thread Starter 

So how do we factor in efficiency over sensitivity and power level difference?

 

I know that:

 

[Required Power] + 1mW (for the rated sensitivity) = 10^([Difference of Volume in dB] / 10)

 

And [Difference of Volume in DB] = [Intended Listening Level] - [Rated Sensitivity]


Edited by Bill-P - 10/10/12 at 4:04pm
Reply
post #60 of 81
10/11/12 at 3:56am

[dB SPL @ 1 Vrms, x Hz] - 10*log10(1000 / [Impedance @ x Hz]) = [dB SPL @ 1 mW, x Hz]

 

Plug in the values for all frequencies (x).

 

 

edit: Hz instead of kHz


Edited by xnor - 10/11/12 at 8:40am
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science

Head-Fi's Best Sellers

Head-Fi.org › Forums › Equipment Forums › Sound Science › Would lack of amp power causes degradation to sound quality for high impedance headphones?