Motherboard OPAMP upgrade kit

[Mr Pig]

I have a little Beresford DAC here and I've changed the op-amps in it. The different op-amps do sound different! I'm sorry, but they do. I still prefer the sound of my Micromega CD player but changing the op-amps in the DAC does, or at least it can, change the sound.

You might think I'm a snake-oil drinker but I'm not really. A few weeks ago a friend of mine brought out his new DAC to try in my system. He bought the DAC for £600 but the new price was double that so it was about a thousand-pounds more than the Beresford cost. I was very keen to hear it. I've heard a lot of high-end CD players but not an expensive DAC so I was looking forward to being impressed.

We plugged his DAC in to warm up and played about eight tracks on the Beresford. Then we switched to his new DAC and played the same tracks again. I say the same tracks but actually, I gave up after about four. It sounded pretty much exactly the same! Maybe slightly slower, less toe-tapping, but really nothing you could be sure about. I told him flat out that I thought they sounded the same but he didn't concede it. He reckons it works better in his system, but as he's never tried the Beresford in his system I have no idea how he could possibly know that.

Differences in digital audio do exist but in my opinion they are generally as lot smaller than in analogue and are very often not price related at all.

[SeanB]

On a motherboard playing a lossy encoded format, with a best guess of 10% tossed because of it being below the preset thresholds, does the choice of opamp matter? Possibly with a CD player with 16 bit digital audio and low jitter and a good DAC yes, on a PC not even going to be noticed.

[Phaedrus]

I have a little Beresford DAC here and I've changed the op-amps in it. The different op-amps do sound different! I'm sorry, but they do. I still prefer the sound of my Micromega CD player but changing the op-amps in the DAC does, or at least it can, change the sound.

You might think I'm a snake-oil drinker but I'm not really. A few weeks ago a friend of mine brought out his new DAC to try in my system. He bought the DAC for £600 but the new price was double that so it was about a thousand-pounds more than the Beresford cost. I was very keen to hear it. I've heard a lot of high-end CD players but not an expensive DAC so I was looking forward to being impressed.

We plugged his DAC in to warm up and played about eight tracks on the Beresford. Then we switched to his new DAC and played the same tracks again. I say the same tracks but actually, I gave up after about four. It sounded pretty much exactly the same! Maybe slightly slower, less toe-tapping, but really nothing you could be sure about. I told him flat out that I thought they sounded the same but he didn't concede it. He reckons it works better in his system, but as he's never tried the Beresford in his system I have no idea how he could possibly know that.

Differences in digital audio do exist but in my opinion they are generally as lot smaller than in analogue and are very often not price related at all.

If you use the term "toe tapping" in a comparison between two audio systems, you might be a snake oil drinking audiophile.

[ciccio]

Fundamentally there reaches a point where all of the audio specifications become numbers, the point is argued amongst enthusiasts with some claiming to hear well past 0.00001% distortion and microvolt noise levels. Most good audio op-amps are so good that you simply will not hear any difference.

Well, some enthusiasts an some experts, as my grandfather used to say, speak for the only reason that  there is a tongue in their mouth.
My experience  is that, in a well designed circuit, most op-amps (I refer to the ones suitable for the application, some very good op-amps like OP-07 are not) sound the same (noise performance is another story).
The BIG difference is when you use them near their limits (like output swing approaching supply voltages),  with non compressed input (live music, speech reinforcement). Here you HEAR the difference: different op-amps saturate in a different fashion, some with only an increase of distortion, some with more negative effects (try to hard clip a TL072  in a  non inverting buffer, or a NJM4580 in a high pass filter: you will get a frequency doubler).
To hear subtle differences, you must have a very low noise level, and most of the times you don not have this..
Best regards

[Mr Pig]

If you use the term "toe tapping" in a comparison between two audio systems, you might be a snake oil drinking audiophile.

You don't think that some audio systems are more involving than others?

[alm]

On a motherboard playing a lossy encoded format, with a best guess of 10% tossed because of it being below the preset thresholds, does the choice of opamp matter? Possibly with a CD player with 16 bit digital audio and low jitter and a good DAC yes, on a PC not even going to be noticed.

