advertisement


fo.q tape

@spt, we can only dial up distortion until we can hear it and note its audibility. Even then we only know it's audible, one man may prefer it the next not, that's how taste is.

I'll settle for kit that's measurably blameless, with all distortion below audibility and I'm prepared to accept that's just it sounds, honest.

Might I prefer something else, quite possibly, but life's too short for me to chase that via endless box swapping, hunting sympathetic combos.
 
@spt, we can only dial up distortion until we can hear it and note its audibility. Even then we only know it's audible, one man may prefer it the next not, that's how taste is.

I'll settle for kit that's measurably blameless, with all distortion below audibility and I'm prepared to accept that's just it sounds, honest.

Might I prefer something else, quite possibly, but life's too short for me to chase that via endless box swapping, hunting sympathetic combos.
It's not just about distortion though. I'm also thinking about those other attributes that go into a subjective SQ assessment, like imaging, timing, soundstaging, dynamics, 'solidity' and so-on. And yes, distortion will account for some of what is/isn't perceived there, no doubt, but how do we know what distortion, what parameters, are associated with which attribute?

And moreover, distortion doesn't have to be actually audible to have a negative effect. Look at what happens to perceived SQ when the noise floor raises from one inaudible value to a higher inaudible value. People report things like 'greyness', lack of dynamics or tonal colour, and stuff.
 
But then again, even if a blind test shows* a correlation with measurements, that’s not causation.
Why should there be?

So how can we be confident that the measurements are telling us what we think they are? Is it not more likely that we measure what we measure because we can, and we talk ourselves into believing that the measurements tell us something meaningful about sound quality?
There is this thing called science: about electricity, hearing, acoustics, sound and psychoacoustics - to name a few areas relevant to this discussion. You know, it has been going on for some time now, accumulating the body of knowledge these measurements are based on.

*Edit: and let's not lose sight of the fact that blind tests just ask the listener to say 'different/not different', there's no qualitative aspect to a blind test.
So assuming the blind test shows (to a statistically valid degree) a 'different' result, then you're still left with taking on trust any subjective report as to what the actual audible difference in sound quality is.
Let's not lose sight of the fact that there are blind tests with qualitative aspect, too. It is also good to realize that existence of a mere difference is a necessary condition for existence of a qualitative difference.
 
There is this thing called science: about electricity, hearing, acoustics, sound and psychoacoustics - to name a few areas relevant to this discusssion. You know, it has been going on for some time now, accumulating the body of knowledge these measurements are based on.
Okay, so what science tells you which measurement relates to the perception of, say instrumental solidity? How does it explain this correlation? How does increased distortion affect perception of sound staging? What role does the noise floor play in perception of tonal colour, say? What perceptual science proves the correlations?
 
Honestly, I don't know what any of that wordy stuff really means. Instrumental solidity, timing, sounstaging.
 
Honestly, I don't know what any of that wordy stuff really means. Instrumental solidity, timing, sounstaging.
Well, these are quite common parameters, presumably them, or something equivalent, appears in Octavian’s scientific literature. But the point is, at some stage, even under scientific conditions, you know, lab coats and everything, there comes a point where somebody in a lab coat has to take on trust the listener’s subjective description of what they heard.
 
That's the thing they aren't parameters, they're opinions of something nebulous, more bass, less bass, forward, edgy recessed, that's all FR stuff and well correlated.

I wouldn't even know where to start with something like timing. I just don't believe such difference exist with blameless kit. Better than -120db for thd, imd, ruler flat FR and able to drive the connected load.
 
I guess the root is that some people think that if they feel they enjoy something more that in should correlate to some aspect of measured performance.

But should it? There's no reason why we should prefer the least distortion there's no reason why one person's preferences should align with another's. I mean, people love the ls35a, but it was only ever a 2nd tier monitor intended for voice. Yet they'll take it over anything else, even though its obviously frequency curtailed and severely volume limited.

Get much better than a badly distorting design and it only really comes down to preference. I prefer to know I'm using measurably blameless gear and am happy to accept that what I hear, speaker and room choice excluded, as pretty much what was recorded.

I don't blame the gear if I don't like the music, I just accept the artist, or maybe the engineer doesn't make stuff I like and I move on. I don't look for hifi that will make me like it by being less faithful. But each to their own.

One view isn't more right than the other as long as we are honest with ourselves about what we are actually striving for
 
Like: ‘how much I’m enjoying the music’.That sort of thing? I guess we can manage without knowing that, if we must. Does rather negate the whole exercise, though.
Measurements tell us quite a lot about how much you are enjoying the sound of music vs. how much you are enjoying the sound of your equipment. Many subjective hobbyists seem to be unwilling or unable to make this distinction.
 
I guess the root is that some people think that if they feel they enjoy something more that in should correlate to some aspect of measured performance.

But should it? There's no reason why we should prefer the least distortion there's no reason why one person's preferences should align with another's. I mean, people love the ls35a, but it was only ever a 2nd tier monitor intended for voice. Yet they'll take it over anything else, even though its obviously frequency curtailed and severely volume limited.

Get much better than a badly distorting design and it only really comes down to preference. I prefer to know I'm using measurably blameless gear and am happy to accept that what I hear, speaker and room choice excluded, as pretty much what was recorded.

