O2 vs TOTL

[chrislangley4253]

I suppose I've never understood how there could be big sonic differences between SS headphone amps anyway. There are bad ones, for sure. My E9 is a POS with a high noise floor, but I often cringe when I hear talk about how amps expand sound stage or do other things to sound that just aren't plausible.

I, uh. Didn't claim big sonic differences. In fact, I can tell you how turn test went. It was dbt and it was the O2 off my mobo vs the m stage off a gamma 2 dac. I couldn't pick the differences using my Grado 225i my friend couldn't pick the difference using his hd595. Say we are terrible ears or w/e but the difference was really minimal.

Worth noting that later on after more listening I trained my ears to be able to tell my mobo from the gamma 2. Utilizing the switch on the matrix. I could consistently pick the difference, blind.

 

[zzffnn]

Quote:

I avoid commenting on this because there's a chance I might just have no idea what I'm talking, but the idea of an amp changing soundstage kind of puzzled me. Soundstage is in itself a perceived effect. That would mean altering the frequency response to include more mid-to-low frequencies from the left channel on the right, and vice-versa, basically a hardware crossfeed. This does not happen by mistake, if an amp would change the channel balance and crossfeed it's stunning how it actually was able to do it exactly in a way that would be interpreted by the human brain as a spacial perception effect. And yet I've read countless report of amps improving soundstage. This would be a good example of how subjective evaluations are flawed, since the chances of an interference being this specific are AFAIK minimal. Of course I might be wrong, so if someone knows how this is possible I'm open to being wrong.

@ LizardKing1,
I told you, over at the other O2 thread, that crossfeed REDUCES soundstage by blending right and left channels. Please do some further research on this. Also I do not think an amp has to change frequency response to modulate soundstage.
 
What twodeko said makes more sense and is the exact opposite to your channel blending crossfeed. Minimal crosstalk of an amp indicates better soundstage.
 
 
Quote:

 
Quote:

 
So amps that 'improve' soundstage change this timing/phase somehow?

I always understood that the perceived soundstage is better (wider/more defined) with amplifiers that have minimal crosstalk, since low crosstalk prevents coupling of the two input channels.  This would explain why some amplifiers may have a 'better' soundstage as it all boils down to the amplifier design, PCB layout, and somewhat on the quality of components used.

 
 

 

[scootsit]

Regarding the "better" or "worse" statement. The points folks made in opposition are well taken. But, that accepts that given the definition of "better" in specific measurements or contexts that we understand that they in no way reflect on enjoyment. Which is to say, I think those particular words engender the contentious and somewhat vitriolic nature of discourse. Were everyone to accept that their favorite may not be favored by everyone, and that there is "better" equipment in the world, though it may or not be more enjoyable, it would seem that much of the really combative threads could die down a bit.
 
Regarding the so-called "objectivists". I suppose by most standards, I would be among them. I would call myself an empiricist, but that's really just splitting hairs. As an empiricist, data and facts drive my decision making process. These facts reflect upon my judgement, and no one else's. I really feel zero compulsion to convert anyone to my way of thinking. That said, I do have some equipment that may very well perform no better than other equipment. I chose some of it because it looked cool. Others, I built because I thought the build sounded like fun (and it was!).
 
I guess performance and what-not are the topic of this forum. Criticizing one's decision making process is not really the topic of this forum, and I guess that's what I was attempting to stab at last night. It's okay to buy a $1,000 amplifier, or even a $10,000 amplifier. It's okay to be perfectly content with a $10 pair of headphones. Much of my rambling last night pertained to the scientific method, which is somewhat irrelevant, and I apologize for that. I suppose that the real point is that your decision to spend whatever you did is your business and there is absolutely no reason for anyone to question that.

 

[maverickronin]

Quote:

I told you, over at the other O2 thread, that crossfeed REDUCES soundstage by blending right and left channels. Please do some further research on this. Also I do not think an amp has to change frequency response to modulate soundstage.

Crossfeed actually increases the depth of the soundstage at the expense of its width.  Since pretty much all headphones have too much width, not enough depth, and 99% of music is mixed for speakers which introduce tons of  frequency dependent crosstalk between the channels I think its rather obvious that crossfeed is a good idea.
 
