Master Thread: Are measurements Everything or Nothing?

[Robin L]

How is being less faithful to the original waveforms somehow being more faithful to the music?

1: It's a koan.
2: The ear is easily fooled.
3: The ancient alchemists were getting off on solder fumes.

 

[Robin L]

The musicians I know all are more interested in melodies, the 'art' around it, instruments and their tonal differences than interested in high fidelity sound reproduction.
Most of them do not even have h-fi gear either. They enjoy the music not the sound quality.

The musicians I know begin to worry once they hear their acoustic instrument starting to buzz.

 

[rdenney]

Ok. I guess now I would be interested to know your opinions on digital emulation of various instruments such as the tuba, violin etc? Are they accurate?

All models are false, though some are useful.

Yes, they are accurate in that they fall within a very wide space that encompasses the breadth of performance practice. If recorded and played back as part of an ensemble, they sound like a tuba. They do not sound like a musician in a solo setting, because by the time you have modeled all the necessary tools of expression, it's easier just to play the damn tuba, or hire a tuba player.

A lot of "orchestral" music heard on television is digitally created, and in the context of that playback envelope and use case (background music) does not demand a suspension of disbelief. Nobody is using it to create a standalone performance, but what it lacks is what the humans provide, not what the instruments provide.

But this is not the same thing at all as using digital tools to modify the artist's sound in ways the artist desires, versus using analog tools to make the same modifications, so it's a fun straw man but it's the wrong straw man in the wrong field.

Rick "acknowledging this does not imply approval" Denney

 

[rdenney]

Sorry, I assumed the sarcasm would have been obvious. It is (or at least was - I've stopped paying attention) something of an Audiophile talking point - "you're measuring the wrong things. technical accuracy doesn't matter - musical accuracy does"

Fair enough--your sarcasm was too dry even for me.

Rick "" Denney

 

[pkane]

Fair enough--your sarcasm was too dry even for me.

Rick "" Denney

Nothing’s worse than a wet sarcasm!

 

[Mnyb]

I would like a better definition of where we stands in regards of audibility of the issues we see in the measurements.
In normal listen conditions with music at sane levels at typical distances from the speakers.

You can probably differentiate 24bit vs 16bit if you set a reverb tail -70db down on repeat and cranked the volume to max and pressed you ear against the driver .

No contrived listening conditions that’s designed to highlight problems ( this a thing a manufacturer or designer should do btw ) I’ve learned on this forum that with the rigth test tone you can find problems that music in many cases masks ?

Has someone for example blindly ABX’ed 0,01% THD ?

Ok contrived controls are ok as a contrast to a real use case .

Example . I made a -112dB test tone at 1kHz or 2k it can faintly be heard in the noise floor if I press my ear against a driver on my digital active speakers with the volume at maximum . Siting in the sofa it’s more around -70 or -80 dB.

The most solid thing thing people said here is that -115dB is good for everyone in every situation all the time ?

In short I would like if the “transparent window” was better defined ?
For most of the DAC’s it’s a done deal as they so much better than our very human ability to discern them ? But then again where is the line or the gray area where things can get dicey ?

 

[kyle_neuron]

Pro violinists fail to spot Stradivarius in blind test.

I’ve been to the museum and spoken to artisans who are baffled by this (pun intended). I still enjoyed hearing a 400+ year old Stradivarius being played by a virtuoso.

The whole point of measurements is to quantify our subjective experience. The late Heyser spoke much on this, and perhaps it would be less ‘taboo’ if it was more normal. Do we need more people with ‘literal rocket scientist’ on their CV to make that normal? Why did it take so long for the seminal works of Toole and Olive to be recognised, let alone standardised?

More importantly, the fundamental question is ‘why are we so afraid of giving up the intangible?’ And that goes into the point where it’s gotta be some book reading. I find so many of these aspects fascinating, and analyse myself more than others - I have been constantly fallible in my ‘career’ as an audio engineer by my own metrics, yet my clients and peers have generally been impressed by the results I achieve, even across cultures or musical tastes.

Is that because I have ‘golden ears’? I don’t think so. It’s because I listen to the person who’ll be using the system, apply the accepted science to the use case, and more importantly - try to explain what and why I am doing these things throughout. A huge part of the audiophile bollocks is the ‘journey’ right? We all like to know something others don’t, or perceive that at least. Predatory sales tactics are as old as time. It just feels nice to sometimes turn them on their head and do it with myth and science combined.

Can’t beat em?

