The Ten Biggest Lies in Audio

[paskinn]

I read some of this stuff with a sense of astonishment....the contributors so sure that 'measurements' tell us all we need to know are as ridiculous as someone who says that if you examine the chemical constitution of food you can be certain of what it will taste like.....without any acknowledgement of the role of the person who actually eats it!
Even if the 'Sergeworld' view was true it must require two (necessary) conditions:
1) That measurements are complete and involve the actual conditions of use (ie, distortion under dynamic conditions and so on.). In fact this set of conditions is never met...there are far too many variables.
2) Even if condition (1) were ever met, you would then need complete knowledge of the way the human brain intrepets and 'understands' these signals. (which might vary from person to person.).
These two conditions are never met, because it would be far too expensive and complex to do so. Which is why, Serge et al repeatedly make claims they cannot substantiate, even in the terms they claim to understand and accept (science.)
Of course there is another way.....it is called listening.

 

[sergeauckland]

Sorry, I don't subscribe to the dumb pipe analogy.

Great hi-fi is made by people with great taste in audio.

I mean come on. What hi-fi manufacturer produces equipment without every listening to it first?

There was an interview many years ago with Peter Walker of Quad, going into their engineering and design procedures. He said that they never listened to music on any of their products as part of the design process. They listened to tones and "funny noises" as he charmingly put it, but not to music. All design was done by calculation and measurement.

When I was designing audio equipment for the broadcast industry, it would never have occured to me to listen to music as part of the design process. Calculations checked by measurements were the way I did it. It was only at the very end of the process, once I had a product ready to hand over to production that I might listen to music on it, just as a final "reality check" in case I'd missed anything really silly, and for the satisfaction of hearing music played on something I'd designed. Not once did I ever find anything from listening to music that I didn't already know about from the measurements.

S.

 

[Joe P]

Baz,

I haven't had cable TV since 2005, so I rarely know what's current or what has made it from the other side of the pond over here. I'm essentially a DVD watcher.

Joe

 

[Basil]

Baz,

I haven't had cable TV since 2005, so I rarely know what's current or what has made it from the other side of the pond over here. I'm essentially a DVD watcher.

Joe

If you like 'black humour' the 'The League of Gentlemen' should be right up your street...

http://www.amazon.ca/League-Gentlemen-Collection-Mark-Gatiss/dp/B0007Y0982/ref=sr_1_1?s=dvd&ie=UTF8&qid=1318861794&sr=1-1

 

[positive_energy]

There was an interview many years ago with Peter Walker of Quad, going into their engineering and design procedures. He said that they never listened to music on any of their products as part of the design process. They listened to tones and "funny noises" as he charmingly put it, but not to music. All design was done by calculation and measurement.



S.

Well there you go - kinda explains why I have never liked Quad. Cold and boring. Lifeless.

 

[Mescalito]

I read some of this stuff with a sense of astonishment....the contributors so sure that 'measurements' tell us all we need to know are as ridiculous as someone who says that if you examine the chemical constitution of food you can be certain of what it will taste like.....without any acknowledgement of the role of the person who actually eats it!
Even if the 'Sergeworld' view was true it must require two (necessary) conditions:
1) That measurements are complete and involve the actual conditions of use (ie, distortion under dynamic conditions and so on.). In fact this set of conditions is never met...there are far too many variables.
2) Even if condition (1) were ever met, you would then need complete knowledge of the way the human brain intrepets and 'understands' these signals. (which might vary from person to person.).
These two conditions are never met, because it would be far too expensive and complex to do so. Which is why, Serge et al repeatedly make claims they cannot substantiate, even in the terms they claim to understand and accept (science.)
Of course there is another way.....it is called listening.

Everybody auditions before buying. But measurements are invaluable for shortlistng kit to audition. In my experience, if kit measures badly, it will not sound good, different, maybe, but never "right".

So, if someone suggests I audition, say, a CDP with a "doctored" frequency response, or a SET which measures badly, I will politely refuse.

