Accuracy Part 3.

[LiloLee]

oedipus and Pete, yes I do agree that there must be a point at which distortion is irrelevant, so I suppose the real question I would like answered is "At what actual level of distortion does it become irrelevant. 1%, 0.1%, 0.001% ....or what?"

 

[michaelab]

Lee, I think oedipus is referring to room effects, which have a rather greater impact on the accuracy of a system than the measured electrical accuracy of its electronic boxes.

 

[PeteH]

No they don't measure the same. The measurements show that they don't. They may sound the same or they may sound different, but the measurements won't show this.

Using instruments sufficiently accurate and sufficiently precise you can in principle measure the difference between anything and anything else - for example, the same recording played back twice in succession through the same system, due to tiny variations in air temperature and so forth. This doesn't mean it's audible though.

Human hearing has been studied in great detail and an awful lot is known about its capabilities and limitations. Accordingly, we can in most cases say whether or not a difference is large enough to be audible, and borderline cases which might sometimes be audible can be identified as such. For example, the mains hum measured in the Arcam's left channel at -120dB could with certainty be described as inaudible, as could the crosstalk at similar levels.

 

[oedipus]

This is the one I really want a reply to

So what is the rest of the story if it isn't measurement?

You need to take the measurements AND what is known about human hearing (of distortion) to be able to claim that the differences are audible. The onus is now on you to show "that the ear is capable of detecting these very small levels of distortion".

 

[PeteH]

I suppose the real question I would like answered is "At what actual level of distortion does it become irrelevant. 1%, 0.1%, 0.001% ....or what?"

I haven't got any figures to hand, nor do I actually know anything about it. I'm sure oedipus does though.

It has been shown though that artificially adding enormous amounts of jitter is for all practical purposes inaudible up to the point where the DAC is unable to lock onto the data stream. IIRC amplitude differences are subjectively detectable at around the +/- 0.2dB level (? - again, not entirely sure) within a few octaves of 1kHz, although they're frequently not identified as such.

 

[LiloLee]

I accept that I was getting carried away think I can hear difference of 0.0015%. But lets face it I must be deaf because I like valve amps that frequently distort up to 3%.

You need to take the measurements AND what is known about human hearing (of distortion) to be able to claim that the differences are audible.

So how this appears to me is that measurements aren't everything because of the way human hearing works, and since my ears are different from your ears what may appear significant to you might be inaudible to me.

 

[PeteH]

I accept that I was getting carried away think I can hear difference of 0.0015%. But lets face it I must be deaf because I like valve amps that frequently distort up to 3%.

Valve amps definitely do give audible distortion, but many people like the resultant sound and that's fine. ISTR that harmonic distortion is generally audible at something like the 0.3% level (?) or thereabouts - not 0.0015% anyway. Again I'm sure oedipus would have more to add on the subject.

So how this appears to me is that measurements aren't everything because of the way human hearing works, and since my ears are different from your ears what may appear significant to you might be inaudible to me.

It's not out of the question, but another thing that's been shown is that there's isn't actually a great deal of variation in what people can or can't hear - the 'golden ear' is definitely a mythological creature, at least in terms of being able to hear very fine distinctions that lesser mortals cannot (obviously this doesn't include people who actually have hearing damage).

 

[julian2002]

lee, give it up they have moved the goalposts so many times and redefined what they 'have been saying all along' that the 'playing field' is now nothing but holes where the goalposts used to be.
this seems to be a standard tactic of those who enjoy music via an oscilloscope and multimeter.
cheers


julian.

 

[PeteH]

lee, give it up they have moved the goalposts so many times and redefined what they 'have been saying all along'

Not redefined - clarified numerous misunderstandings that have arisen on the way through.

this seems to be a standard tactic of those who enjoy music via an oscilloscope and multimeter.

Oh for heaven's sake . Actually, you know, we enjoy music much more because we're not constantly worrying about whether or not we need to get our CD players reclocked or change our cables.

 

[7_V]

oedipus and Pete, yes I do agree that there must be a point at which distortion is irrelevant, so I suppose the real question I would like answered is "At what actual level of distortion does it become irrelevant. 1%, 0.1%, 0.001% ....or what?"

It's not quite that simple. The published data that I have read over the years indicates that:

There are different types of distortion - intermodulation distortion is not the same as 2nd harmonic distortion is not the same as 3rd harmonic distortion, etc., etc., etc.

