Head-Fi.org › Forums › Equipment Forums › Headphone Amps (full-size) › How do/can DACs and amps change sound signature?
New Posts  All Forums:Forum Nav:

How do/can DACs and amps change sound signature? - Page 2

post #16 of 60
Hey guys it's all good. No point in being hostile towards each other.
You both have valid points.
post #17 of 60
Quote:
Originally Posted by Rawrbington View Post

Hey guys it's all good. No point in being hostile towards each other.
You both have valid points.

 

He hasn't made any points at all actually, other than that wikipedia has a "tube sound" entry. 

post #18 of 60
Quote:
Originally Posted by Eisenhower View Post

 

He hasn't made any points at all actually, other than that wikipedia has a "tube sound" entry. 


Then he won't have any problems defending his "claims", will he?

post #19 of 60
Quote:
Originally Posted by Hooster View Post


Then he won't have any problems defending his "claims", will he?

 

I was referring to a lack of points in defense of his claims.

"Tube amps are warm" is a claim, "even order harmonic distortion from tube amps is pleasing to the ear" is a point.

post #20 of 60

Eisenhower, instead of posting "you [this]" and "you [that]" how about you discuss the OP's question? smile.gif

 

A bunch of people arguing whose claim is more correct doesn't help people. Discussing the different aspects a question, with caveats about the limitations of one's knowedge (as scientists actually do) does.

post #21 of 60

To the OP:

Two things, you'll have to look at.  

The first is the electrical characteristics of the headphone and the amp. The amp is the source, driving the load (headphone) and the way their impedances interact with each other can boost/reduce some part of the spectrum in the headphones' sound.

Second is the signal itself. This is where harmonic distortion and noise comes in, and then other EQ effects.

 

Typically, the first one is an unwanted characteristic, and should be minimized. You don't want the headphone underpowered, and want to make sure it gets the power it needs at all times.

The second one is a matter of preference.


Edited by proton007 - 11/11/12 at 5:18pm
post #22 of 60
Thread Starter 

Great info, especially by mikeaj. Much appreciated. I'm going to read over these posts a few times to make sure I really understand it. Prelim idea is that nothing is perfectly efficient and the inefficiencies are what creates changes in sound. It seems to me that these inefficiencies in a DAC are detrimental but in an amp might simply alter the sound instead of degrade it.

 

At any rate, like I said, time to reread and absorb.

post #23 of 60

"Inefficient" isn't really the right word.  Efficiency would be something like "this amp converts 100% of its input power into the output signal and dissipates no heat".  That's unrelated to sound quality.

 

A picture might be more useful than words.  Scroll down to the images on http://www.amb.org/audio/gamma2/highlights.html and you can see some different examples of how a high-end DAC might choose to translate a digital signal like an impulse or square wave into an analog signal.  As you can probably imagine, the different filters here sound subtly different due to the different tradeoffs they make in terms of pre-ringing versus post-ringing, etc.  Note in particular that filter B is in some sense "more distorted" on an absolute basis, but as the accompanying text explains, may sound better due to the characteristics of the distortion.

 

More pictures are available on https://passlabs.com/articles/cascode-amp-design .  The figures here demonstrate how real-world components like transistors respond compared to the theoretical "ideal" amplifier as input voltage scales.  The basic idea is that while you'd like to have linear response, that's not what you get; but you can construct more complex circuits which do a better job of reproducing that response over a specific input range than a single component will.  Now given that circuit complexity means increased component cost (among many other tradeoffs), you might see why amplifiers at different price points perform differently.

post #24 of 60
Quote:
Originally Posted by pkasting View Post

A picture might be more useful than words.  Scroll down to the images on http://www.amb.org/audio/gamma2/highlights.html and you can see some different examples of how a high-end DAC might choose to translate a digital signal like an impulse or square wave into an analog signal.  As you can probably imagine, the different filters here sound subtly different due to the different tradeoffs they make in terms of pre-ringing versus post-ringing, etc.  Note in particular that filter B is in some sense "more distorted" on an absolute basis, but as the accompanying text explains, may sound better due to the characteristics of the distortion.

 

More pictures are available on https://passlabs.com/articles/cascode-amp-design .  The figures here demonstrate how real-world components like transistors respond compared to the theoretical "ideal" amplifier as input voltage scales.  The basic idea is that while you'd like to have linear response, that's not what you get; but you can construct more complex circuits which do a better job of reproducing that response over a specific input range than a single component will.  Now given that circuit complexity means increased component cost (among many other tradeoffs), you might see why amplifiers at different price points perform differently.

 

It is important to quantify such differences, though, and to compare them against known limits of human hearing, to find out how much likely it is that they actually matter. For example, in the case of a DAC filter, if it has a reasonably chosen response, the pre- and/or post-ringing are both very short and at a very high (to most people, ultrasonic) frequency, and the group delay of a minimum phase filter is also not high enough to be audible (compare this graph - the vertical scale is in milliseconds - to this table).

