Master Thread: Are measurements Everything or Nothing?

[nagster]

Have you seen the scale of that graph?

yes.

0.1dB or even 0.2dB (the biggest deviation from flat in this graph) is not hearable by humans.

It depends on the person and the case. Not common to all humankind.
The production side controls less than 0.1dB, but the listener side does not need that strictness.
That said, my post is about measured differences, not audible differences.

 

[nagster]

Cool.
Is that AC coupled?
Would be helpful to see full input and analysis parameters.

Yesterday's post was a graph that had measurements from a few months ago, and it was a AC and relative value, but in order to make it more correct, but I tried again with DC and made it an absolute value display.

 

[Ken Tajalli]

Yesterday's post was a graph that had measurements from a few months ago, and it was a AC and relative value, but in order to make it more correct, but I tried again with DC and made it an absolute value display.View attachment 212290

Zooming in to 0.1dB level, maybe irrelevant to subjective results, but at least we can see those filters in action within audio.
Showing a graph where all filters overlap, is not informative, unless it includes wider bandwidth to 50kHz to see after effects of such filters.
This is an informative graph, thank you. A zoomed out wideband graph of the same, completes the picture.

 

[nagster]

Zooming in to 0.1dB level, maybe irrelevant to subjective results, but at least we can see those filters in action within audio.
Showing a graph where all filters overlap, is not informative, unless it includes wider bandwidth to 50kHz to see after effects of such filters.
This is an informative graph, thank you. A zoomed out wideband graph of the same, completes the picture.

Wideband is another measurement, and measurement by AP has already been published, so take a look.

D90SE Fully Balanced DAC -TOPPING

www.tpdz.net

Topping D90SE Review (Balanced DAC)

This is a review and detailed measurements of the Topping D90SE flagship balanced USB DAC. It was sent to me by the company for testing and I believe costs US $899. The D90SE comes in the same clothes as current class of Topping products which is just fine with me: We have the usual inputs...

www.audiosciencereview.com

Measurements of Topping D90SE/LE DAC - L7Audiolab

台式设备 两个月前，我更新了站里的DAC SINAD Chart，在天梯图的前面出现了几个莫名其妙的测量结果。这是什么妖魔鬼怪……乃至于都被搬运到ASR……好吧，无论如何NDA已经过期了…… 今天就让我揭晓…… Topping D90SE （工程机）的测量结果吧……

www.l7audiolab.com

 

[MattHooper]

QUESTION:

How would you go about blind testing various audio gear?

We all appreciate the place blind testing can have in vetting claims about audible differences.

The problem is that, in practice, some types of equipment can be quite clunky to try and blind test. Especially if we are trying to have the ability to quickly switch between the components.

So I'm thinking first of expensive audiophile AC cables vs cheap, competently built/spec'd "off the shelf" cables.

If you are trying to blind test for the purported sonic effects of different AC cables on an amplifier, it would require powering down the amp and physically swapping the cables for each change. Pretty clunky.

One way I can see making for a smooth blind test of expensive AC cables would be to have two of the same unit of a DAC. Feed the analog output of each DAC to a preamp/switcher that has at least two inputs, and then one of the DACs has a "cheap" or off the shelf AC cable, the other with the fancy expensive AC cable.
Then the listener can quickly switch between the two source inputs under blinded conditions. (Audiophiles of course claim high end AC cables "work" with virtually anything, including improving DAC signals). So you can test: does the signal coming from the DAC with the fancy AC cable sound different or not?

Not sure how to do this for amps. Maybe if you had two sets of the same amplifier, same thing - one uses the cheap AC cable, the other the expensive AC cable - then have a switch box switch between each amp powering the speakers, for quick switching?

Any of that make sense...or are there any other ideas or ways you've blind tested equipment?

 

[Robbo99999]

QUESTION:

How would you go about blind testing various audio gear?

We all appreciate the place blind testing can have in vetting claims about audible differences.

The problem is that, in practice, some types of equipment can be quite clunky to try and blind test. Especially if we are trying to have the ability to quickly switch between the components.

So I'm thinking first of expensive audiophile AC cables vs cheap, competently built/spec'd "off the shelf" cables.

If you are trying to blind test for the purported sonic effects of different AC cables on an amplifier, it would require powering down the amp and physically swapping the cables for each change. Pretty clunky.