I want my hi fi to have no other effect than to retrieve & amplify the signal recorded on the CD or whatever as faithfully as possible, without imposing character of it's own.

Chris

 

[sergeauckland]

Interesting, some questions:

1. Which combination of measurements and thresholds would indicate the performance has moved from being transparent to non-transparent?
2. What do you measure to prove a doubling of temporal accuracy?
3. Do manufacturers publish this information?
4. Did you use this information to form your purchasing decision?

Andrew

For a device to be transparent, it needs to have the following.

a) Distortion of all sorts less than 0.1% at all levels, frequencies and permitted loads.

b) A frequency response that is flat to within 1dB 20Hz -20kHz

c) Noise < 80dB below peak level unweighted

If it meets the above criteria, then it will pass a straight-wire bypass test and hence be transparent. It is possible to relax some parameters under certain circumstances, if one knows the conditions of use, but in general, if a deveice meets the above criteria, it will be transparent.

2) I have no idea what you mean by temporal accuracy, so have no idea how to prove a doubling of it. It is not a term I have come across.

3) Some do, some don't. Some publish them in a way that makes it impossible to know how they were measured, or with incomplete data. For example, if an amplifier is specified as having <0.01% distortion, but without further qualification, is that at full power only or at any power, within what range of frequencies, into what loads. In other words it's a meaningless specification that however looks good on a spec sheet.

4) Yes, I bought all my kit on the basis of specs. I've never chosen anything by first listening.

S.

 

[ultrawomble]

you would then need complete knowledge of the way the human brain intrepets and 'understands' these signals. (which might vary from person to person.).

If we accept the above is true then what does it do the validity of your subjectivist approach?

Of course there is another way.....it is called listening.

 

[Basil]

Well there you go - kinda explains why I have never liked Quad. Cold and boring.

Rubbish, unless you just happened to play something that was 'cold & boring'...

 

[Joe P]

Baz,

Rubbish, unless you just happened to play something that was 'cold & boring'...

Like Coldplay?

Joe

 

[Basil]

Baz,


Like Coldplay?

Joe

'xactly...

 

[Tony L]

So why are you willing to let item post such rubbish about how networks and computers work without comment?

It is not my position to correct every error on the internet! I only went after Ash (in a light-hearted manner) as he wrote some (deliberately?) misleading stuff about pfm on his / Darren's advertorial site. FWIW I don't think I've even noticed Item's posts outside of the LP12 suspension 'Spoke' thread - I don't read everything on pfm by any stretch, there are well over 1000 new posts a day, which is way more than I can get through so I tend to concentrate on topics that look like they might interest me (unless someone reports a post).

 

[sergeauckland]

Everybody auditions before buying. But measurements are invaluable for shortlistng kit to audition. In my experience, if kit measures badly, it will not sound good, different, maybe, but never "right".

So, if someone suggests I audition, say, a CDP with a "doctored" frequency response, or a SET which measures badly, I will politely refuse.

I want my hi fi to have no other effect than to retrieve & amplify the signal recorded on the CD or whatever as faithfully as possible, without imposing character of it's own.

Chris

Chris, I don't, as I don't see what I can tell more by listening than I can from the specs. If the product is transparent, then listening won't tell me anything about it, and if it isn't then I wouldn't want to buy it.

Even with loudspeakers, provided they have a decently flat frequency response, I find I can live with most 'speakers after a brief time to acclimatise. I bought my present Meridians without even hearing them, although I was very familiar with Meridian products.

S.

 

[Basil]

It is not my position to correct every error on the internet! I only went after Ash (in a light-hearted manner) as he wrote some (deliberately?) misleading stuff about pfm on his / Darren's advertorial site. FWIW I don't think I've even noticed Item's posts outside of the LP12 suspension 'Spoke' thread - I don't read everything on pfm by any stretch, there are well over 1000 new posts a day, which is way more than I can get through so I tend to concentrate on topics that look like they might interest me (unless someone reports a post).