Are you aware of any double blind study comparing uncompressed CD audio with decent quality MP3 files that showed a difference between these two? Because sites like HydrogenAudio have been doing it for years, and above a certain bitrate (which might be as low as 128 kbit/s for some codecs / types of music) people become unable to distinguish them (they don't perform better than picking at random). Even musicians, 'experiened listeners' and the few audiofools that take the test.

[Phaedrus]

If you use the term "toe tapping" in a comparison between two audio systems, you might be a snake oil drinking audiophile.

You don't think that some audio systems are more involving than others?

I think that how "involved" you are in a given piece of music depends 40% on how you feel at that moment, 29% on what you had for breakfast, 19% the price of ham in Belgium, 11% the arrangement of the planets, and 1% the choice of op-amps used in the audio system.

If it was less toe-tapping the second time through, it's more likely because you just listened to the song already, rather than the audio system under test.

I'm just reminded of an audiophile review I read a while back, where someone complemented a $50 picture of a frog because it made her audio system "pop" and that it made everything more toe tapping. I mean, seriously? 

[Mr Pig]

Are you aware of any double blind study comparing uncompressed CD audio with decent quality MP3 files that showed a difference between these two?

How sure are you that double-blind listening tests work?

[rexxar]

Are you aware of any double blind study comparing uncompressed CD audio with decent quality MP3 files that showed a difference between these two?

How sure are you that double-blind listening tests work?

What do you mean? How could it not work? 

[Mr Pig]

I think that how "involved" you are in a given piece of music depends 40% on how you feel at that moment, 29% on what you had for breakfast, 19% the price of ham in Belgium, 11% the arrangement of the planets, and 1% the choice of op-amps used in the audio system.

Ah, you said systems. I agree that the differences between op-amps is probably very small, but in a revealing enough system you can hear it. It's the same with CD players.

I would go as far as to say that a lot of amplifiers sound quite similar too with major sonic differences often being down to how well they drive your chosen speakers. But they're not all the same.

But as far as systems are concerned then yes, some are 'toe-tapping' and some are boring. Big time.

If it was less toe-tapping the second time through, it's more likely because you just listened to the song already, rather than the audio system under test.

Well, I've been into Hi-Fi for thirty-years and I'm pretty honest and critical about what I hear. I've sold, or not bought, quite a few expensive items because similar performance could be had for much less money, I'm not a Hi-Fi snob who deludes himself. If I was I wouldn't have thought the really expensive DAC sounded the same as the cheap one.

I'm just reminded of an audiophile review I read a while back, where someone complemented a $50 picture of a frog because it made her audio system "pop" and that it made everything more toe tapping.

Yeah, sounds like stupidity to me too but I don't see what it has to do with actual sound reproduction? Just because one guy feels better after a sugar pill does not mean that drugs which actually word do not exist.

[Mr Pig]

What do you mean? How could it not work? 

Ok. Why is it claimed that sighted listening tests are flawed?

[dexters_lab]

I was going to post that... it's real:
http://www.neoseeker.com/Articles/Hardware/Reviews/aopenax4btube/

I remember reading reviews of that when it was released, i was too much of a abit junkie to try it

[Phaedrus]

What do you mean? How could it not work? 

Ok. Why is it claimed that sighted listening tests are flawed?

Because of fairly obvious and thoroughly demonstrated biases. If you have an affinity for Brand X, and are asked to compare Brand X to Brand Y, you will choose Brand X almost every time. It's psychological. You can't help it. Even if the one you are told is Brand X, is actually Brand Y, you will still choose what you thought was Brand x.


Double blind testing is the gold standard for psychological testing, where neither the tester nor the testee are aware of which product/sample/whatever is which. That way all personal preference and bias is removed, except for the variable under test. That's the basis of a scientific experiment: removing all variables except the ones you want to test. And when it comes to studies that could involve human preference, such as psychological, sociological, medical, and product studies, the double blind test is the gold standard, as it is the best as narrowing down variables in those situations.

[tom66]

Double blind testing is very difficult to do with audio. Can you compare something you heard 10 seconds ago to something just now? Sure, maybe you can compare the odd thing but you'll neglect the rest.