Anyway. I appreciate this site and others for the genuinely inquisitive nature of the members. Doing a measurement badly is better than not at all. Same for simulations or estimation. The idea is to LEARN. If I, a network operations manager who first encountered FFT as applied to DSL broadband frequency binning in Fujitsu MSANS, can achieve multiple stages at Glastonbury and gigs of accolade around the world by basically prodding friendly nerds for ten years, you can do anything

 

[dwkdnvr]

Sorry to cut you off

But I challenge this notion that objectivists distrust their hearing and subjectivists trust only their hearing. I rather think it's the reverse. Measurements-driven proponents trust hearing--it's the interpretation of the hearing based on the other senses that they don't trust. So, they applaud subjective testing when it is done with the proper controls, because that eliminates the biases introduced into the perceptual process by those other senses.

Well, I still think that saying they understand the limitations and flaws is accurate. You're correct that subjective testing is recognized as valuable when controlled, but IMHO that quite the same as 'trusting' their hearing.

Feelings-driven proponents, however, only claim to trust their ears. But in fact they don't. They refuse (usually) to conduct controlled testing, and thus allow their other senses to influence their interpretation of what they hear. If they are comparing $10,000 Mark Levinson amps against a $700 Buckeye, they will hear a difference, but the difference they hear is really the difference they see, coupled to the assumptions they make about what those differences will be. Those assumptions bias their interpretation of what they hear without the option, and without any specific acknowledgement or even awareness of the assumptions.

The only way to claim one judges based on hearing alone is to apply controls to subjective testing--the one thing "subjectivists" seem unwilling to do "because the difference is so obvious". If it's that obvious, then why are they so reluctant? I doubt it's all explained by laziness. Fear that their $10,000 investment might be invalidated must play a role in those biases, too.

Rick "who trusts his ears, but not his brain" Denney

This is a can of worms. My original post was certainly on what subjective audiophiles 'claim' or would report they believe. I do agree though that there is an obvious contradiction if not outright hypocrisy in their behavior. I don't tend to frequent the typical subjective audiophile forums much any longer, but the position now seems to be that ABX testing is inherently invalid because the stress of being tested disrupts the subjective listening experience and that is why audiophiles routinely fail to identify differences - the test is the problem, not their hearing acuity. I think they arrived at this one after previous attempts to explain away the ABX failures were debunked - e.g. "the switch box obscures the differences", "fast a/b switching is artificial", "I was unfamiliar with the room" etc. The obvious irony is that this absolutely validates the 'objectivist' position that external factors influence subjective impressions, but they don't want to really fully admit that so they try to have it both ways - sighted, uncontrolled listening is the only way to be free of the stress of being tested, and therefore the only valid listening evaluation; but don't worry - there aren't any other possible aspects of sighted listening that could possibly influence the experience.

Of course, even the language they use - 'being tested' - reveals the inherent bias - their 'audiophile cred' is on the line, and if they fail the test their reputation would take a massive hit. Which is too bad in some senses - IF there are actually audiophiles with higher hearing acuity, it's exactly those folks that it would be useful to include in properly controlled tests.

 

[Alice of Old Vincennes]

In the US, the SUV craze grew from the fleet mileage requirements. SUVs were considered “light trucks” and didn’t count towards the passenger car mileage average. So larger passenger models were phased out and SUVs ushered in.

Also, we are fatter, have lots of stuff, and boomers had to ferry their kids around to all their activities because the stickball in the street we grew up with was no longer safe.

Mini Van can hold more. Women don't like them anymore.

 

[rdenney]

Mini Van can hold more. Women don't like them anymore.

Haven’t seen a minivan yet that I can slide a sheet of plywood into. But, yes, minivans hold a lot.

Rick "so does my Expedition" Denney

 

[gallantus]

Fundamental principle of market-based economics: Price and cost are unrelated. Price is dictated by the market; cost is what it is. If the market will bear a price that exceeds costs (writ large), then the product will produce profit and should be made. If the costs exceed the price the market will bear, then the product isn't feasible and should not be made.

People with wealth often look for ways to display (or, more charitably, express) that wealth. Harman provides a range of products at a range of price points to find markets, not to qualify products. People who buy Mark Levinson amps will have something that creates an ownership experience those people would likely not get from a Buckeye, even if the Buckeye sounds as good in objective terms. That's worth something, or it isn't. If it isn't, the brand won't last at that elevated price point.

Rick "there is nothing wrong with this" Denney

Yes, there is a tier level. But are you saying Mark Levinson is really not better quality than Arcam? The definition of quality can take part in say, more luxurious chassis with a 4mm think aluminum, or Mundorf film capacitors, having dual transformers for a dual mono design, perhaps better engineering. Now whether all of these better quality in engineering results in better sound or not, is the topic of discussion here.

Someone use a GM and Cadillac as an analogy, but the fact is, a Cadillac has better quality leather, maybe a messaging chair. So are Cadillac buyers spending more money but not getting more?

