13 hours ago, tmtomh said:

But the op-amp based Magni DAC measures better, and so I would be inclined to get the op-amp version of the Magni. Of course if I got the op-amp based Magni and didn't like it, I wouldn't try to convince myself that I did like it. I'd get rid of it - but I wouldn't buy the discrete-stage Magni instead, because the measurements suggest to me that if I don't like the Magni op-amp version, it's probably not the op-amps that are causing the sonic problem

Or, it could be the problem is the op-amps.  Even the designer probably does not know exactly why there are sonic differences.  You could take Schiit up on the offer and order both and return the one that sounds worse.  The thing is, the measurable differences are really really slight, really low level differences.  It's just not "objectively" likely that the sonic differences are attributable to these usual gross measurements.  More likely the differences are in what's not measured, and what is measured is "very very good" for both and a not a factor.

edit:  tmtomh said "DAC" but we both meant the Magni HP amplifier.

6 minutes ago, crenca said:

It's just not "objectively" likely that the sonic differences are attributable to these usual gross measurements.  More likely the differences are in what's not measured, and what is measured is "very very good" for both and a not a factor.

+100! Wow! I had to double check multiple times who wrote this. I could not believe it.

I purposefully stepped away from the thread. In that interim something was nagging me. 

It was the title and the point that it remains fundamentally unanswered:

Why did audio stop being about audio?

When products are championed in the market place that are proven empirically to have no possible impact on audio.

1 hour ago, plissken said:

I purposefully stepped away from the thread. In that interim something was nagging me. 

 

It was the title and the point that it remains fundamentally unanswered:

 

Why did audio stop being about audio?

 

When products are championed in the market place that are proven empirically to have no possible impact on audio.

 

 

I'd say this began soon after the first phonograph was invented for home use. There just wasn't an online forum for people to complain back then.

2 hours ago, crenca said:

 

Or, it could be the problem is the op-amps.  Even the designer probably does not know exactly why there are sonic differences.  You could take Schiit up on the offer and order both and return the one that sounds worse.  The thing is, the measurable differences are really really slight, really low level differences.  It's just not "objectively" likely that the sonic differences are attributable to these usual gross measurements.  More likely the differences are in what's not measured, and what is measured is "very very good" for both and a not a factor.

 

edit:  tmtomh said "DAC" but we both meant the Magni HP amplifier.

That's a good point - in particular with something as inexpensive as the Magni, the practical choice probably is indeed to get both and compare.

RE op amps, I'd be curious to hear from @mansr and other members with significant engineering knowledge about the question of whether or not the main existing measurements can reasonably be considered to capture the performance of op amps vs a discrete stage. My understanding (which could be mistaken) is that they do.

I also worry about the argument that sonic differences are attributable to the as-yet unknown parameters that we don't measure. If you are going to put forth that argument, the burden is on you propose some type of factor or aspect that would need to be measured, which currently isn't. Absent that, we're left with open-ended "it could be anything and I have no idea what it could be but I know it's something and if you disagree you are just close-minded" arguments - like the arguments people make elsewhere on these forums that the conditions under which a digital music file was copied and transmitted over the internet creates a sonic signature that is not contained in the 1s and 0s but still somehow "travels" with and is permanently encoded in the file; or that different USB cables that produce identical outputs from a DAC nevertheless have some as-yet-unknown difference that enables better or worse bass control and soundstage to emerge from the DAC in defiance of all measurements that say otherwise.

Finally, I also think the burden is on you to establish some basis for your claim that existing measurements are "gross" aka too rough or rudimentary to capture sonic differences. That is not my understanding. Differences in frequency response, phase, noise, distortion, and so on are clearly audible so long as the measurements show differences within the experimentally determined audible range of humans. And in that regard I think it's worth noting that while we are remarkably insensitive to sonic differences below certain thresholds, we are conversely very sensitive to many of those same sonic differences once they are above our threshold. For example, most people probably can't hear the difference between a -90dB SINAD DAC and a -110dB SINAD DAC, especially in real-world listening conditions. But by the same token, studies have shown that humans can consistently identify a sonic difference between two musical signals when the volume difference between them is as small as 0.2dB (not directly - they can't necessarily pick out which one is louder or even figure out that the volume is the factor that's been changed; but apparently that tiny volume change does enable the perception that something is different between the two).

11 minutes ago, tmtomh said:

An IC can be designed to tighter tolerances than a discrete-component stage (that much I know; others can elaborate beyond my level of understanding).

Right, and you can fit a silly number of components in a small area, enabling much more advanced circuits. The smaller size also makes the IC less sensitive to radiated noise.

18 minutes ago, Jud said:

More information would be interesting and helpful. What voltage and current levels would be reasonable cutoffs?

It's hard to pick a particular number. One factor is thermal management. I guess you've seen power amp output transistors bolted to a heatsink. Not only is it easier to get rid of heat from a discrete component, keeping that heat away from other components is also a good thing. Another aspect is component size. At high currents (hundreds of amps), a diode might need to be an inch in diameter. No sense in trying to integrate that.

1 hour ago, tmtomh said:

I also worry about the argument that sonic differences are attributable to the as-yet unknown parameters that we don't measure. If you are going to put forth that argument, the burden is on you propose some type of factor or aspect that would need to be measured, which currently isn't.

But this also rises the issue of a lack of skills, interest, time, etc... on the part of the person making the argument. He may have no clue what to even propose. I'd say this is 99.9999% of people making the claim (including myself).

I certainly hear what you're saying, but the burden shift is never going to lead anywhere. 

On 12/9/2019 at 10:20 AM, firedog said:

If you want me (us) to spend big bucks on a device that costs 2X, 5X or 50X, the price of a "conventional" equivalent, shouldn't you have an ethical obligation to base those claims on more than sighted and - by definition biased -listening?

 

If I understand you correctly, I find that sometimes exactly the opposite happens.  I went to a audio dealer this summer to listen to some speakers from a respected brand thinking I would like them.  I didn't.  They sounded somewhat dull and congealed.  Then the dealer played the same songs I requested with a different set of speakers, and I thought I would not like them.  I did.  They sounded dynamic and balanced and had good tone and a fantastic soundstage.

We don't always have expectation bias when we listen without double blind testing.  Sometimes we are surprised by what we hear.  And when we have a chance to compare, we wind up getting what what we want or close to it.

I have experience using the scientific method with studies, and so I understand the argument for using it.  If I submit a study to an academic journal, I will have used the scientific method with coders, rigor, etc. to keep my opinions out of it as much as possible.  When I listen to music, however, I want to feel something.  Emotion.  Goosebumps.  Awe.  Acoustic instruments and vocals played over equipment that can convey good tone, decay and attack make me feel compassion and appreciation for the artistry and the audio equipment.

I'm listening to Sondra Sun-Odeon right now, and she is carrying me away.  That can't be measured (or can it?).  I think it would be cool if someone would invent a device that could measure how we are affected by music and the equipment we use to listen to it.

1 hour ago, thyname said:

‘That’s a great point, I mean all this discussion you are having. A breath of fresh air from the general radical objectivists opinion that “all amps sound the same as long as they have enough power and current for a certain speaker”

A more ... objective ... characterization might be that all amps (say of a given class, power rating, and gain) should sound identical, because none should introduce coloration when they amplify the signal, and the degree to which they don't sound identical is easily measured and quantified objectively, usually in terms of harmonic and/or intermodular distortion.