Why Do People Come To Computer Audiophile To Display Their Contempt For Audiophiles?

[mansr]

1 minute ago, Albrecht said:

""The buffer stores digital sample values as received over the wire. There is no mechanism by which it could store anything else, including noise.""

However much ground loop noise, & poorly timed those digital samples are....

The buffer stores only the sample values, not their times. The timing is inferred from the known sample rate.

1 minute ago, Albrecht said:

""A saturated link is just as reliable as a barely used one.""

Reliability is not the issue, - timing is.

What timing?

[plissken]

3 hours ago, Albrecht said:

No, - the noise "could" be carried into the buffer.

 

 

No it can't. Period, end of discussion. You have an awfully steep hill to climb if you thing a file stored on a hard drive or other static storage is encapsulating any noise.

[plissken]

2 hours ago, Jud said:

Blinding the audition certainly makes it better at determining who's the superior player regardless of gender.  But there are other qualities you want in a new hire besides gender neutrality, and so there are better and worse blind audition setups to get the results you want.

 

It's only ONE metric. The are controlling for THAT metric as best they can. That metric may or may not be combined with others in a final hiring decision.

 

I wouldn't know what else gets considered but I'm sure there are other traits.

 

Maybe they have a history of not showing up on time for rehearsal. Maybe they are some sort of diva to deal with. Maybe they like to pick their nose with their bow.

[semente]

1 hour ago, Ralf11 said:

 

Floyd Toole and others have proven it works.

 

What is your def'n of "work"?

 

Toole's listeners were comparing loudspeaker frequency response and if you look at the measurements he published you'll see wildly varying response between samples.

Big differences are easy to spot.

[pkane2001]

1 hour ago, mmerrill99 said:

Not sure what buffers are being talked about here but does the input to & output from buffers happen concurrently or does the buffer get filled & then filling is turned off while emptying happens?

Just a consideration about the electrical noise happening while filling & emptying are happening & whether the concomitant electrical noise can have any subsidiary effect?

 

In isochronous USB audio, one packet consisting of many samples is sent every 125µs. These packets have to be stored in a buffer, so that they can be doled out to the DAC at an appropriate sampling rate. For an asynchronous USB interface, this output rate is controlled by a separate clock. The quality of this clock dictates the timing (and jitter) of the output to the DAC, as long as the incoming data does not cause buffer overruns or underruns. When buffer under/overruns do occur, you'll hear obvious clicks and drop-outs.

 

Ground loops and leakage currents can potentially affect the accuracy of the clock which might result in jitter, but this is not related to the buffer. The buffer is not the mechanism for these effects to sneak through to the DAC input.

[Lebouwsky]

3 hours ago, Sal1950 said:

Last I understood, this "hobby" was about obtaining High Fidelity

Not just "sounds good to me"

Who told you that? If you can led that go you'll notice your "hobby" wil change in an enjoyable hobby (no quotation marks)

 

[semente]

11 minutes ago, Lebouwsky said:

Who told you that? If you can led that go you'll notice your "hobby" wil change in an enjoyable hobby (no quotation marks)

 

 

Why would you think that someone pursuing accuracy will not enjoy himself, his music or the sound he is getting from his system?

[esldude]

31 minutes ago, semente said:

 

Toole's listeners were comparing loudspeaker frequency response and if you look at the measurements he published you'll see wildly varying response between samples.

Big differences are easy to spot.

Small ones are not that hard either.

 

When something sounds really different I start trying to figure out why.

 

Item number one to check is level.  Lots of audible differences end up being a level difference in playback.  Get things within a .1 db on level, and many differences go away.

 

Item number two is frequency response.  Blind listening to files of gear I have picked out devices that have a 1/4 db or half db difference spanning a couple of octaves.  Get much closer than this and I can't hear it.  Hearing 1/4 db is rather tough, and some days I can't do that.  

 

After I do these two things it is rare genuine differences that can be perceived are left, and the most common one is an overload or high distortion on peaks. 