The true test is to use equipment to measure the difference: double blind tests in this instance are about as useful as getting people to compare the amplifier performance by how much it weighs or what it smells like.  In the same way as double blind tests on wine often show that cheaper wine tastes both better and worse than more expensive wine, depending on who tastes it.

But, when you present this idea to audiophiles, the result is anguish -- no, your oscilloscope cannot be used to check the output for distortion, because the human ear is more sensitive than high end test gear... bollocks. A good instrument could measure the difference.

[Phaedrus]

The way to do double blind testing with audio gear is not to have the same people listen to the same thing twice on different setups; but to have two sufficiently large sample groups listen to one thing on one setup and rate it. Given large enough sample sizes, this will give a good indication of how "good" a particular sound system sounds.

[tom66]

The way to do double blind testing with audio gear is not to have the same people listen to the same thing twice on different setups; but to have two sufficiently large sample groups listen to one thing on one setup and rate it. Given large enough sample sizes, this will give a good indication of how "good" a particular sound system sounds.

That doesn't really work though, because it's so hard to tell the difference between well-matched systems, most people forget the difference and will put a random choice down. Audio memory of the brain is poor, it works on details (speech, general sound) but can't remember slight nuances in sound quality.

Perhaps one way to do a test would be to play a track and then slowly mix between two amplifiers set up to equal amplitude. If the user can judge when one amplifier is playing over another, then there is a difference, but it won't tell you if one is better than the other.

[alm]

A-B-X testing is common in audio. First you hear sample A, then sample B, and then X (which may be either A or B). Then the subject is asked whether sample X was sample A or sample B. Repeat this for multiple samples. This allows you to test whether people can distinguish A and B without cognitive biases playing a role or having to declare a best recording. I guess ideally you would also randomize the order of A and B to prevent people's memory from introducing a bias.

[krivx]

Effects circuits are a whole different ball game. You're actually trying to distort the signal somewhat, so a "crappy" device is actually quite useful there, it adds the color of its crap to the signal. Germanium diodes and transistors, vacuum tubes, 200-year-old op amps, they all have their uses in analog effects. I've even gone so far as to experiment with intentionally abusing parts (like running a signal transistor above absolute max power for an hour or two, carefully controlled to keep the smoke in).

But for high fidelity audio... there are precision op amps from the likes of Linear and AD that have such good noise performance and low offset voltage that they really ought to be performing as pretty much ideal op amps in an audio situation. Surely you can't tell those apart?

Nah.  Unless you're intentionally trying to overdrive the op-amp, and most don't, the op-amp is running in it's linear range.  Hearing the differences really depends on the circuit it's sitting in.  Sometimes it doesn't really matter much, and sometimes there are very noticeable differences.  They all have different curves for things like output impedance vs frequency, for example, and that's either important or not depending what you're driving.

Running op-amps outside their linear range is one of the most common applications in effects circuits. Op-amps with non-linear components (often diodes) in their feedback loops, over-driven inputs and deliberate slewing distortion can all be useful. You can definitely hear a difference in some cases. You can see it on a scope too.

[John Coloccia]

Running op-amps outside their linear range is one of the most common applications in effects circuits. Op-amps with non-linear components (often diodes) in their feedback loops, over-driven inputs and deliberate slewing distortion can all be useful. You can definitely hear a difference in some cases. You can see it on a scope too.

When you put clipping diodes in the feedback loop, that doesn't usually drive the op-amp out of it's linear range.  A typical op-amp clipping gives you hard clipping and generally sounds like crap though some effects do rely on opamp clipping.  If that's what you want, you would generally run diodes after the pre-amp opamp and then you truly get hard, diode clipping because the output is clamped to the forward voltage of the diodes.  The whole point of running diodes in the feedback loop is to get soft clipping and avoid hard clipping.

That said, the typical feedback clipping circuit (i.e. TubeScreamer style overdrive) has an RC circuit in the feedback loop along with clipping diodes, an RC circuit in place of the traditional gain resistor to Vref, and an RC circuit after the opamp for a treble bleed (i.e. the "tone" control).  Any change in opamp impedance will change the response of these RC circuits in a rather complicated and interactive way, even though the opamp is still operating well within it's linear range.

