Op-amp Rolling Using Sparkos on Fosi V3 Mono

[xanalog]

Not to mention it is a flawed scenario. Amir isn't just measuring one parameter (in this scenario, he wouldn't just be measuring calorie count to determine taste differences)

Valid point; especially since @amirm measures a wealth of standardized test routines to characterize audio I/O performance in a quantitative manner.
He also supplements his analysis of the measured results with [his certified 'golden-ear'] listening tests, for that subjectivity thingy!
But:

....The sound engineer has recorded both sets and the graphs appear to be identical.
8 out 10 people listening to both sets live couldn't hear a difference. The remaining two could hear a difference. (They preferred the newer violin to the Strat)

My question is about:
What measurements can be done to differentiate/show the minute tonality [?] differences between the two musical instruments and/or two reproduction hardware; Not the subjective/perceived notions but the measurable differences, no matter how subtle such delta may be.
Would that be a more fair question?

I still have not fully digested that there are zero (like -273.15°C ) performance differences while op-amp rolling.

 

[pma]

The differences are non-zero. Feed the same music signal into two identical link level channels with 2 different op-amps, digitize the outputs and compare them in AA or use Deltawave. You will ALWAYS find a difference signal, though very low in level. The speculations will be on its audibility in the presence of the main music signal.

 

[im_gumby]

I'm afraid if you have to ask that then there's no hope.

Apparently not.

And in this case, its not me, its you.

I asked a serious question of @Buckeye Amps ... why he chose the op-amps he did instead of going with the cheaper option if after all there are no sonic differences.
In a different discussion I asked Amir why Bruno P would tell a YTer interviewing him and building an amp off his boards to switch out the op amp if there's no sonic difference. (Note: This is a surface mounted chip, which is much harder to switch out. ) Amir gave a nonsensical answer that Bruno has to keep his foot in both camps. I had no response from BuckeyeAmps.

 

[antcollinet]

and by your logic they should taste the same.

Not by my logic - nor anyone else here.

Calories have nothing to do with taste. To measure taste, you'd have to precisely measure the complete chemical constituents of the candy, plus every characterstic of the texture. Bear in mind also you are talking about the taste of the candy here (the equivalent of the music for audio gear) Not the jar the candy is contained in to transfer it from the factory to our mouths. (the equivalent of the gear). I'm sure you'd admit, measuring the transportation devices to see how it might alter the flavour is much easier than measuring the candy itself to determine how it tastes to start with.

Similarly the measurements Amir makes tell us EVERYTHING about how a device ALTERS the music.

And again you missed the point.
The first time I attempted to discuss this I got banned...

So why are you carrying on with the same utter nonsense - if you know what the inevitable result will be.

 

[solderdude]

Amir gave a nonsensical answer that Bruno has to keep his foot in both camps.

This is NOT nonsensical and in fact the only correct answer.
Of course... Bruno will not admit that but am sure he wants to sell not only to a certain 'group' of people but rather to subjectivists, objectivists and inbetweenivists so it makes sense to please as much potential buyers.

Sometimes choosing a certain op-amp in any design also has a non-technical reason.
For instance... when people 'believe' that a certain op-amp has 'better' tonal character (which none can have by definition) as the performance is largely determined by the components around the op-amp. It would be wise for a manufacturer to either offer an option or choose the 'well reviewed' one or the one with marginally 'better' specs (distortion number wise and/or noise.
A reason could also be availability/price.

Similarly the measurements Amir makes tell us EVERYTHING about how a device ALTERS the music.

Well... technically Amir's measurements tells us a LOT about the signal fidelity of a device under specific circumstances.
The thing is... for technical people that know how to interpret the various measurements it sure tells us more than enough.

The measurements do not say everything nor about altering of music.
Music is not altered anyway, distortion will be ever so slightly different.

The differences are non-zero. The speculations will be on its audibility in the presence of the main music signal.

^^^ this ^^^
There will be measured differences, even between measurement runs and even between channels.

People want to see explanations/verification for what they 'perceive' and when there are minute differences even below any audible thresholds they will assume that such small differences thus are audible as measurements show there is a minute difference.
Add to that such individuals always have superior ears and gears than those silly guys looking at wave forms and plots and never listen to/enjoy music which also confirms the line of thinking that 'everything matters' and 'can be heard'.

In the case of the reviewed amp both op-amps are both lower in distortion and wider in frequency response than the other component (TPA3255) that actually determines the 'behavior' of the amplifier (at a certain test level/load).

 

[Heinrich]

Yet again more in the series on op-amp rolling.