 

Now what I do isn't scientific research, but it is repeatable blind.  If you practice blind listening you become comfortable with it though boring and tedious it will be.   If some gear gets thru these checks what is left, could be audible, but it is a very small difference or only occurs under unusual specific conditions. 

 

The short cut is to quit being taken in by all the phoolery.  A good well done set of measurements for basic level, FR, distortion and noise might not catch everything. What is left is small and of little consequence at best.  One can probably find a few unusual things that might be missed by this approach, but it will be a rarity.

[Ralf11]

44 minutes ago, semente said:

 

Toole's listeners were comparing loudspeaker frequency response and if you look at the measurements he published you'll see wildly varying response between samples.

Big differences are easy to spot.

 

I think you are either talking about a different study or have missed the point.

[esldude]

17 minutes ago, Lebouwsky said:

Who told you that? If you can led that go you'll notice your "hobby" wil change in an enjoyable hobby (no quotation marks)

 

Saw this after my last post.  Read it.  This other approach to simple definable fidelity without the smoke and mirrors and mystic magic is so much easier, satisfying and calming.  You cross your t's and dot your i's and move on.  No magical suspicious activity required. Much more time to simply enjoy music.

[plissken]

46 minutes ago, semente said:

 

Toole's listeners were comparing loudspeaker frequency response and if you look at the measurements he published you'll see wildly varying response between samples.

Big differences are easy to spot.

 

I agree. When someone says 'Night and Day', 'Easy to Discern', 'Readily Apparent' those are easy to spot. Blinded.

[semente]

3 minutes ago, esldude said:

Small ones are not that hard either.

 

When something sounds really different I start trying to figure out why.

 

Item number one to check is level.  Lots of audible differences end up being a level difference in playback.  Get things within a .1 db on level, and many differences go away.

 

Item number two is frequency response.  Blind listening to files of gear I have picked out devices that have a 1/4 db or half db difference spanning a couple of octaves.  Get much closer than this and I can't hear it.  Hearing 1/4 db is rather tough, and some days I can't do that.  

 

After I do these two things it is rare genuine differences that can be perceived are left, and the most common one is an overload or high distortion on peaks. 

 

Now what I do isn't scientific research, but it is repeatable blind.  If you practice blind listening you become comfortable with it though boring and tedious it will be.   If some gear gets thru these checks what is left, could be audible, but it is a very small difference or only occurs under unusual specific conditions. 

 

The short cut is to quit being taken in by all the phoolery.  A good well done set of measurements for basic level, FR, distortion and noise might not catch everything. What is left is small and of little consequence at best.  One can probably find a few unusual things that might be missed by this approach, but it will be a rarity.

 

I do believe that listening is as important as measurements, particularly with speakers, and blind makes all the sense to remove bias.

It's the A/B switching comparisons that I don't find particularly useful because I as say they are boring and tiring, and also because I end up focusing on different things on second listenings.

Lately I just play music that I know will reveal problems or present specific challenges: I don't listen so much to the music but instead focus the sound.

 

Foo is a big problem in audio but to me it's lost the attraction of those early years when I read magazines.

I am also quite immune to looks, finish and price, to trends and safe buys.

I just try to combine observational listening with whatever measurements I can find.

Unlike most, I can hardly ever find any useful information in reviews.

[Jud]

1 hour ago, plissken said:

 

It's only ONE metric. The are controlling for THAT metric as best they can. That metric may or may not be combined with others in a final hiring decision.

 

I wouldn't know what else gets considered but I'm sure there are other traits.

 

Maybe they have a history of not showing up on time for rehearsal. Maybe they are some sort of diva to deal with. Maybe they like to pick their nose with their bow.

 

Yup, we are in agreement.

[Jud]

3 minutes ago, semente said:

Unlike most, I can hardly ever find any useful information in reviews.

 

In “official” reviews, and anywhere else that has the Lake Wobegon effect (“where all the women are strong, all the men are good-looking, and all the children are above average”), I agree with you. Only those who have a variety of opinions are helpful if I am looking for subjective input.