One way I can see making for a smooth blind test of expensive AC cables would be to have two of the same unit of a DAC. Feed the analog output of each DAC to a preamp/switcher that has at least two inputs, and then one of the DACs has a "cheap" or off the shelf AC cable, the other with the fancy expensive AC cable.
Then the listener can quickly switch between the two source inputs under blinded conditions. (Audiophiles of course claim high end AC cables "work" with virtually anything, including improving DAC signals). So you can test: does the signal coming from the DAC with the fancy AC cable sound different or not?

Not sure how to do this for amps. Maybe if you had two sets of the same amplifier, same thing - one uses the cheap AC cable, the other the expensive AC cable - then have a switch box switch between each amp powering the speakers, for quick switching?

Any of that make sense...or are there any other ideas or ways you've blind tested equipment?

(Is there really any point in blind testing cables though, we already know they measure the same for frequency response & distortion whenever Amir has measured them.)

Good general point about how to blind test equipment though, I think it would be quite difficult. Probably easier if you've got some other people willing to help you for a while who are doing the switching, but it's not straight forward......all variables would need to be controlled as possible, I'll let someone who's done it before flesh out advice on the details, and it would vary with what piece/type of equipment you would be testing.

 

[SuicideSquid]

QUESTION:

How would you go about blind testing various audio gear?

We all appreciate the place blind testing can have in vetting claims about audible differences.

The problem is that, in practice, some types of equipment can be quite clunky to try and blind test. Especially if we are trying to have the ability to quickly switch between the components.

So I'm thinking first of expensive audiophile AC cables vs cheap, competently built/spec'd "off the shelf" cables.

If you are trying to blind test for the purported sonic effects of different AC cables on an amplifier, it would require powering down the amp and physically swapping the cables for each change. Pretty clunky.

One way I can see making for a smooth blind test of expensive AC cables would be to have two of the same unit of a DAC. Feed the analog output of each DAC to a preamp/switcher that has at least two inputs, and then one of the DACs has a "cheap" or off the shelf AC cable, the other with the fancy expensive AC cable.
Then the listener can quickly switch between the two source inputs under blinded conditions. (Audiophiles of course claim high end AC cables "work" with virtually anything, including improving DAC signals). So you can test: does the signal coming from the DAC with the fancy AC cable sound different or not?

Not sure how to do this for amps. Maybe if you had two sets of the same amplifier, same thing - one uses the cheap AC cable, the other the expensive AC cable - then have a switch box switch between each amp powering the speakers, for quick switching?

Any of that make sense...or are there any other ideas or ways you've blind tested equipment?

Honestly, I've just accepted that aside from speakers, the differences between properly-designed home audio equipment are so small that even if I could detect a difference in blind testing, I just don't really care at this point. I buy the stuff that measures well and has the features I want at a reasonable price and am done with it.

But for recording equipment, I make recordings with whatever gear I need to test, clearly label everything (say, four mics in a mic shootout), then ask my wife to name everything "A/B/C/D" and record in a separate file which letters correspond to what gear. Then I can compare without knowing which is which, make my notes, and then look up what equipment is what. This method works (and is very revealing) for most recording equipment.

 

[MattHooper]

(Is there really any point in blind testing cables though, we already know they measure the same for frequency response & distortion whenever Amir has measured them.)

From my own point of view I certainly agree.

But there is still a lot of "measurements don't tell you everything, we can hear things you can't necessarily measure" pushback, so blind testing is one way to address that
claim. I know, I know...those same people are likely to reject blind testing as well. But, still, blind testing is a part of the tool kit for investigating claims which is why I'm asking about solving the problem of practical blind tests on various types of gear.

 

[Newman]

For speakers, see the setup in papers by Toole and Olive.

For electronics, record the outputs and set up a level-matched ABX using the files.

 

[MattHooper]

For speakers, see the setup in papers by Toole and Olive.

For electronics, record the outputs and set up a level-matched ABX using the files.

Thanks.

If one is testing, say, the purported sonic effects of an expensive AC cable on an amplifier vs a generic AC cable, how do you suggest to record the output?

 

[Newman]

I assume you are talking about mains power cables? (Speaker cables and analog interconnects are “AC cables”, hence my point of clarification). In that case I would record the output of the amp.

 

[MattHooper]

I assume you are talking about mains power cables? (Speaker cables and analog interconnects are “AC cables”, hence my point of clarification). In that case I would record the output of the amp.

Do you mean record the musical signal from the output of the amplifier? I've never heard of someone doing that (since it's usually just leads sent to drive the speakers).