Here's an analogy ...

You rephrase a notion that should be widely accepted. That there is more (or in general: different) into hearing than amplitude spectra.

As an example we have MP3, that investigated the human hearing in regard to what is perceivable and what not. Is the missing part in an MP3 coded piece considered to be distortion? Anyway, it was investigated a lot, and some people were more susceptible for those objectively undeniable differences. All of them had specific hearing losses, as far as I remember. Because MP3 relies on masking a lot, the outcome didn't appear unexpected, in hindsight.

The violinist you mentioned, o/k, timing in playing and all such subtleties, too easy.

MP3 was easy. Now op-amp-rolling. What we have is amplitude over time, there are no other features in such an analog device (MP3 is not analog to a very high degree), that's why it is called "analog". We might define a difference between tested transmission lines, OpAmp A versus B respectively.

Before I start a sermon, what do you expect in technical terms, from the side of the devices under test, so that we can relate that all together to the hearing? I'm eager to do that.

I acceptthe notion, distortion measurements (of any kind) were only proxies, they were not the full story, that they form a plausibility argument at best, as fully true. Insofar I'm completely with you, HD, IM and such measurements need to be interpreted, there is no clear threshold and so on, they indicate, but do not prove. But the plausibility for 0,0xxx5% of HD to be utterly insignificant, hence 0,0xxx1% should be insignificant likewise, hence there is no difference between the two, is very strong. Devices characterized by those numbers are equivalent to the heraring. Very strong plausibility speaks for that, confirmed by so many "blind tests". It is you to formulate a valid counter argument. Go ahead, appreciated!

 

[Ricardus]

Apparently not.

And in this case, its not me, its you.

Your example is a classic use of the Straw Man fallacy.

Here, free education:

Straw man - Wikipedia

en.wikipedia.org

 

[Heinrich]

I asked a serious question of @Buckeye Amps ... why he chose the op-amps he did instead of going with the cheaper option if after all there are no sonic differences. ...

In reply:

Your example is a classic use of the Straw Man fallacy. ...

No, it is not. It is a very good, and may I say, a scientific argument. As objectivists we shall not overshoot.

As scientifically oriented people we must not be sure of anything, accept uncertainties and open questions (that is were the funding comes from, to begin with ;-).

In answer to @im_gumby, the choice is made between conflicting targets. Cost versus the least distortion, Distortion is seen as an indicator foe good sound for variaous reasons. Not that distortion could ever be perceived as such, for that they are hilariously too low. It is only utterly implausible, that the underlying non-linearity of an analog (!) device described by that distortion (the proxy) would generate artifacts, that are big enough to spoil the subjective experience. To long for even lower distortion is kind of a sport, nothing else. Like a Porsche in city traffic's stop and go. It's a sports car, right? LoL

Or to put it more gracefully, the least distortion underlines the qualty of the plausibility argument around it. We celebrates science, we do not know, we try the best we can with a wink of the eye, the sportsmen we are. It neither speaks for nor against the audibility of anything connected to the usualy very (unreasonably) low numbers.

Objectively the plausibility is strong as rock. It is on the deniers, an honorable role model really, to formulate counter arguments beyond saying: "no proof". We cannot provide proof, the scientists we are, no proof to be given.

If you need some grounds to justify your next purchase, well, it is up to you. Let me know once you come up with a new or otherwise interesting twist. If you feel attracted by costly opamps, go for it. Beauty is in the eye of the beholder. Taste cannot be disputed. I think its ugly to waste work and payment on a fata morgana.

 

[wwenze]

White chocolate actually has same or lower calories than normal chocolate. Hard to convince people who think otherwise because of the taste difference.

 

[amirm]

I asked a serious question of @Buckeye Amps ... why he chose the op-amps he did instead of going with the cheaper option if after all there are no sonic differences.

That has little to do with the topic at hand. When you are designing something new, you can strive to achieve the lowest noise and distortion with proper circuits all around the op-amp. Here, we are talking about taking an existing design and blindly replacing the op-amp, hoping that the more expensive the part, the better the sound. My data shows this to be plainly wrong. Leave it to the designer to do the right thing. If they do, it show sup in measurements relative to their competition. Don't confuse these topics.

 

[amirm]

There are two issues w Amir's tests.
1) Do they actually test for the information or is there something else?
2) Are Amir's interpretation of the results correct?