Of course because the level of crossfeed has to be frequency dependent and time delayed appropriately so some random level of crosstalk introduced from a bad layout or something has a very minimal chance of actually achieving this.  More likely IMO is that higher levels of second harmonics are interpreted by the brain as a slight reverb and an increase in soundstage, though that's just a hypothesis at the moment.

 

[LizardKing1]

 
Quote:

@ LizardKing1,
I told you, over at the other O2 thread, that crossfeed REDUCES soundstage by blending right and left channels. Please do some further research on this. Also I do not think an amp has to change frequency response to modulate soundstage.
 
What twodeko said makes more sense and is the exact opposite to your channel blending crossfeed. Minimal crosstalk of an amp indicates better soundstage.

Sorry, I didn't notice your comment on that thread. If it was the one that got closed, maybe that was why.
What he said was that less crosstalk meant less channel blending. Basically the opposite of what I said. I still don't understand how that means less soundstage. And I always understood digital crossfeed added a bit of the lower frequencies from one channel to the other to give the impression of soundstage, which corresponds to what I said. I think Jude mentioned it on his SPL Phonitor video. Of course it's not very relevant since it's not actual soundstage, instead it's a simulation, but it's the only way I can think of influencing spacial perception with an amp. I'm not saying I know I'm right, I just haven't realized how it's the other way around. Interesting discussion to have, and since we're on Sound Science is it considered OT?

 

[zzffnn]

Quote:

Quote:

I told you, over at the other O2 thread, that crossfeed REDUCES soundstage by blending right and left channels. Please do some further research on this. Also I do not think an amp has to change frequency response to modulate soundstage.

Crossfeed actually increases the depth of the soundstage at the expense of its width.  Since pretty much all headphones have too much width, not enough depth, and 99% of music is mixed for speakers which introduce tons of  frequency dependent crosstalk between the channels I think its rather obvious that crossfeed is a good idea.
 
Of course because the level of crossfeed has to be frequency dependent and time delayed appropriately so some random level of crosstalk introduced from a bad layout or something has a very minimal chance of actually achieving this.  More likely IMO is that higher levels of second harmonics are interpreted by the brain as a slight reverb and an increase in soundstage, though that's just a hypothesis at the moment.

I somewhat agree with maverickronin here, although not completely.  I apologize for not specifically distinguishing soundstage width vs. depth.
 
Crossfeed does provide better center image and a more "natural" listening, at the expense of width. Crossfeed probably does add some depth, although I would rather go with a better amp with better design and component. Personally I only use crossfeed for very early Beatles songs that were recorded in mono; I rarely use crossfeed for music/songs that were well recorded in stereo.
 
Re "pretty much all headphones have too much width, not enough depth,..........I think its rather obvious that crossfeed is a good idea"
To me, some IEMs, such as HJE900s, provide decent depth but not enough width; reducing soundstage width further with crossfeed may not sound good in that case.
Also some full size headphones are designed to give an upfront vocal presentation (purposely less depth in vocal range, making singer sounding closer). ATH-AD2000 and Grados are examples. If headphone is designed to sound upfront in vocals, I am not sure using crossfeed will be a good idea with stereo recordings.
 
 

 

[zzffnn]

^ Also do not forget that most open headphones already provide some degree of natural crossfeed. Because your left ear can hear some notes from right headphone driver (eidt: and vice versa). Although open headphones do not provide as much natural crossfeed as speakers.
 
If you want ear speakers, there are always AKG K1000s, Song MDR-F1s, Stax Sigma, most of which will perform better than amplifier crossfeed with stereo recordings.
 
One of the few clear advantages I can see from amplifier crossfeed is with mono recordings (wherein vocal is in one channel and instruments are in the other channel, for example). Other than that, it is more of a personal preference thing, with regards to whether or not amplifier crossfeed is beneficial.
 
To summarize my view, I think it is less likely that an amplifier designer would build in crossfeed by default (crossfeed will reveal poor spec / high channel crosstalk). More likely he/she would have amp crossfeed as an optional feature. Or simply use your open-back headphones, or avoid mono recordings.