Anyhow, I don't really mean to argue about this.  I'm just trying to explain a bit better how they work and why different opamps can make the entire circuit sound different even though the opamp itself may essentially be operating at ridiculously low frequency and practically zero distortion of any kind.  You have to consider the entire circuit, and that's the point.

[madshaman]

I've noticed some comments about opamps sounding different when overdriven, well, that simply means you need to fire your sound engineer.

Seriously, if you're going to overdrive your amplification, there is only *one* simple choice, use freaking vacuum tubes, vacuum tubes have the nicest response when driven to saturation (make sure to bias far away from cutoff), you'll emphasise the even harmonics and get a nice "warm" sound.

Vacuum tubes also have much lower gain than an opamp so vacuum tube amps can't employ much negative feedback so they will be even *less* linear, yay!  So you'll get additional harmonic distortion; double yay!

So, throw out your opamps and get to work on those tube circuits.  And no, I don't hate vacuum tubes in audio amps, I *love* them, for the only thing they're good for: guitar and bass amps.

Maybe it's my mood this morning, but this thread is driving me ****ing nuts.

If the circuit they're seated in is properly designed, the amp is being fed with proper levels, and the opamp has specs which exceed human perception, it make no ****ing difference at ALL what opamp is in there, period; none whatsoever.

**** (profanity omitted)
**** (profanity omitted)
**** (profanity omitted)

edit: fixed spelling/grammar, added more profanity and anger

[madshaman]

I can't help it, I just can't...

If anyone is interested, I'm selling toasters which use your choice of tungsten or molybdenum heating elements.

The black body radiation emitted by these special rare elements contains special nuancing photons which flavour the toast in their own special way.

Tungsten is most appropriate for rye bread, the tungsten photons add a subtle nutty flavour and do not cause premature infant death syndrome.

Molybdenum is best used on multigrain bread, the molybdenum photons protect the health benefits of multigrain and the photons themselves contain no MSG.

[johansen]

I can't help it, I just can't...

If anyone is interested, I'm selling toasters which use your choice of tungsten or molybdenum heating elements.

The black body radiation emitted by these special rare elements contains special nuancing photons which flavour the toast in their own special way.

Tungsten is most appropriate for rye bread, the tungsten photons add a subtle nutty flavour and do not cause premature infant death syndrome.

Molybdenum is best used on multigrain bread, the molybdenum photons protect the health benefits of multigrain and the photons themselves contain no MSG.

i feel you brother.

furthermore: it is not hard to detect the difference between two different audio unknowns by changing only one channel of the unknown. but its not quantitative. most of the FFT plots and noise amplitude plots don't include phase, which is extremely important to the human ear. certain phase shifts are actually quite painful.
but in any case..

most of the usual adjectives will be met with "wtf you talking about bro?"

[GK]

furthermore: it is not hard to detect the difference between two different audio unknowns by changing only one channel of the unknown. but its not quantitative. most of the FFT plots and noise amplitude plots don't include phase, which is extremely important to the human ear. certain phase shifts are actually quite painful.
but in any case..

LOL. I just love the way people continuously invent and pull facts from their arses in these audio threads. It is a psychoacoustic fact that our hearing is particularly insensitive to phase. This fact has been established for about 2 million years already. Just one reference, immediately at hand:

http://www.douglas-self.com/ampins/wwarchive/wwarchive.htm#phase

 

[madshaman]

I have many of Douglas Self's books, very excellent, very practical, very intolerant of audiofoolery.

He goes into detail of every aspect of psychoacoustics and does devote time to talking about stereo image and the human perception of phase.

Human's aren't insensitive to phase, it's how we determine left-right direction, and is an integral part of hearing.

What Self does point out is that the phase offset caused by filtering in audio equipment is largely undetectable by the human ear and he does describe at what point it becomes a problem.

As for stereo image, subtle differences in the first pole of explicit or implicit filters (which is tied to the phase response) in each channel generally won't be enough to impact the perception of the stereo image.

For those not using headphones, the room you're listening to your music in and your speaker enclosures are probably adding more phase shift and wacky filtering than your amp ever will.

[tom66]

This is also interesting
http://ethanwiner.com/believe.html