Different types of distortion have different levels of audibility at different source sound levels and at different frequencies. Moreover these can be masked to differing extents by other stuff going on at the same time in the music.

Different people have different sensitivities to detecting different distortions.

What complicates the issue still further is that the reverberation, echos and reflections in any given concert hall will mimic distortion in the decay of each note.

 

[PeteH]

Different types of distortion have different levels of audibility at different source sound levels and at different frequencies. Moreover these can be masked to differing extents by other stuff going on at the same time in the music.

True. However, ISTR that there is data to indicate what is and isn't audible under ideal conditions - presumably if you can't hear a given amount of a given type of distortion in a test tone under anechoic conditions in a laboratory, then it's probably reasonable to say that you're unlikely to be able to hear it in music under less ideal conditions. And of course lots of experiments have been done using actual music too.

 

[oedipus]

I accept that I was getting carried away think I can hear difference of 0.0015%. But lets face it I must be deaf because I like valve amps that frequently distort up to 3%.

It's not quite that simple. The published data that I have read over the years indicates that:
<...>

It's certainly not simple. The factors you mention are all real. But, as PeteH pointed out, those factors tend to make it more difficult for the ear to detect distortion. For real music, at domestic loudness levels, you'd have a tough time justifying numbers lower than 1% for 2nd Harmonic and 0.1% for higher order harmonics. Note that those numbers are for individual harmonics rather than the THD+N style numbers used for component measurements, but it follows that no matter what the harmonic distribution, if the THD+N is less than 0.1%, then the distortion isn't audible.

 

[Saab]

Oedipus

Your knowledge of all things science is impressive,as is your determination to persuade all non-believers that science and measurements offer the only route to HIFI nirvana.Except you forget one thing,religion is a faith and can't be bought with your science.In the same way you can't persuade the Naim forum that a DVD player costing £200 will only be bettered if a 5k Naim CDP is made with flawed engineering,you can't persuade anyone in here to forget their ears and thus their own tastes,and rely on scientific measurements.
You seem happy when listening to music that your gear compys with your own scientific view of how it 'works',and thus 'measures',which is good for you,it clearly keeps you happy.
But please,can you give it a rest in here? Its obvious to a blind man and a galloping 'hoss,that this forum is frequented by people who use HIFI as an ends to a means,i.e to enjoy music,but with some enjoyment of owning some nice equipment and admittedly,a smidgeon of scientific gratification.But you take this to a level that is getting a tad nauseating.
As has just been said in PF

'edipus,

I haven't heard bollox like this since the 80s, when they were hyping the first CD players.

Regards

just as they don't want to hear someone trash their expensive CDPs,some in here I'm sure can't be arsed to argue anymore about sciencific measurements of distortion/accuracy,whatever (I for one have lost the plot when it comes to figuring what point you and BD are trying to make).

I speak for myself obviously,the rest may disagree.God help me if i ever sit listening to music and start thinkng about this sort of stuff and forget to enjoy the music.You are the classic 'equipment' listener and as a result the classic HIFI bore.FFS man get a grip and listen to some Dylan,or any other music where lyrics are more important than the reproduction quality.Maybe soem flicker of emotion will batter the science into submission,although I doubt it. (Try 'Masters Of War',from The Freewheelin',its about the Cuban missile crisis,if that track has you trying to identify strengths in your scientific equipment you are fvcked,a basket case,and can't be helped)

 

[notaclue]

This is a hi-fi forum. I expect discussion of hi-fi. oedipus is both knowledgable and interesting on the subject of hi-fi. Heck! He may even be right...

 

[PeteH]

Its obvious to a blind man and a galloping 'hoss,that this forum is frequented by people who use HIFI as an ends to a means,i.e to enjoy music,but with some enjoyment of owning some nice equipment and admittedly,a smidgeon of scientific gratification.

I can't speak for oedipus (in fact I can barely speak for myself because I'm away from my proper internet connection for the next couple of weeks - I hate fecking dial-up ), but how does an understanding of the engineering basis of hifi preclude using it as a means of enjoying music? In fact, if you aren't constantly trying to tweak your system to perfection and checking your favourite test discs over and over again, wouldn't that actually give you more free time to enjoy music?

just as they don't want to hear someone trash their expensive CDPs,some in here I'm sure can't be arsed to argue anymore about sciencific measurements of distortion/accuracy,whatever (I for one have lost the plot when it comes to figuring what point you and BD are trying to make).