So if I spend $10k on a Mark Levinson, what exactly more an I getting than Arcam that is not just the brand, or is it just the brand?

 

[tmtomh]

I would like a better definition of where we stands in regards of audibility of the issues we see in the measurements.
In normal listen conditions with music at sane levels at typical distances from the speakers.

You can probably differentiate 24bit vs 16bit if you set a reverb tail -70db down on repeat and cranked the volume to max and pressed you ear against the driver .

No contrived listening conditions that’s designed to highlight problems ( this a thing a manufacturer or designer should do btw ) I’ve learned on this forum that with the rigth test tone you can find problems that music in many cases masks ?

Has someone for example blindly ABX’ed 0,01% THD ?

Ok contrived controls are ok as a contrast to a real use case .

Example . I made a -112dB test tone at 1kHz or 2k it can faintly be heard in the noise floor if I press my ear against a driver on my digital active speakers with the volume at maximum . Siting in the sofa it’s more around -70 or -80 dB.

The most solid thing thing people said here is that -115dB is good for everyone in every situation all the time ?

In short I would like if the “transparent window” was better defined ?
For most of the DAC’s it’s a done deal as they so much better than our very human ability to discern them ? But then again where is the line or the gray area where things can get dicey ?

I agree with you, and would only emphasize one point: the performance standards and recommendations @amirm makes - and that anyone would make - are necessarily based on multiple possible use-cases, along with a best practice of having a little buffer/wiggle room.

So is a -115dB noise floor necessary in most real-world home listening situations? No, you are correct, it is not. If you take the maximum safe SPL in dB for sustained listening periods and subtract from that the dB level of a typical quiet home listening environment, you get something like 70dB.

But then consider headphone listening, and near-field listening, and the possibility that instantaneous peaks 3-10dB louder than the SPL level, and unit-to-unit variation from published measurements. And when considering a DAC or source component, add in the recommended 10dB additional SINAD compared to the downstream power amp, and perhaps add a buffer for DSP, and -115 to -120dB doesn't sound like such an excessive benchmark anymore.

Finally, none of this is to disagree with your larger point that it helps to have clear information about the limits of audibility. In this regard we could always have more and better info - but it is worth noting that Amir's reviews are filled with comments about that, especially with regard to distortion and jitter levels.

 

[ahofer]

I think the basic primary gulf between the 'objectivist' leaning crowd and the 'subjectivist' leaning crowd rests in the difference in perspective on the human auditory system. The 'objectivist' leaning cohort accepts the results from psychology that our hearing is actually rather flawed, and that bias (both conscious and unconscious) has a large influence on subjective perception. i.e. they understand that we really can't trust our ears in many cases. Conversely, the subjective camp believes strongly in both the resolution and repeatability of our hearing. They have a perspective that seems to be rooted in the belief of an 'invariant' listener, leading to the belief that the only possible explanation of a difference in subjective experience is that there is an objective difference in the sound field.

This is a really good point: The gulf is as much over-confidence in hearing and aural memory as it is a lack of confidence in measurements. A bit of chicken and egg, though. Proper respect for measurements and over-confidence in hearing would lead to cognitive dissonance.

 

[tmtomh]

Yes, there is a tier level. But are you saying Mark Levinson is really not better quality than Arcam? The definition of quality can take part in say, more luxurious chassis with a 4mm think aluminum, or Mundorf film capacitors, having dual transformers for a dual mono design, perhaps better engineering. Now whether all of these better quality in engineering results in better sound or not, is the topic of discussion here.

Someone use a GM and Cadillac as an analogy, but the fact is, a Cadillac has better quality leather, maybe a messaging chair. So are Cadillac buyers spending more money but not getting more?

So if I spend $10k on a Mark Levinson, what exactly more an I getting than Arcam that is not just the brand, or is it just the brand?

You answered your own question: If you spend $10k on a Levinson you are not just paying for the brand. You are paying for other aspects that add to the cost, and that are real things that you can see and touch and that might give you enjoyment and/or make you feel better than if you owned the Arcam. But as to whether or not any of those things make the Levinson sound better, that's a different question that does not necessarily have a super-tight correlation with the increased cost.

And remember: if we're going to compare apples to apples sonically and performance-wise, then the Levinson and Arcam would have to have the same (or similar) wattage, the same gain, similar (or similarly inaudible) noise and distortion levels, similar behavior with low or complex/"difficult" impedance loads, similar current capacity, and perhaps comparable damping factors (and maybe other things I am forgetting). The fact that both might read the same on a frequency-response graph into a dummy load at 1 watt or 5 watts does not tell the whole sonic story - there are, as @solderdude has noted, multiple measurements that must be taken together to paint a full picture of whether or not two pieces of gear really can be expected to perform similarly from a sonic perspective.