 

[mikeaj]

Quote:

Originally Posted by scootsit /img/forum/go_quote.gif
 
[snip]
 
Questioning whether a device is better or worse is meaningless without a rigorously defined definition of quality. Without a mathematically defined optimum, we cannot define a device is closer or further from optimum. That said, different devices perform differently with respect to different analyses. There is a fruitful discussion in these specific analyses, selecting the appropriate device is ultimately personal. 
 
[snip]
 
I think it would be useful to eliminate the terms "better" and "worse" from our vocabulary and replace them with "more enjoyable for me" and "less enjoyable for me."

 
Without an ideal and a definition of quality, there's no saying what's better or worse, as you say.  This is an important point to bring up, so thanks for doing that.
 
In some discussions, including here, some people have been reasonably careful in establishing the definition of quality.  We're taking the "wire with gain" as the ideal, so anything that behaves as y(t) = A x(t) is the optimum, where x(t) is the input which we constrain to be of audio frequencies (and bounded), y(t) is the output of the amplifier, and A is some scaling constant that may or may not be greater than 1.  (In practice there may also be some ultrasonic noise and so on too.)  In many applications for amplifiers in general, exact accuracy and precision is not that important, but for many others, a "wire with gain" is exactly what you want.  Some applications quite arguably require even higher accuracy than audio.  It's mathematically the simplest intuitive definition as well.  For audio playback, if you want highest fidelity in reproducing whatever signal the source is sending, that's also what you want.  This is not necessarily what people prefer to have for their audio playback systems, so not necessarily optimal in that sense, but it's hardly an arbitrary reference point.
 
With this framework, we can begin to define what "better" and "worse" mean, in the context of "wire with gain" (accurate) signal reproduction.  One thing we could examine is the error, y(t) - A x(t).  One measure of quality could be something like average mean-squared error over time, given some specific input x(t), with a certain load, at a certain output level.  Then we could rank different amplifiers in terms of best to worst.
 
However, that is a very crude measure and may not well correspond to what sounds most like the ideal.  Also, what kind of input x(t) would you use?  That's why in practice, people have mostly settled on standardized test signals x(t) to use for comparison purposes, to stress the amps in different ways to try to elicit bad behavior, and more relevant audio metrics such as the noise, THD, frequency response, and so on.  As is suggested by theory and by listening results, if an amplifier has good "scores" on a wide variety of lab test inputs, it should do well with just about any allowable input, including music.
 
Some people seem to be very convinced that standard test data has very limited meaning.  I'll grant that it's not easy to come up with some 100% psychoacoustically valid weighing function that takes in all the test results and spits out some kind of final score that corresponds to what people hear.  However, that's not particularly the best approach anyway.  It may be more instructive to say that one amp is better at X, Y, Z, and worse at J, K, L.  But if one amp scores better than another in pretty much every test, by a reasonable margin, it is probably safe to just cut out some details and conclude that one amp is "better" than another, with respect to accurate signal reproduction.
 
Of course, people could also be talking about output power levels, features, ergonomics, aesthetics, price, and more, when they are being looser with the definition of "better".  Some of those attributes are difficult to define; others are not.
 
Evaluating goodness with respect to individual preferences is obviously more difficult.  However, we need not consider personal preferences when evaluating how close amps come to the ideal.

 

[maverickronin]

Quote:

Crossfeed does provide better center image and a more "natural" listening, at the expense of width. Crossfeed probably does add some depth, although I would rather go with a better amp with better design and component. Personally I only use crossfeed for very early Beatles songs that were recorded in mono; I rarely use crossfeed for music/songs that were well recorded in stereo.
 
Re "pretty much all headphones have too much width, not enough depth,..........I think its rather obvious that crossfeed is a good idea"
To me, some IEMs, such as HJE900s, provide decent depth but not enough width; reducing soundstage width further with crossfeed may not sound good in that case.
Also some full size headphones are designed to give an upfront vocal presentation (purposely less depth in vocal range, making singer sounding closer). ATH-AD2000 and Grados are examples. If headphone is designed to sound upfront in vocals, I am not sure using crossfeed will be a good idea with stereo recordings.

