Computer Storage - Best Practices

[alfe]

13 hours ago, mansr said:

 It's also possible that he simply got lucky. Given the odds involved, I'd consider that the more likely explanation.

 

Or  may be a storage influence.

 

[mansr]

8 hours ago, STC said:

Too bad.  You took part and allowed the single test to proceed. It is no use now to cry that the methodology was invalid, especially when it was conducted by an objectivist. 

I never said it was invalid, although I would have liked for some aspects to have been better controlled. Given the surprising outcome, the next step should be an attempt to replicate the result in a more controlled environment. Until that happens, all we have is an unexplained anomaly.

[manisandher]

20 minutes ago, mansr said:

Given the surprising outcome...

 

Get this into your head - the positive ABX result was the expected outcome. I'm not in the habit of wasting my, or anyone else's, time. If I wasn't confident of a positive outcome, I wouldn't have invited you to my place... and paid for your train fare. Got it?

 

For the n^th time:

 

1. I was certain that I could hear differences between two different but bit-identical playback means

2. I was confident I could demonstrate this to you, and invited you over to my place to do just that

3. I scored 9/10 in a blind ABX test that you controlled (1% probability of achieving this by guessing alone)

 

Now stop sounding like a bloody flat-earther and accept the only logical conclusion to this series of events, i.e. there was an audible difference between the two bit-identical playback means.

[STC]

Where can I find the original thread? If I remember correctly, you couldn't identify in the first three attempts.  I don’t see why @mansr shouldn’t be skeptical. I do hope another round of testing will go on. 

[manisandher]

4 minutes ago, STC said:

If I remember correctly, you couldn't identify in the first three attempts.

 

No, that's not correct. Here's what we did:

 

Test 1:

A, B, X, X, X, X, X, X, X, X, X, X

 

Test 2:

A, B, X, X, X, X, X, X, X, X, X, X

 

In both of these, I scored ~5/10. I was totally aware that I was guessing. After the first X, I no longer had a reference. But these tests actually proved very useful - they proved that there were no 'tells' or 'cheats' going on.

 

Test 3:

A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X

 

I now had a reference before each X. I scored 8 straight correct answers, got #9 wrong and then #10 correct. This third test took ~20 minutes to conduct. That's quite a long time to remain totally focused, especially seeing as I was already quite tired from the first two tests.

 

Mani.

[STC]

5 minutes ago, manisandher said:

Test 3:

A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X, A/B/X

 

That reminds me of another blind test that you posted here. I too couldn’t tell the difference initially but had picked out them mostly correct under similar condition. 

 

This probably warrant further investigation. 

 

 

[sandyk]

Y

56 minutes ago, manisandher said:

now had a reference before each X. I scored 8 straight correct answers, got #9 wrong and then #10 correct. This third test took ~20 minutes to conduct. That's quite a long time to remain totally focused, especially seeing as I was already quite tired from the first two tests.

 It always pays to have a Reference file .

More than  10 Years ago, (back when I was originally sparring with Marce in DIY Audio ) I performed a series of 3 lots of 20 A/B comparison pairs  made by DIY Audio member Greg Erskine, which were created  using 20 pairs of original files of mine on a USB memory stick.
 He shuffled them around between his 2 HDDs using a special program he had written, and renamed all of them, as well as removing Properties information from them and saved them to the memory stick. 
 As I was having difficulty at the time with BP related issues. I then added a brand new reference using the original files on my PC.,making 3 copies of each, I was then more easily able to identify which was which, BUT neither of his 2 new copies sounded as good as the original REFERENCE file !!!

 As Ripley would say " Believe It or Not"

[manisandher]

52 minutes ago, STC said:

Where can I find the original thread?

 

I'll start a new thread with a full summary of the test, when I have some time.

 

32 minutes ago, STC said:

This probably warrant further investigation.

 

I'd be happy to take part in any endeavour that investigates of the mechanism that could be causing bit-identical replay to sound audibly different - a lot of good could potentially come out of this, to the benefit of everyone in this hobby.

 

But I have zero desire to take part in any further listening tests. To my mind, whether bit-identical replay can sound audibly different is firmly established. It can and does. Something that I'm certainly not enamoured with, but have come to accept, because it's simply the way things are:

 

"If we want to solve a problem that we've never solved before, we must leave the door to the unknown ajar."

[marce]

16 hours ago, numlog said:

you mean the 95% who parrot what a handful of actual researchers have proved for them over the years.

 

No I mean those of us who work in high end electronics doing signal integrity and critical layout, that's who... basically the whole of the electronics industry, apart from a few misinformed audiophiles...

 

So if bits are not bits what are they...

Oh jitter/timing is a critical part of signal integrity, that's what we use to ensure the BITS get to the destination in time...

 

[marce]

Also I ask Motberg to enlighten us as he seems to think "bits are bits" is funny, of course this digital phenomena you may have heard of is based on binary and bits being bits....

[sandyk]

5 minutes ago, marce said:

No I mean those of us who work in high end electronics doing signal integrity and critical layout, that's who... basically the whole of the electronics industry, apart from a few misinformed audiophiles...

 

Seems that you all still have a lot to learn then, and perhaps Rajiv's huge thread could be a good place to start !

[marce]

Just now, sandyk said:

 

Seems that you all still have a lot to learn then, and perhaps Rajiv's huge thread could be a good place to start !

No you have a lot to learn, I work with it in the real world with real results, not some meandering thread where people are playing and having a good time, but it is not research, it is hearsay.