But yes, if all those things are equal or sufficiently similar, then from a sound-reproduction point of view, one could reasonably say that a Levinson is not really better than an Arcam - or at least the Levinson would not be better sonically out of the box. Years down the road, there might be a statistically significant difference in each unit's likelihood of maintaining its original performance level and continuing to work without breaking in any way.

 

[gallantus]

But yes, if all those things are equal or sufficiently similar, then from a sound-reproduction point of view, one could reasonably say that a Levinson is not really better than an Arcam - or at least the Levinson would not be better sonically out of the box. Years down the road, there might be a statistically significant difference in each unit's likelihood of maintaining its original performance level and continuing to work without breaking in any way.

So it sounds like what you are implying is that better parts used, caps, diodes, etc will net you longevity not sound quality?

 

[MakeMineVinyl]

The definition of quality can take part in say, more luxurious chassis with a 4mm think aluminum, or Mundorf film capacitors, having dual transformers for a dual mono design, perhaps better engineering. Now whether all of these better quality in engineering results in better sound or not, is the topic of discussion here.

In a higher end amplifier vs a more conventional one, both with the same power rating - more and beefier output transistors on a larger heatsink. Higher capacity power transformers. More power supply filter capacitance. General upgrade in parts quality (Wima capacitors vs something else etc.).

Thicker aluminum front panel with laser etching.

 

[ahofer]

When top jazz guitarists and folk like Jeff Beck go the way of purely SS and DSP, I think that the technology will have arrived.
Just my personal opinion.

You can 'play' tube distortion with subtle or major changes in position, form, and strength of string attack (not to mention moving closer to the amp). The tubes feel like they are responding directly to tactile inputs, although this is with VERY large amounts of overdrive and distortion relative to audio. While this touch-sensitivity seems like a modeling challenge, I've seen good players use modeled amps, even the basic Line 6 variety, and other guitarists are left wondering whether it was a real amp or modeled. I have a cheap fender-modeling pedal that is surprisingly good, and offers the right sound at low volume (impossible with the amp itself).

Tube warmth/distortion in music playback is of such a different scale from audio reproduction that I don't think it's a great analogy to guitar amps, whether modelers could in fact pass that blind test or not.

I googled around. This test has obvious treble balance differences, so it's really a preference between different sounds.

Here's another. I got the first two right pretty easily, but mixed up the software and SS in the last. But again, there are obvious treble balance differences.

So I guess people are going at this.

btw, Michael Landau, like Jeff Beck, is a genius at manipulating distortion with attack.

 

[tmtomh]

So it sounds like what you are implying is that better parts used, caps, diodes, etc will net you longevity not sound quality?

There’s no need to infer what I’m implying. I’ve explicitly stated that some enhanced parts are likely to impact sound quality, others might, and others are unlikely to do so. I do not claim to have comprehensive knowledge of which parts fit into which category in every piece of gear.

However, part of my point about the multiple different measurements we need to take in order to determine if two amps really do measure similarly is that some parts do clearly impact measurements. So a more robust power supply with higher-rated, more reliable caps will likely produce more wattage, better frequency response at high wattage, and better current-handling capacity (and possibly lower noise, although that depends on many factors).

But that is not the same thing as saying silk capacitors create a smoother high end than paper ones - and it’s also not the same as saying the Levinson’s super-heavy casing reduces vibration thereby sharpening the soundstage presentation.

 

[escksu]

Understood, but something like this could work in order to give an overall rating:

View attachment 172832

too much work though, perhaps ...

Ask you a question. How do you determine whether is excellent or fair??

 

[gallantus]

There’s no need to infer what I’m implying. I’ve explicitly stated that some enhanced parts are likely to impact sound quality, others might, and others are unlikely to do so. I do not claim to have comprehensive knowledge of which parts fit into which category in every piece of gear.

However, part of my point about the multiple different measurements we need to take in order to determine if two amps really do measure similarly is that some parts do clearly impact measurements. So a more robust power supply with higher-rated, more reliable caps will likely produce more wattage, better frequency response at high wattage, and better current-handling capacity (and possibly lower noise, although that depends on many factors).

But that is not the same thing as saying silk capacitors create a smoother high end than paper ones - and it’s also not the same as saying the Levinson’s super-heavy casing reduces vibration thereby sharpening the soundstage presentation.

So if some enhanced part might impact sound quality, is that sound quality always measurable by an Audio Precision Analyzer?

This is a honest question, some here say anything related to the sound can be measured by the Audio Precision Analyzer. So say no, some is I don't know, need evidence.