Could you be specific about exactly how you'd do the recording, for example, from this Bryson amp:

4B³ – Bryston

bryston.com

 

[Newman]

I’m struggling to see the issue? Run cables from the amp’s speaker terminals to your recording device, set the music signal level from the amp to be compatible with the recording device, and go.

 

[MattHooper]

I’m struggling to see the issue? Run cables from the amp’s speaker terminals to your recording device, set the music signal level from the amp to be compatible with the recording device, and go.

No normal recording devices are made to take speaker level inputs.

That video explains what most folks in the recording end of audio would tell you: if you send a speaker level input to a mic, instrument or line level input "can damage the device." "Make sure you only connect speaker level outputs to speaker level inputs." I just talked to some folks at a pro sound store and they said "ah, yeah...you REALLY don't want to do that." Nor did they offer a way to do it.

That's why you'd be hard pressed to find any examples of anyone actually recording a music signal from the amplifier output.

So since I've never done this, nor has anyone else I know, and I no doubt lack some of your technical knowledge, could you be more specific: "Set the music signal level from the amp to be compatible with the recording device" part could use some more detail.

Thanks.

 

[BDWoody]

So since I've never done this, nor has anyone else I know, and I no doubt lack some of your technical knowledge, could you be more specific: "Set the music signal level from the amp to be compatible with the recording device" part could use some more detail.

Any reason a super simple voltage divider wouldn't work?

Voltage Dividers - learn.sparkfun.com

 

[Newman]

@MattHooper It’s perfectly safe as long as you match levels, which is what I mean by your bold text above.

If your recording device has, say, a maximum input voltage of 2 volts RMS, then make sure the power amp isn’t putting out more than 2V RMS at its speaker output.

A simple, fail-safe* way is to set the power amp to a gain of 1. That is, play a sine wave through the DAC, measure its voltage output from the DAC outputs, then connect it to the power amp, and adjust volume until the power amp has the same voltage output at its speaker output.

* I suppose, technically, if the DAC connected to the power amp has so much output voltage that it could damage your recording device by itself, then this isn’t as fail-safe as I thought. But seriously, in that case the DAC could technically damage the power amp too…this would be very unusual.

@BDWoody the level matching I describe achieves the same thing more simply.

 

[LTig]

@MattHooper It’s perfectly safe as long as you match levels, which is what I mean by your bold text above.

It's perfectly safe in theory, but I would never ever connect an expensive ADC directly to any source which is capable to emit higher voltages than the ADC can safely handle. One unwanted spike and the ADC is gone. Therefore one should always use a proper voltage divider, possibly including a pair of zener diodes across the ADC input.

 

[MattHooper]

Yeah, these are all the messy details I was figuring. No doubt part of why nobody has done such tests.

Of course the paucity of blind tests - like for high end AC cables - can simply come down to the fact those who believe the cables make a difference aren't compelled to bother blind testing, and those who believe they don't make a difference (justified on technical grounds) are similarly not compelled to bother with the testing.

Which is too bad.

(I'd certainly like to see more "subjectivists" especially engaged in blind tests).

 

[Vive la dissonance!]

In sighted tests, I couldn't hear any difference between cheap and expensive interconnects, nor between cheap and expensive power chords, so I never felt any reason to do blind tests. BTW, I was looking to hear a difference, but did not... so, I mean, why bother?

It is those with the extraordinary claims going against understood scientific principles who should be presenting valid evidence.

I admit to have bought slightly more expensive cables, just because they look slightly nicer to my eye. But I am not under any illusion that they improve my system sound-wise, compared to the stock cables.

 

[Andysu]

syn08 seemed to be complaining of audiophiles treating stereo like it could be a "hologram" suggesting a total immersion with height as realistic as depth etc. The movie screen analogy shows that we haven't deemed this necessary to accept the description "3D" to movie content, nor do we need to with stereo content. Both can be constrained by their respective "windows" on to the illusion, yet still have characteristics of depth and image placement and environment, to use the short hand "3 dimensional."

3d movies suck . just a cash cow to lure the sheep in . i followed one such for "titanic 3d" it sucked . threw away the glasses and watched blurred . only went see if the sound mix was still as same when i projected it at warner bros village 12 plex 1998 . digital cinema rubbish . i saw colour banding noise at the start of the hard-drive-movie . this wasn't film it wasn't 35mm i once recall .