I don't blame the gear if I don't like the music, I just accept the artist, or maybe the engineer doesn't make stuff I like and I move on. I don't look for hifi that will make me like it by being less faithful. But each to their own.

One view isn't more right than the other as long as we are honest with ourselves about what we are actually striving for

It seems to me that the opening paragraph above has it the wrong way round. The argument from the objective viewpoint is that measured performance is sufficient to assess the capability of the equipment to create a pleasurable experience for the listener.

I do agree that what people find pleasurable differs - like you I don’t venerate the LS3/5a - and that fits with my comments upthread about SETs too.

From which it follows, I think, that there’s no clear correlation between perceived SQ (considered as the amount of pleasure the listener derives) and the measured engineering performance. That’s essentially what I’ve been arguing upthread.

This does suggest that measurements aren’t the whole story, and can’t be a reliable predictor of how people will respond to the equipment. The scientific method takes understanding and makes predictions, which it tests to validate or disprove a hypothesis. So if you can’t make reliable predictions on the basis of what you think you know, where does that leave you?
 
... This does suggest that measurements aren’t the whole story, and can’t be a reliable predictor of how people will respond to the equipment. ...
"How people will respond to the equipment" is a very rare objective indeed when professionally making any measurement on a piece of equipment. There are many other much more common, perfectly valid objectives.

You do get a few tests with such an objective. Such as Floyd Toole and Sean Olive setting out to connect what they measure from loudspeakers with how people perceive them. Mostly the industry simply does not go there because of the difficulty. Test results frequently get published without any accompanying context to say in what way they may be useful (or not useful). That leaves things open to people quite naturally over-interpreting what measurements may tell you by assuming a purpose they were not set up to achieve.

But there won't ever be a resolution to the issue and discussion will continue ...
 
"How people will respond to the equipment" is a very rare objective indeed when professionally making any measurement on a piece of equipment. There are many other much more common, perfectly valid objectives.

You do get a few tests with such an objective. Such as Floyd Toole and Sean Olive setting out to connect what they measure from loudspeakers with how people perceive them. Mostly the industry simply does not go there because of the difficulty. Test results frequently get published without any accompanying context to say in what way they may be useful (or not useful). That leaves things open to people quite naturally over-interpreting what measurements may tell you by assuming a purpose they were not set up to achieve.

But there won't ever be a resolution to the issue and discussion will continue ...
Yes, I’d agree. The reliance on test results as sufficient is, IMV, misconceived. Tests are first and foremost an engineering tool. How many times do we read designers saying that they use measurements to get to a functional point, then listening to tweak the product to its final form. So if that’s the case, it’s evident that measurement is not the be-all and end-all that some claim it to be.
 
This does suggest that measurements aren’t the whole story, and can’t be a reliable predictor of how people will respond to the equipment. The scientific method takes understanding and makes predictions, which it tests to validate or disprove a hypothesis. So if you can’t make reliable predictions on the basis of what you think you know, where does that leave you?
A big part of the problem is that - particularly in the subjectivist hobby - the subjective reactions are very much based on imagination and expectation if not downright superstition - rather than what is actually being heard. You can never have a correlation between pink unicorns and the objective reality.
 
A big part of the problem is that - particularly in the subjectivist hobby - the subjective reactions are very much based on imagination and expectation if not downright superstition - rather than what is actually being heard. You can never have a correlation between pink unicorns and the objective reality.

That’s a rather colourful, but not particularly inaccurate, description of expectation bias!
 
A big part of the problem is that - particularly in the subjectivist hobby - the subjective reactions are very much based on imagination and expectation if not downright superstition - rather than what is actually being heard. You can never have a correlation between pink unicorns and the objective reality.
The subjective reactions are The Whole Point. I sometimes wonder if you may have forgotten that the purpose of hifi is the enjoyment of music, not 'the accurate reproduction of a source file'. The accurate reproduction seems, in some minds, to have become the master rather than the servant.

I maintain that it is important to attempt an understanding of what causes the subjective response in people; I know people who adore SETs, despite their less than state of the art measured performance, and others who can't abide them. I know people who are exquisitely sensitive to timing, or pitch, and others who really don't mind about them. It seems clear that the human experience is on a continuum (as with so many human things) so one set of measurements is never going to be sufficient to define everything for everybody. It follows that a slavish devotion to measurement as the ultimate arbiter of quality or value is ultimately doomed. The dismissive response to reported, subjective experience, as exemplified in the post above is, well let's call it unhelpful.
 
Yes, I’d agree. The reliance on test results as sufficient is, IMV, misconceived. Tests are first and foremost an engineering tool. How many times do we read designers saying that they use measurements to get to a functional point, then listening to tweak the product to its final form. So if that’s the case, it’s evident that measurement is not the be-all and end-all that some claim it to be.
However, I perceive the opposite end of that spectrum - the "measurements mean nothing" end to be equally absurd. And I even see the occasional absurd² example (as I perceive it, anyway) of extreme objectivity in mutual support of extreme subjectivity.

IMHO there are deep rabbit-holes at the extreme-subjectivity and extreme-objectivity ends of the spectrum into which people may venture. If that's how they enjoy the hobby then fine, so perhaps my "deep rabbit-holes" characterization may not be appreciated. Maybe I should find a less contentious term. However it is how I see it from my PoV without any intention to criticize anyone else's way of having fun. People are very diverse and I won't tell them how to enjoy their hobby.
 


advertisement


Back
Top