I posted a blind test on the Sound Science forum a while ago, where music was passed through a high quality software sample rate converter (44.1 kHz to 96 kHz/24 bit with a linear phase filter) in one case, and the other file was a loopback recording (96 kHz/24 bit) of the same music through a cheap sound card (the signal went through a Cirrus Logic D/A and A/D converter using minimum phase lowpass filters, several cheap - mostly NJM5532 - op amps, a "no name" electrolytic coupling capacitor, and cheap cables and connectors). Most audiophiles would probably expect the loopback recording to sound horrible, but in fact no one was able to tell the files - which were accurately level matched and synchronized - apart with an ABX comparator.

post #25 of 60

Here are some interesting graphs about how a DAC can change sound. If you look closely on som of the graphs you can se that som of the DACs actually recreates a 11,025hz signal as a aprox. 11,1k signal, thus pitch the original signal aprox 100hz . (Look at the arcam and HRT dac)

 

(the thes is originaly in Norwegian) 

http://translate.google.no/translate?sl=auto&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&u=http%3A%2F%2Fwww.lydogbilde.no%2Ftest%2Fhandfaste-bevis&act=url

 

Arcam dac 11025hz signal

 

 

Electrocompaniet 11025hz signal:

 

 

 

And look at the recreation of a sinusiodal signal:

Nad m51:

 


Wavelenght Proton:

 

I would assume the wavelanght color the sound quite a bit, making it sound more stressed. 


Edited by paara - 11/14/12 at 10:49am
post #26 of 60
Quote:

Originally Posted by paara View Post

 

And look at the reacreation of a sinusiodal signal:


Wavelenght Proton:

 

Is the vertical scale in microvolts ? I would guess so, which means the signal level is only about -90 dBFS, so it basically just shows that the DAC is quite noisy (roughly comparable to average PC motherboard audio).

 

Quote:
Originally Posted by paara View Post

Here are some interesting graphsa about how a DAC can shange sound. If you look closely on som of the graphs youcan se that som of the DAS actually recreates a 11,025hz signal as a aprox. 11k signal thus shifting the actual freaquency of the original signal. (Look at the arcam and HRT dac)

 

That is rather poor, but some "audiophile" DACs are indeed objectively worse in some aspects than onboard audio. Most of these graphs look worse than my Xonar D1 recorded with another sound card, and many of the devices are probably not even cheap.


Edited by stv014 - 11/14/12 at 10:53am
post #27 of 60
Quote:
Originally Posted by stv014 View Post

 

Is the vertical scale in microvolts ? I would guess so, which means the signal level is only about -90 dBFS, so it basically just shows that the DAC is rather noisy.

 

 

That is rather poor, but some "audiophile" DACs are indeed objectively worse in some aspects than onboard audio. Most of these graphs look worse than my Xonar D1 recorded with another sound card, and many of the devices are probably not even cheap.

 

If the DAC is noisy, wouldn't that change the sound? Making it sound worse? (just asking,because I am not sure how it would affect sound)

 

Does that mean that the sound would be fine if you do not "dampen/lower" the input signal?


Edited by paara - 11/14/12 at 10:54am
post #28 of 60
Quote:
Originally Posted by paara View Post

If the DAC is noisy, wouldn't that change the sound? Making it sound worse?

 

It means that there is likely some (very quiet) background hiss, but is probably not audible while playing actual music that does not have a very wide dynamic range, at least assuming that the level of the noise does not increase much with the signal (sometimes it does).

post #29 of 60
Quote:
Originally Posted by stv014 View Post

 

It means that there is likely some (very quiet) background hiss, but is probably not audible while playing actual music that does not have a very wide dynamic range, at least assuming that the level of the noise does not increase much with the signal (sometimes it does).

Aaah. I guess my assumption was wrong then. Thank you for correcting me :) Couldn't figure out why in the world anyone would buy the Wavelenght ;)

post #30 of 60

Quote:
Originally Posted by stv014 View Post

 

It is important to quantify such differences, though, and to compare them against known limits of human hearing, to find out how much likely it is that they actually matter.

 

Absolutely.  One of my favorite articles about this sort of thing is http://people.xiph.org/~xiphmont/demo/neil-young.html , which demolishes the argument that distributing 24-bit 192 kHz files for end-user listening leads to a sound quality improvement.

 

Just because one graph looks different than another does not necessarily mean they will sound perceptibly different.  I was mostly trying to give concrete examples how differences in amps and dacs can, concretely, affect the output.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Headphone Amps (full-size)
Head-Fi.org › Forums › Equipment Forums › Headphone Amps (full-size) › How do/can DACs and amps change sound signature?