fao null tester believers

[julian2002]

td,
as you say a perfect copy of the picture subtracted from the original will indeed result in a blank page but what does this tell me about the picture, whether i like it, does it move me, is it colour or black and white, is it a materpiece or the inside of the lens cap. this is my problem with measurement as the sole means of judging hi-fi, it just seems to miss the point of the whole thing.
cheers


julian

 

[Tube_Dude]

td,
as you say a perfect copy of the picture subtracted from the original will indeed result in a blank page but what does this tell me about the picture, whether i like it, does it move me,

If it moves you or not must depend of the original ...the music and the performance .
Not from the medium, that brings the original to you...

 

[Saab]

No Saab i don't underestimate the human senses.

There are no sound ,as live music...but in a copy of the live music , the more similar (less null) ....the more like the original the sound will be.

except the emotional reponses required to make it sound like it did when you watched it.I understand your view,its an entirely scientific stance,I take a simpler emotional veiwpoint to music listening.

 

[Tube_Dude]

I understand your view,its an entirely scientific stance,I take a simpler emotional veiwpoint to music listening.

And I understand your view...is why the world is a beautiful place to live...so many views...so many opinions...and in a hundred years on , nothing is important...

L'important...c'est la rose... :MILD:

 

[spxy]

"Imagine a photo... if you can make a copy of that photo ,that any trace of that photo and any color coincide with original photo...you have a perfect copy of the original.

That is what the null test do...if the output coincide at any point to the input sound ,there is no sound in the null point."


But you cant do that in reality with a very high resolution photo.
Also note with a photo the resolution is so low compared to vision, most cameras only use a few million pixels, I'm not sure what the resolution of the eye is but its far far higher.So you colud be matching low rez photos but as the resolution becomes higher it gets harder to match, then you have to consider that we see things in 3-D and a photo camptures none of that so although we can equalise low levels of distortion asa the resolution gets higher it gets harder and harder to do as the is so much information to match.
You would have to have very accurate equipment to do so.
A distortion adding device would have to be very complex in order to do this.

The sound that come out of out speaker is more comparable to vision than a photo as it contains extra degrees of information. A badly performing and low resolution amp baybe can be matched , but a higher resolution one would be far harder to match as it is not just distortions we are trying to match but accuracy.
Of course too amp that have the same distortion will sound the same.But we haven't yet got a grip on measuring distortions accurately.Nor do we also know which distrotion we should measure.

So going back to the stereophile test we can see that the system was not likely to be high resolution or it would have required the same high resolution to match it.What it shows is that some amps produce distortions that some people like, that can be matched more cheaply than you may think.

 

[ChrisPa]

I've joined this discussion late, but my twopenneth...

And remember that the objectivists of the adb-tendency have only ever said that ... the sound is explainable by standard engineering and there is no magic

that's the problem - "standard" engineering. What's standard engineering today is different to standard of 30 years ago. So the advancement of engineering comes from taking human observations and reapplying them to the engineering model; whilst being very careful not to dismiss human observations because they contradict the current engineering model

but have never said that the more accurate or flatter the frequency response/phase dispersion/whatever the better sounding the kit

possibly not in this forum/discussion, but (IME) that's not the general viewpoint of objectivists.

Of cource two things that measure the same will sound the same

Yes. It's then a question of which measurements to make.

Have you not read any of the hundreds of posts made on this forum claiming that measurements are pointless because the unmeasurable effects of DCT, umpteen 9s pure copper and various other varieties of fairy dust, etc. etc. etc. make so much more difference.

Of course the effects are measurable if they're there. If the "subjectivists" are right then the human hearing system is providing an effective measurement device - quite right. The question then is how to emulate the human measurement system with man-made measurement devices.

I do find it interesting that virtually none of the subjectivists have anything to say about this test - which seems to completely destroy their argument.

I don't think it does. Subjectivism (to my understanding) says that there are (clearly) audible effects that have not yet been measured by "objective" (ie. "artificial" or man-made) measurement devices.

Lee, I don't think the distortion in valve amps is irrelevant. On the contrary, it's probably the distortion they exhibit that makes some people prefer them over other amps. They may not be getting "accuracy" or "high fidelity" in the strict sense of the term but if they prefer it, who cares?

Generally agreed. At this level I feel it's much like preferring the tone of one violin versus another, or one guitar versus another (select your own musical instrument of choice), or one concert hall versus another.

What I don't agre with is this,that all differences in equipment are due to currently measurable distortions.

Agreed - "currently measurable" being the important words

And I don't think you can equalise the sound of equipment just by changing these distortions to match each other.

Disagreed - make the distortions match and I think you've done the Carver trick and CAN make 2 amps (note - amps, not necessarily anything else) sound the same.

Exactly at what resolution do we hear and exactly at what resolution is current measuring equipment?
If the resolution of measuring equipment is lower than the resolution of our hearing, then or current measurments are not good enought.
This is of cource assuming we are measuring for the right things in the first place.

I suspect it's not resolutuon per se. it's knowing the right parameters to measure.