"Better equipment" won't ever replicate what crossfeed does.  Most all music is mixed for speakers in a conventional “stereo triangle” set up.  That's what stereo is.  Its not just anything with two channels and its not just anything designed for two speakers.  An Ambiophonic mix is designed for two speakers but it's not stereo.  Stereo is designed to be played back over speakers in a conventional set up where the left ear receives significant acoustic crosstalk from the right channel and vice versa.  We're talking in the neighborhood of only -10dB at certain frequencies and then decreasing as the wavelength combines with the size and shape of your head to affect the wave's diffraction.  Open headphones or "earspeakers" of any description don't even come close to providing the right amount amount let alone at the right frequencies.
 
Usually I find the better that something is recoded and mixed the more I need crossfeed because better recoding and mixing make for a larger soundstage.  At least on speakers.  With headphones it just gets wider and wider instead of mostly being a blob in the middle like on a lot of modern rock and pop.  IRL or with speakers you don't normally just hear one section of the orchestra in one ear with just about none of it in the other.  I usually can't get through one movement of a symphony without crossfeed before I get a headache.
 
The same goes for open headphones.  In general, though not always, the more open a headphone is the wider its soundstage and the more unnatural and headache inducing hard pans become.  I can go a lot longer without crossfeed when using IEMs then I can with something very open and wide like my recently acquired AD1000s.
 
I'm not saying that anyone has to like crossfeed, and I may be unusually sensitive to asymmetry in a sound field and need to use crossfeed more than others, but it's simply a matter of fact that using crossfeed gets you closer to what you'd hear if you played the same music over an ideal speaker system.

 

[estreeter]

Quote:

I guess performance and what-not are the topic of this forum. Criticizing one's decision making process is not really the topic of this forum, and I guess that's what I was attempting to stab at last night. It's okay to buy a $1,000 amplifier, or even a $10,000 amplifier. It's okay to be perfectly content with a $10 pair of headphones. Much of my rambling last night pertained to the scientific method, which is somewhat irrelevant, and I apologize for that. I suppose that the real point is that your decision to spend whatever you did is your business and there is absolutely no reason for anyone to question that.

Agreed. I suspect that what gets a lot of empiricists like yourself hot under the collar (particularly the DiYers) is this :
 
Rebrand a pair of $10 headphones, complete with a marketing campaign, and try to sell them for $100 in Apple stores and most here will call BS, yet putting a few 'audiophile' parts in a nice case and charging considerably more than $100 seems to be accepted.  I've never subscribed to the spreadsheet approach to evaluating a component's worth - anyone who wants to build their own iPhone can go right ahead - but whatever value the manufacturer places on design and implementation needs to far outweigh the marketing effort. Head-Fi provides a (large) captive audience for some of the most cost-effective marketing anywhere - word of mouth. If I want to build up a buzz around my new Acme Wonderamp,great, but the least I can do is publish a meaningful set of measurements to accompany that shiny new case ......
 
(If I have a criticism of some DiYers. its that they seem to value other people's labor at roughly zero dollars per hour and thats a big part of my objection to the spreadsheet approach to evaluating a components worth - I have to wonder how they would react if their daytime employers took a similar approach to their own efforts)
 
 

 

[Magick Man]

For me, can't say the same for others, the DIY process itself is the most fun. The end product is like icing on the cake. My modded T50RPs and D7000s are pretty much the only items I've kept.

Anyway, that was neither here nor there, I just wanted to say that I'm a form and function guy, usually in fairly equal amounts. For me, and I suspect most other audiophiles, a piece of gear needs both. While the O2 may measure very well, and can drive 95% of headphones out there, that front input, its tiny size, and plain look just does nothing for me. I'll pay more for more appealing aesthetics, usually quite a bit more. But then, I recently paid $5,000 for a Lovecraftian-style sculpture, so maybe my views are skewed. I blame my Steampunk loving wife.

 

[maverickronin]

1) Is that your wife? 

 
2) You can always put the O2 in a fancy looking enclosure.

 

[sridhar3]

Quote:

 
Whoa.

 

[Anaxilus]

Measurements or it didn't happen.

 

[Magick Man]

Yeah, she's a cutey (that is a wig BTW, she's actually blonde), and she's totally altered my views on aesthetics. Things I never considered before, such as visual composition and the symmetry of an object, it all appeals to her meticulous nature. In terms of gear, I see it now in two parts; how the component sounds and how it "communicates". Beautiful things are appealing on many different levels.