What can measurements show/not show?

[jkeny]

As I said we don't have ALL the answers. If you for instance look on the work of Siegfried Linkwitz for the last few decades he has made a lot of progress to translated audible differences into measurable parameters into design parameters for things like soundstage and realism. His speakers show he's got it pretty well nailed.
It's a complex of factors, not just a single number. It's like asking for a single number that indicates how well a country is of financially. There is no single number, but you can come up with a set of factors that indicate the status.
Similarly, you can come up with a set of factors that will indicate which speaker will give better sound stage than another, and these factors can be measured. But you can't hang a multimeter off a speaker and say, aha, it's 4.6 and the other is 7.2 so this one has better soundstage.

jd

Again, I agree that there is no simple single figure but why would I submit a unit for testing in a test that has not been shown capable of discerning the effects that we are speaking of? Is this not doomed to failure? Why would anybody demand a unit to be tested & not be willing to show how good their tests were? Is it not a complete tautology?

Simple Analogy, you have developed a car that can travel at a fantastic speed. I don't believe you even though a lot of other people can testify to the fact that the car has travelled faster than anything they have ever seen. These are anecdotal & not measured tests. I ask you to submit your car for testing & proof. You first want to see that my test will be able to measure the high speed that your car can travel but I refuse to offer this proof. Instead I just insist that you produce the car for testing. Your reluctance to do so is meant to imply that you are afraid of being exposed & your car really doesn't do the speed you claim. Others now begin to try to say that you are bluffing. Your attempts to get some positive examples of the test showing it can measure at the speed you know your car travels are ignored!

What do you think?

 

[jkeny]

Sorry JK, you are just bluffing now. Your bluff has been called. Man up, or back out.
Anything else is just silly.

I'm not sure what you mean we are just getting to the point now so please don't close down this thread!

I don't know what bluff you mean - I have scope shots that will be published (done by a reviewer) but these are being rejected as unsuitable????. Instead analogue plots are being asked for.

I don't believe these analogue plots can possibly show the differences as they are not yet sophisticated enough. I'm asking for proof that they are sophisticated enough by seeing examples of such plots & the effects noted on them. These are not being produced. Instead a constant demand to make my units available for this test is made. As if this is a test that I'm afraid to undergo! Prove your test is valid & we are good to go. Simple

What do you think/would you do?

 

[jkeny]

JKeny: I think you are tilting at windmills. Or perhaps red herrings.

I, for one, do believe that if your modifications change the spdif signal in a significant way, then we will see the results in a measurement of the analog signal. It is the analog signal that matters, isn't it?

So why not measure it? You may be delighted with what you find. I did so on the DCX2496 and was quite pleased with the results.

Ok, I didn't see this post - can you show us the plots - this would be most interesting? This would be the first evidence produced!

 

[jkeny]

Deleted!

 

[SY]

15 seconds of googling. Really, why is this so hard for you?

http://www.analog.com/static/import...509029308534504114752208671024345AN_756_0.pdf

Stereophile: Jitter, Bits, & Sound Quality

http://www.thewelltemperedcomputer.com/Lib/Troisi.pdf

ScienceDirect - Measurement : Sampling clock jitter effects in digital-to-analog converters

http://www.nanophon.com/audio/jitter92.pdf

How many more references do you need?

 

[jkeny]

15 seconds of googling. Really, why is this so hard for you?

http://www.analog.com/static/import...509029308534504114752208671024345AN_756_0.pdf

Stereophile: Jitter, Bits, & Sound Quality

http://www.thewelltemperedcomputer.com/Lib/Troisi.pdf

ScienceDirect - Measurement : Sampling clock jitter effects in digital-to-analog converters

http://www.nanophon.com/audio/jitter92.pdf

How many more references do you need?

I've seen or read most of them before - are they measurements of analogue waveforms from the speakers showing jitter? I don't think so! Try again!

Surely if you're saying that you can measure the units you want me to send to you, you should have plots from your own measuring system showing jitter on the analogue out from speakers? This is what you are saying counts, right - the sound we hear?

Why resort to google?

 

[SY]

I've seen or read most of them before - are they measurements of analogue waveforms from the speakers showing jitter?

Moving the target? Your mods either change the analog signal or they don't. If you can't be bothered to demonstrate that, send me the units and I'll do it.

 

[Pano]

What do you think/would you do?

Me? I would offer up the units to be tested. Because I think that there would be measurable differences to be found.
The SPDIF signal is of much less importance to me, as I would not be able to easily interpret it. There are some on this forum who may be able to.

Why ask for proof of the measurement validity when no proof of the listening test validity has been shown? Why the double standard?

I will see if I can dig up my DCX tests. They were not jitter related, but an FFT of different output stages. (same DAC and digital sections).

 

[jkeny]

Moving the target? Your mods either change the analog signal or they don't. If you can't be bothered to demonstrate that, send me the units and I'll do it.