[Don Hills]

3 hours ago, mansr said:

I listen to veils. I especially like the sound of them being lifted.

 

I prefer to watch them being lifted.

[jabbr]

2 minutes ago, Don Hills said:

 

I prefer to watch them being lifted.

I prefer skirts

 

[mmerrill99]

1 hour ago, pkane2001 said:

In isochronous USB audio, one packet consisting of many samples is sent every 125µs. These packets have to be stored in a buffer, so that they can be doled out to the DAC at an appropriate sampling rate.

There are spikes at 8KHz (125uS) seen in some measurements - are these spikes due to electrical noise from the buffer getting samples every 125uS or from the USB receiver handling these packets or both?

 

As far as I know, anytime there is bursty electrical activity (packet processing) the electrical noise generated will also be bursty & may effect other sensitive areas i/e clock, Dac

[semente]

4 minutes ago, jabbr said:

I prefer skirts

 

Patience is a virtue 

[mansr]

16 minutes ago, mmerrill99 said:

There are spikes at 8KHz (125uS) seen in some measurements - are these spikes due to electrical noise from the buffer getting samples every 125uS or from the USB receiver handling these packets or both?

The noise is from the USB interface handling the incoming frame. The subsequent buffer has nothing to do with it.

[pkane2001]

17 minutes ago, mmerrill99 said:

There are spikes at 8KHz (125uS) seen in some measurements - are these spikes due to electrical noise from the buffer getting samples every 125uS or from the USB receiver handling these packets or both?

 

As far as I know, anytime there is bursty electrical activity (packet processing) the electrical noise generated will also be bursty & may effect other sensitive areas i/e clock, Dac

 

Most likely due to electrical crosstalk between the two circuits. This shouldn't happen in a well-designed interface. If the path from the USB input is not properly isolated, the 8KHz signal can potentially leak all the way to the DAC output.

[mmerrill99]

27 minutes ago, mansr said:

The noise is from the USB interface handling the incoming frame. The subsequent buffer has nothing to do with it.

OK, but is the buffer filled in a bursty manner & does this not result in bursts of electrical noise, not necessarily at 8KHz ?

[mansr]

2 minutes ago, pkane2001 said:

Most likely due to electrical crosstalk between the two circuits. This shouldn't happen in a well-designed interface. If the path from the USB input is not properly isolated, the 8KHz signal can potentially leak all the way to the DAC output.

In my good ADC (Tascam UH-7000) there is some 8 kHz noise visible if the pre-amp gain is turned up high. With standard line level inputs, this amount of gain is not needed.

[mmerrill99]

5 minutes ago, pkane2001 said:

 

Most likely due to electrical crosstalk between the two circuits. This shouldn't happen in a well-designed interface. If the path from the USB input is not properly isolated, the 8KHz signal can potentially leak all the way to the DAC output.

Ah yes "well designed" is a great phrase which I always see mentioned but seldom do people give the specifics of what it means.

just because we don't see this 8KHz noise spike on the outputs of the DAC doesn't mean it's not affecting other chips/clocks & intermixing with their own self-noise.I wouldn't subscribe to such a linear view as you seem to take.

[mmerrill99]

14 minutes ago, mansr said:

In my good ADC (Tascam UH-7000) there is some 8 kHz noise visible if the pre-amp gain is turned up high. With standard line level inputs, this amount of gain is not needed.

Does this mean that the 8KHz noise may affect low level signal linearity in your Tascam?

[fas42]

6 hours ago, semente said:

The way I see it, system performance can only be achieved if we can recognise shortcomings, identify possible causes and look for replacements that improve on what we currently have.

This requires unbiased or emotion free listening performed from an observational perspective with adequate methodology supported by measurements.

Upgrading or system building through "tasting" is a trial and error affair and good results are accidental.

 

Spot on! What I have noted is that a remarkable number of audio enthusiasts seem to be totally oblivous of the fact that the system in front of them is outputting very high levels of obvious anomalies - they seem to have learned to shut down their hearing to these problems. This is somewhat apparent when they mention that no-one else in the house is interested in listening ...