I've seen oedipus put an awful lot of backs up for one reason or another, but I can't remember seeing anyone actually convincingly dispute anything he said. And I'm not sure anyone knows what point BD is trying to make

God help me if i ever sit listening to music and start thinkng about this sort of stuff and forget to enjoy the music.

See above. I'd like to write a bit more about this, but unfortunately my connection will probably drop again in a minute

FFS man get a grip and listen to some Dylan,or any other music where lyrics are more important than the reproduction quality.

In Dylan the lyrics are more important than the actual music, which frankly is fortunate as the music's pretty much non-existent

Maybe soem flicker of emotion will batter the science into submission,although I doubt it.

Or you might start wondering if your mains quality is worse today, or worrying if your interconnects are good enough, or if you should really get a new rack for your amp, or if your CD transport is good enough, etc. etc. etc. which might be a tad distracting from the music too.

 

[wadia-miester]

Well Pete based on the above, you need to sell your gear, get a $100 dac, a dvd transport a rotel 1062 & some Kans, and you'll happy happy listening for enternity.

 

[Saab]

I should have backed out of this when the anal bush made an appearance.Won't argue with anything you have said Pete,said my piece.But Dylan is a genius,you are just wrong

 

[michaelab]

I don't see why oedipus should "give it a rest". He's making good arguments, backed up with sound scientific knowledge and references. No one's saying you have to agree with him. This is a thread specifically about accuracy so his contributions are particularly relevant.

Also, don't go lumping BD together with oedipus. BD doesn't even know himself what he's talking about or what point he's trying to make half the time and rarely makes a rational argument backed up with evidence for anything .

And I agree with PeteH, just because someone's making rational scientific arguments doesn't in any way imply that they enjoy their music any less. They probably enjoy it more since they don't have to worry about all the pseudo-science audiophile voodoo.

Michael.

 

[GrahamN]

Pete has said most of what I wanted to say in reply to Saab - whose post irritated me beyond belief. Sorry we're having a go at you Saab - but this was just the most recent of the seemingly wilful misinterpretation of the engineering approach. I really cannot understand Julian's misunderstanding of this point - unless it's confirmation that (despite the best efforts of the profession) software engineering (or at least his version of it ) has nothing to do with real engineering!

Unfortunately, when rational argument and evidence comes up against blind faith, blind faith will frequently win and happily continue with the dark ages - although will probably succumb in the long run (qv. one of the most famous cases, Galileo). It's still staggering to hear though that over 40% of US citizens are more inclined to believe in the creation story of Genesis that in evolution. It's also interesting to note that some of the best scientists in the country are devoutly religious and others are devoutly atheist (and while I'm not in that august group under consideration I've been both). It all depends on your approach to that you don't understand.

However, in most cases progress is only made by the application of scientific and engineering disciplines. If you don't understand what is causing the differences you perceive (always assuming they are real and not just due to subconscious suggestion :yikes: ), you are stuck in the expensive and time-consuming regime of chance (aka in scientific circles as "monte-carlo" methods or, when combined with a feedback mechanism, as "genetic"). It's a pity that so much of the information we seem to spend so much bandwidth arguing about is only available if you have access to the AES archives or a specialist audio-engineering library.

Why is it the subjectivists keep on characterising those with an interest in understanding the "why" questions as a bunch of nerds sitting listening to silence with no appreciation of music? From what experience I have, I'd say the exact opposite.

[Long and incredibly boring discourse on what constitutes good science and engineering deleted - the engineers know it already....and the religionists will never understand it ]

Developing new products does require the engineering approach (unless you have a staggeringly good marketing department - or a mind-numbingly gullible captive market). Unfortunately engineering does frequently get corrupted once in gets into the marketing and sales departments (I frequently cringe to see the distortions there - significantly greater than 0.015%, or even 2%). Once you've done the engineering though, the important thing is to enjoy the final product.

Dylan may indeed be a genius (I wouldn't really want to comment in either direction) but, if so, it's such a shame that doesn't extend in any direction musical!

 

[julian2002]

graham,
just want to say that my 'version' of software engineering (games) is a far cry from what you'd want from say a bank or spreadsheet. it would be a positive advantage when trying to get squads of enemies to react reasonably realistically if occasionally 1+1 != 2. heavy use of random numbers isn't something most software engineers have to utilise that often either i would think. therefore if my outlook is rather 'fuzzier' than you would expect from someone in the bit mines then i appologise but i'm a product of my environment.
cheers


julian