Here's what you said to me already "So if there's jitter, noise, whatever (and that's what you've claimed), these are MEASURABLE effects on the analog output stream. They are either there or they aren't. It will take you about 15 minutes to get some data."

& when rejecting the SPDIF measurements: "No, what I'm asking for is measurements of the results, that is, changes in the analog output. I don't listen to SPDIF signals, nor (I suspect) do you. If there are analog changes, then it's plausible that there could be audible differences."

I've consistently asked you, ad-infinitum to show analogue plots so no moving target here (often we are guilty of what we accuse others of - transference I think it's called?)

I believe my above posts expose your ruse & posturing - it's there for all to read! Goodbye!

 

[jkeny]

Me? I would offer up the units to be tested. Because I think that there would be measurable differences to be found.
The SPDIF signal is of much less importance to me, as I would not be able to easily interpret it. There are some on this forum who may be able to.

I bet the same lack of interpretation would apply to the analogue plots if they were ever produced!

Why ask for proof of the measurement validity when no proof of the listening test validity has been shown? Why the double standard?

I never said the listening tests proved anything other than consensus of opinion. No double standard here on my part - I have tests already - I don't need to submit to the demands of anybody else's test, particularly when they refuse to show how valid their tests are! Did you not read the simple analogy above? I would be happy to if they would show the validity of their tests - not forthcoming, I'm afraid.

I will see if I can dig up my DCX tests. They were not jitter related, but an FFT of different output stages. (same DAC and digital sections).

Great, would love to see them.

 

[Pano]

You are simply dodging the issue. I see no need to comment further.

 

[jkeny]

You are simply dodging the issue. I see no need to comment further.

Maybe this will satisfy you - To anybody who can demonstrate that their tests can show sound stage, jitter, timbre, sonic tails. I will send them a modified & unmodified unit for testing, OK? Simply what I have been asking SY for all along but if there is anybody else then lets do it!

 

[jan.didden]

Maybe this will satisfy you - To anybody who can demonstrate that their tests can show sound stage, jitter, timbre, sonic tails. I will send them a modified & unmodified unit for testing, OK? Simply what I have been asking SY for all along but if there is anybody else then lets do it!

You're mixing things up. Jitter can easily be measured, in fact is a standard type of measurement, both in the analog as well as in the digital domain.

Soundstage cannot be measured by a simple one-number thing - it depends on several factors, but each of those can be measured. Same with timbre, you can't say: "what is this equipment's timbre on the scale of 1 to 10?" but spectral analysis will reveal it quite well.

I'm pretty sure you very well know that, you're not stupid. So why ask this type of questions?

jd

 

[Netlist]

I took some spectrum shots of a female voice, a trumpet and a violin. You can easily see the 'timbre' of each instrument with their harmonics. I would think this is a nice example of what you want to see. All shots are from vinyl records.

 

[jkeny]

OK, Netlist, you''re in Europe, makes things a lot easier - do you want to do some measurement tests & some listening evaluation (nothing scientific expected here just what you hear!)?

 

[Netlist]

Thanks, honestly, my ears are too old and I have no gear to do proper tests.

 

[jkeny]

You're mixing things up. Jitter can easily be measured, in fact is a standard type of measurement, both in the analog as well as in the digital domain.

Soundstage cannot be measured by a simple one-number thing - it depends on several factors, but each of those can be measured. Same with timbre, you can't say: "what is this equipment's timbre on the scale of 1 to 10?" but spectral analysis will reveal it quite well.

I already said I agree that these are not single figure or even single plot measurements but nobody here has shown any plots whatsoever - so why so shy? All I wanted to see is that it's possible to measure these things, as you say it is.

Again, I was being asked to submit a unit for testing without any idea what the test could/couldn't show! Would you be interested in that offer?

I'm pretty sure you very well know that, you're not stupid. So why ask this type of questions?

jd

Explained above!

 

[jkeny]

Thanks, honestly, my ears are too old and I have no gear to do proper tests.

OK, anybody else in Europe with the equipment to do this? I'm in Europe & it makes customs issues so much easier (I don't want to have to pay shipping & customs duty charges).

 

[SY]

If you have a soundcard, then YOU have the equipment to do this. Really, that's all it takes.

 

[J.R.Freeman]

I already have measurements of the differences in SPDIF out before & after changes

And yet, you refuse to produce said measurements.

For those such as SY, who have graciously contributed to this thread their thoughtful replies, I would warn you of the possibility that this is a wasted effort. This thread was started under the guise of an honest discussion, but I suspect its true motivations include an intentional stalemate - deliberate break-down of reasonable communication, to make it appear that certain magical technologies can stand toe-to-toe with proven, scientific theories. It has also become apparent that commercial interests are involved. Take what you want from that!

I truly enjoy DIYAudio.com - great people, who are willing to help, and teach, for the most part. It is a shame to see this sort of thing, as it detracts from the honest, meaningful discussions one usually finds here.

JF