[sandyk]

1 minute ago, marce said:

No you have a lot to learn, I work with it in the real world with real results, not some meandering thread where people are playing and having a good time, but it is not research, it is hearsay.

 

 Still ignoring the results of your so called "  Gold Standard" DBTs , eh ?

 Apparently they are only " Gold Standard" when they agree with your own  rigid " black and white" beliefs.

[STC]

8 minutes ago, manisandher said:

 

I'll start a new thread with a full summary of the test, when I have some time.

 

 

I'd be happy to take part in any endeavour that investigates of the mechanism that could be causing bit-identical replay to sound audibly different - a lot of good could potentially come out of this, to the benefit of everyone in this hobby.

 

But I have zero desire to take part in any further listening tests. To my mind, whether bit-identical replay can sound audibly different is firmly established. It can and does. Something that I'm certainly not enamoured with, but have come to accept, because it's simply the way things are:

 

"If we want to solve a problem that we've never solved before, we must leave the door to the unknown ajar."

 

Looking at the other thread by Peter and my own experiment with delays, I suspect there could be some anomalies when doing ABX with the same computer. Perhaps, the computer could be creating some difference when two files are chosen randomly. 

 

Try this experiment. Rip a track from CD twice and do a ABX with Foobar. If you are lucky you could hear the difference.    

 

 

 

 

[sandyk]

15 minutes ago, marce said:

So if bits are not bits what are they...

 

 Nobody appears to be saying that Bits aren't Bits, just that Checksums are not presently capable of revealing the differences that numerous members are now reporting.

[sandyk]

4 minutes ago, STC said:

Try this experiment. Rip a track from CD twice and do a ABX with Foobar. If you are lucky you could hear the difference.   

 

I have yet to hear a difference between 2 files ripped using the same Optical device and directly saved to the same folder of the same HDD/SSD, UNLESS something like Norton Security starts doing a System Check while ripping or another Processor intensive program is running. e.g. QuickPar when processing RAR files.

 You really need something much better than the lacklustre Foobar Comparator.

JRiver for example, when playing from System Memory  is way more revealing a player than Foobar 2K.

[manisandher]

13 minutes ago, STC said:

... I suspect there could be some anomalies when doing ABX with the same computer. Perhaps, the computer could be creating some difference when two files are chosen randomly.

 

This will be clear when I do the summary...

 

We captured the digital input to the DAC in real-time during the 3 tests.

 

I have 12 digital captures from the non-ABX test 1.

I have 12 digital captures from the non-ABX test 2.

I have 30 digital captures from the ABX test 3.

 

Analysis shows that all 54 digital captures are bit-identical. The DAC was receiving exactly the same bits on each of these 54 occasions. And yet I managed to identify A and B correctly 9/10 times in the ABX - a 1% probability of achieving this through guessing alone. (Not to mention that this was exactly what I expected to happen, hence the invite to Mans.)

[marce]

24 minutes ago, sandyk said:

 

 Still ignoring the results of your so called "  Gold Standard" DBTs , eh ?

 Apparently they are only " Gold Standard" when they agree with your own  rigid " black and white" beliefs.

What are you on about, what DBT's...

I am on about comments to a post yesterday, Not the listening test others are talking about...

[numlog]

3 hours ago, marce said:

No I mean those of us who work in high end electronics doing signal integrity and critical layout, that's who... basically the whole of the electronics industry, apart from a few misinformed audiophiles...

 

So if bits are not bits what are they...

Oh jitter/timing is a critical part of signal integrity, that's what we use to ensure the BITS get to the destination in time...

 

you are not really following the post correctly, it was about the chance of unsuspecting newbies finding  incorrect information and how 95% of general audio internet population are already spreading what a small amount of experts have told them.

 ''bits are bits'' '' it either works or it doesnt''  etc. are the typical phrases you will see thrown around a lot

[marce]

I think I get you now...

 

[numlog]

17 minutes ago, marce said:

I think I get you now...

 

ok...

[Ralf11]

7 hours ago, marce said:

No you have a lot to learn, I work with it in the real world with real results, not some meandering thread where people are playing and having a good time, but it is not research, it is hearsay.

 

careful - some of the anti-VAXers on this thread think science and engineering is a "disruptive comment"

 

you must never challenge the muffy army with facts

[Ralf11]

7 hours ago, marce said:

What are you on about, what DBT's...

I am on about comments to a post yesterday, Not the listening test others are talking about...

 

 Gold Standard DBTs is a stock phrase he has used for years

[abulr]

boy this was tough to follow...

ok, here's a SUSPECTING Newbie... where did the wise elderly end up?
I'm going Mac Mini as transport, USB out to DAC
If bits are bits then shouldn't matter that i do this from internal SATA HDD (and the only benefit of upgrading to SSD would be Speed)

are all those other posts about hearing benefits running OS off SD Card, storing music on separate bus from DAC output type, and trying to run the music off RAMdisk then moot if bits are just bits?

thanks in advance,

a.r.

[sandyk]

8 minutes ago, abulr said:

are all those other posts about hearing benefits running OS off SD Card, storing music on separate bus from DAC output type, and trying to run the music off RAMdisk then moot if bits are just bits?

thanks in advance,

 If it was that simple, then the 100s of members, and people worldwide  who have purchased the John Swenson designed Linear PSUs for their Mac Minis from Uptone have wasted their money.

 Numerous customer reports indicate otherwise.