These are not the issues at all. As I have repeatedly explained, we know these op-swaps are fruitless based on what we know of electronic design and acuity of listeners. On that basis alone, we can say rolling op-amps is useless. Indeed, I said this in my first review of Douk. Then someone challenged me who is not technical so I thought I do these tests to show data that is much easier to understanding than circuit theory and psychoacoustics.

So what you are facing is three armies of engineering, science, and measurements. All of them dispute your claims of audibility. It matters not if measurements are incomplete. It matters that we know a ton about this area and they all instruct us of truth: that you are wasting your time here. When you swap op-amps, you turn your audio device into a placebo factory. It is that "feature" that makes you think there is a fidelity difference. Of course, placebo wears off after a while and you are back to square one, looking for another useless tweak.

 

[Heinrich]

It is on the deniers, an honorable role model really, to formulate counter arguments beyond saying: "no proof". We cannot provide proof, the scientists we are, no proof to be given.

The harmonic distortion measurement is a proxy for underlying non-linear properties of the device under test (DUT). One could go further, mathematically, to let the HD describe the non-linearity entirely, but would lead too far here.

Let's discuss if we can debunk the plausibility argument, that at some point HD was low enough:

We look for the worst signal to be on the safe side for the others. I suggest: two frequencies far above the individual hearing threshold, say 200001Hz and 20101Hz. The difference is 100Hz. The non-linear characteristic curve of the DUT will produce a differential tone. All the more so as the negative feedback decreases towards the top? No, because the negative feedback reduces ‘dirt’ according to the frequency at which it occurs, in this case 100Hz. (We know that).

Well, let's assume a peak level of 120dB, you don't do anything else. The differential tone mentioned would now have to be suppressed by about 120db in order to be guaranteed to be below the audible threshold, or the sound would be drowned out by noise. Noise levels today are around 110dB - let's stick with 120dB.

This 120dB corresponds approximately to a distortion factor of (6 decades, percentage calculation swallows 2 of them) 0.0001%. Well, that would be a comparative measure. What do we have?

The deniers could now test this (switch off your tweeters, otherwise they will be fried, it's only about the differential tone from the DUT in the amplifier).

I could imagine a sensible argument along these lines. But you never get to read it. There are probably at least two reasons for this. A thorough misunderstanding to the point that science has to provide positive ‘evidence’. No, that's not the case. Secondly, there is often a lack of understanding for the logical development of a technical argument on a strictly rational level. That's perfectly fine, it's an educational profession, really. I can't cover a roof or manage a farm either. We should complement each other, not distrust each other. You should distrust advertising, that's for sure. In that I (we?) want to help out.

 

[RickS]

With regard to recent trolling here, another mod issued a warning (and a ban) for almost identical trolling. I do not have access to all the details, and most of you have not experienced, but the template warning language concludes with the following…

“Any further violations after returning from a ban, may result in a full and immediate Forum Ban.”

Since he continued to troll upon his return, he has earned a full and immediate Forum Ban.

 

[SMen]

That has little to do with the topic at hand. When you are designing something new, you can strive to achieve the lowest noise and distortion with proper circuits all around the op-amp. Here, we are talking about taking an existing design and blindly replacing the op-amp, hoping that the more expensive the part, the better the sound. My data shows this to be plainly wrong. Leave it to the designer to do the right thing. If they do, it show sup in measurements relative to their competition. Don't confuse these topics.

I am not sure that the case of the v3-monos applies here either for the manufacturer striving for lowest noise and distortion, or people blindly replacing op amps, because

1. the Sparkos were both recommended and supplied by Fosi ... and
2. in this case the manufacturer was in fact selling the whole 'tuning' bandwagon as well as op amps...

and this is what was particularly bugging me!

I am presuming Fosi must have tested the Fosi's V3 monos with the Sparkos to be supplying them.

What this test did for me was confirm my own suspicion that the stock NE5532s in the V3 Monos were a perfectly valid choice, and fit for purpose! And much cheaper.
Forgive me if I now scarper from this particular topic!

 

[cgallery]

Has anyone every done an op amp nul test, similar to the nul test Bob Carver did in that amp challenge (I think it was in Stereophile)?

That would be sort of interesting, but I don't know if it is doable/practical.

 

[amirm]

the Sparkos were both recommended and supplied by Fosi ... and

That was a marketing move, not a designed feature. It makes little sense to design around these expensive op-amps when you can't show them doing anything better.

 

[SMen]

That was a marketing move, not a designed feature. It makes little sense to design around these expensive op-amps when you can't show them doing anything better.

Yes - this is really my 'bugbear and where I really see the benefit of calling this out with tests / putting in perspective here.