Chris

 

[michaelab]

whilst being very careful not to dismiss human observations because they contradict the current engineering model

Absolutely, but human observations have to be shown to be real before they merit further investigation. That's the point of DBTs - to see whether these clearly audible effects are actually audible. That requires controlled conditions and enough trials to get a statistically meaningful result.

Of course the effects are measurable if they're there. If the "subjectivists" are right then the human hearing system is providing an effective measurement device - quite right. The question then is how to emulate the human measurement system with man-made measurement devices.

See above. So far no one has been able to demonstrate that these effects are actually there, so the question of man-made measurement devices not being as capable as the human ear becomes irrelevant.

Subjectivism (to my understanding) says that there are (clearly) audible effects that have not yet been measured by "objective" (ie. "artificial" or man-made) measurement devices.

And objectivism says: "prove it". Again, see points above.

It's quite possible that the human ear may be able to detect things that man made measurement devices cannot. However, with regard to differences between cables, no one has been able to show that this is the case in a controlled DBT.

Michael.

 

[GrahamN]

So the advancement of engineering comes from taking human observations and reapplying them to the engineering model; whilst being very careful not to dismiss human observations because they contradict the current engineering model

Agree 100% - and this was the essence of the long and boring stuff I deleted from the earlier post.

Of course the effects are measurable if they're there. If the "subjectivists" are right then the human hearing system is providing an effective measurement device - quite right. The question then is how to emulate the human measurement system with man-made measurement devices.

But a frequent "subjectivist" stance is that there are all sorts of things the ear/brain can detect that we cannot detect with test measurements (and which spxy is verging on saying with his stuff about resolution of hearing and measurement systems). Without having the figures to hand I can't quote them, but I'm pretty sure that there is experimental evidence showing that instrumental measurement resolution/dynamic range etc can be many times better than that of the human hearing system.

I would certainly agree though that (unbiased ) human hearing is the final reference, and the important thing is to develop a quantifiable engineering model that encompasses the salient points of the human hearing system.

I would also be happy to agree with anyone who argued that we have not yet developed an adequate interpretation of the measurements that have been made (or at least an adequate one that can be communicated to consumers in marketing data sheets). The important thing I see about the "Carver" test was that a simple match of outputs to below a level of -70dB or better (I assume measured on an oscilloscope or spectrum analyser) was shown to make two devices indistinguishable, with no "magic" required - and no "interpretation" of the measurements was required.

Subjectivism (to my understanding) says that there are (clearly) audible effects that have not yet been measured by "objective" (ie. "artificial" or man-made) measurement devices.

But I think they go a lot further than that - witness Saab's rabbitting on about emotional responses. If there is a lack of emotion in what is being listened to it is either a) the equipment is losing something in the original recording (lack of dynamics, poor frequency response, high distortion etc, all readily measurable) or b) the original recording didn't have it in the first place. Paul Ranson frequently make the point that adding colouration may be euphonious but it has nothing to do with hi-fi: I don't often agree with him, but I do in this case. I would probably prefer a good quality valve amp on poor source material, but the most accurate amp available on high quality, atmospherically recorded material (and possibly the balance would change depending on my mood too - sounds a good case for a high-accuracy amp and an effects-box!).

Agreed - "currently measurable" being the important words

Disagreed, but

I suspect it's not resolutuon per se. it's knowing the right parameters to measure.

agreed - that's basically my "interpretation" point above.

make the distortions match and I think you've done the Carver trick and CAN make 2 amps (note - amps, not necessarily anything else) sound the same.

But why shouldn't that apply to any component in the system. It would seem to me to be a pretty similar job for tuners, CD transports, DACs (all of which provide a readily comparable electrical output), possibly TTs (although that may well show that the essence of the differences/distortions do lie within the mechanical construction of the system). Speakers may be a tougher job to tweak (spatial dispersions coming in to play as well as the simple transfer-function-related stuff), although since the distortions they generate are so much larger there's a much greater opportunity for making those tweaks.

 

[LiloLee]

I would certainly agree though that (unbiased ) human hearing is the final reference, and the important thing is to develop a quantifiable engineering model that encompasses the salient points of the human hearing system.

I couldn't agree more. When we hear a system and say 'That sounds more open' what the hell do we mean by this and what is it that can be measured to indicate whether something is more 'open'?

An interesting thing to do, following the Carver item above, would be to then alter the Carver amp, by gradual increments until the audience could identify the difference consistently. Then we may have got some idea of the threshold at which things start to become relevant.

 

[Saab]

witness Saab's rabbitting on about emotional responses.

actually Graham,I mentioned once as an antidote for all the science I don't understand.I clearly don't know enough on the subject,so I will bow out.

 

[ChrisPa]

That's the point of DBTs - to see whether these clearly audible effects are actually audible. That requires controlled conditions and enough trials to get a statistically meaningful result.

I was hoping to have replied to this before now, but it deserves a long time spent on the reply, and I'm either working, singing or with family & friends all the way through to Christmas. I'll dedicate some time to a reasoned reply over the Christmas break.

So, for the moment, let me just say that I have serious doubts about the way most DBTs are conducted, and that generally I think they are used to justify a viewpoint which is just as blinkered/narrow minded as extreme subjectivists, without being used to advance engineering knowledge. I'll explain my thinking later....