I spotted that over there as well, almost posted about it here, but thought better of it, as starting a HDD v pink fish slanging match was not a good idea!

 

[Mescalito]

Chris, I don't, as I don't see what I can tell more by listening than I can from the specs. If the product is transparent, then listening won't tell me anything about it, and if it isn't then I wouldn't want to buy it.

Even with loudspeakers, provided they have a decently flat frequency response, I find I can live with most 'speakers after a brief time to acclimatise. I bought my present Meridians without even hearing them, although I was very familiar with Meridian products.

S.

Serge,

I audition because I am perfectly aware that, once I have eliminated the poorly-specced dross, I can indulge the sonic irrelevancies like look & feel, ergonomics and appearance. Most kit is just plain bloody ugly. I suppose audition is the wrong word. Physically & emotionally check out comes closer.

Of course, I end up spending far more than necessary, but hey, why not?

Chris

 

[MVV]

It is a shame that this discussion, and we have had it many times, gets so polarised. I wish I'd heard the views of the likes of Serge years ago it might have balanced the views I tended to listen to. One thing I know is if I had £250 to spend second hand on a thing to play music on I would start with his recommendations. In fact 'Serge's guide to ebay bargains' would be a site I would definitely visit.

 

[AndrewR]

2) I have no idea what you mean by temporal accuracy, so have no idea how to prove a doubling of it. It is not a term I have come across.

Regarding temporal accuracy.

Probably my biggest priority is having a system that will increase the perceived realism of the performance - getting closer to the original sound as recorded to the limits of the recording (i.e. transparency).

That realism can be characterised by many attributes. One that is often neglected is the fidelity of accuracy in the time domain. Perfect fidelity would playback the time domain to 100% accuracy of the recording. No fluctuations and no phase modulation.

Phase response is one of those metrics, but only tells part of the story.

Frequency response is another, in that a wider frequency response will playback more harmonics and the system would be more accurate at resolving EXACTLY WHEN a note starts, stops and does weird things in the middle (e.g. a sharp transient can only be faithfully reproduced with high harmonics present - a la Fourier Analysis)

Then there is consistency of timing. The most well-known metric being wow-and-flutter. That of course applies to turntables, but I am wondering whether such a similar thing exists in non-mechanical electronics. Power supply stiffness and responsiveness (charge & discharge times), noise, RF and heat all have some effect. If everything else is 100% perfect and non-mechanical wow & flutter is 99.8% perfect, then the designer needs to improve the non-mechanical wow & flutter to 99.9% to double the timing accuracy of the hi-fi equipment.

I reckon it is this non-mechanical wow & flutter that needs to be looked into, defined, measured (a very difficult task) and remedies identified.

Andrew

 

[Joe P]

Serge,

4) Yes, I bought all my kit on the basis of specs. I've never chosen anything by first listening.

Top marks for being honest and consistent, but given how easy it is for manufacturers to mislead with specs, not to mention the general dearth of meaningful specs in the first place, how can you be sure that the specs aren't leading you to the wrong conclusions?

Joe

 

[sergeauckland]

Serge,

I audition because I am perfectly aware that, once I have eliminated the poorly-specced dross, I can indulge the sonic irrelevancies like look & feel, ergonomics and appearance. Most kit is just plain bloody ugly. I suppose audition is the wrong word. Physically & emotionally check out comes closer.

Of course, I end up spending far more than necessary, but hey, why not?

Chris

Physical checking is indeed a good idea. Seeing how the stuff is made, and whether it can be serviced. I remember when I bought my Nakamichi I asked the dealer to take the covers off so I could see how bias and record sensitivity and EQ was adjusted.

S.

 

[Tony L]

I spotted that over there as well, almost posted about it here, but thought better of it, as starting a HDD v pink fish slanging match was not a good idea!

I'd have let it be - in the grand scheme of things it doesn't matter in the slightest. The internet is hardly a bullcrap-free zone (I've even noticed some here on occasion!), but when Ash brought it up here it seemed somewhat impolite to leave him muttering to himself when the facts were so easily to hand!