Do you hear what I hear (bit perfect files sounding different)?

[PeterSt]

37 minutes ago, Currawong said:

I have a lot of trouble with this, as would most people. 

 

That part would not be about the file data, but about the program itself. Think of compiler optimization. This really exists ...

... And could vary for result over time and influences.

 

@sbgk (did I quote the correct nick here ?) was all about this, with 10 versions a day, a year in a row. 

I myself ever back started out with this, but this was too much out of control to proceed on.

[PeterSt]

15 hours ago, idiot_savant said:

I’m sure you and your friends are picky, but you can’t be suggesting that the electromagnetic interference continues from here? Or am I being an idiot again?

 

You say you have an EE background, so I’ll patiently await your response,

 

Like in one of the other threads, I would this time like to literally see a quote from the author where there's talk about "electromagnetical interference" in relation to page file hammering. My hint : it does not exist.

In that case there's also no requirement to lay this out, right ?

 

Personally (might that be important): I don't see any EMI by measurement, although that may not tell enough (it is too un-directional to measure properly with what I have for it). However, would I be the creator of such software, I might not be knowing what is happing (nobody does) and I would/could call it EMI influence.

 

Fact is, that the more years I experience with the ^2 and ^3 cables, the more I know that most of the influence goes through air. This works from inside of the cable to the outside, but also from the outside to the inside of the cable. It goes as far as the cable influencing itself.

Now, might you happen to know a device which can picture this for me (as in light through smoke), kindly let ne know and I may obtain it. It will save a lot of trouble and the progression in audio may speed up quite a bit (hopefully).

 

The real message of this post:

1. don't combine elements of which it is not known whether they combine in reality (EMI vs PageFile);

2. the author himself most 99% probably won't know what is happening BUT it happens for result.

 

Ad 2.: Don't underestimate this and know that only after decades this may have become science.

[PeterSt]

28 minutes ago, idiot_savant said:

secondly : are you seriously as a manufacturer of cables and equipment coming on to a public forum and saying you have no idea how they work or why they are beneficial

 

Sure. That's research life.

1. One grabs a couple of 1000 of ears ...

2. etc.

 

29 minutes ago, idiot_savant said:

thirdly : as a manufacturer selling product in the EU, don’t you need to do EMC testing as part of CE? Couldn’t you use this time in the test chamber to get some measurements to support your theories?

 

The idea is great, but the CE certification is explicitly left out. Proof, page 14 of the specs guide of our D/A converter (Dec 2010):

 

 

 

34 minutes ago, idiot_savant said:

saying you have no idea how they work or why they are beneficial?

 

These are your words, which kind of twisted mine. You should not do that (unless you want threads to close down).

You might quote me on these things, however.

I don't know how many dozens of pages I spent on the subject. And what I can't help is that it is my subject. Thus, food for rough ? (I hope this is English).

 

38 minutes ago, idiot_savant said:

firstly, we’re talking about this page-file thrashing being done at a different time to playback ( optimising a file ). So my point still stands - even *big if* this made some difference to the file how does this affect *anything* after it’s been read?

 

No. More firstly, this is your personal conclusion while it is also your personal observation.

Or am I wrong on this one ?

 

Still ... (but here I go again) ... Shall I show you the code in 10+ versions which tries to tame Windows memory and garbage collection ?

Shall I ? It took me 2 years to master it.

Shall I count for you the number of times the term "Page File" occurs in my software ?

I could still say "you know your stuff" (I did so two days ago). But it looks that you must be careful not to use it inadvertently.

 

 

 

 

 

 

This is in a loop with several other stuff preceding it. One small example:

 

 

 

[........]

 

 

 

 

 

 

The latter is a test on the cpu cycles still used on GC.

 

So like me, you better believe too that these matters exist.

Can you now (or anyone) please stop to be suspicious ?

 

On another note, it looks like Eric got hold of my sources. ;-)

 

Anyway, there you have your 2 minute process, possibly (!).

Also, don't underestimate how the first line in the first screenshot relates ("Straight Contiguous Memory"). You can guess it by now, Yes ?

Now don't ask me how smart this Eric is. But in my eyes everybody could be smart, so why not. If they only don't steal ...

---------------------------

 

All what remains is the viability of such applications. Like I said, XXHighEnd is 250K+ lines of VB.net. This one (de Audio Engine itself) is 60K lines of C++ code. The lines that play audio are 16 lines only. And that's 16 instead of 1 because of it sounds better.

But for the viability I myself listen to users and not to myself (or initially I will and when wrong it's not going to the public). All people have a say and the common denominator(s) win. Fix it, and up to the next idea.

With cables no different, as long as we don't understand the processes (and lack the measurements).

[PeterSt]

46 minutes ago, Confused said:

@PeterSt - Have you tried the Jinilabs player or one of it's file optimizations?

 

No. And to be honest, I don't do these things because it would feel like stealing for me. I explicitly don't want to be like that.

But as you just saw, I can guess a lot and I believe all.

 

One thing interests me: Someone quoted from the manual that a copy of the file was made to disk. This asserts that that file is being played, right ? (XXHE works by this means too, never mind after that it goes to memory as well);

Why did I not see anyone diving further into this ?? ... All I see is that the file remains unchanged. OK. So what about that copy then ?

 

PS: In my case the checksums will remain exactly the same. But the organization ... far from it. Remember, before it goes to memory. And all has its reasons.

 

Fun (for those who were there): This is how the JPlay guys split from XXHighEnd and started their own version of it (Marcin and Josef) with XXHE features in the base. We (mainly Josef and me) disagreed on how which means of file/memory approach would sound for the better. ... This is all quite high snuff ...

[PeterSt]

2 minutes ago, idiot_savant said:

I’m talking about this “optimising of a file”

 

I am sorry, but not clear much because you referred to the page file. And otherwise, please quote because I apparently now don't know what you are referring to. By now one thing:

 

2 minutes ago, PeterSt said:

PS: In my case the checksums will remain exactly the same. But the organization ... far from it. Remember, before it goes to memory. And all has its reasons.

 

I won't elaborate on that one. But trust me, it happens (for today the viability is up to listeners).

[PeterSt]

4 minutes ago, idiot_savant said:

I’m talking about this “optimising of a file”

 

Works for an SSD too. Please trust me, even without elaboration.

 

Remember what I just blathered about Contiguous Memory. Got the hint ?

[PeterSt]

47 minutes ago, fas42 said:

Overplaying it? I don't care what the source is of a sound degrading element in a system

 

As long as you understand that a signal is not noise. 🙂 Upsampling that thus also does not remove noise. Unless, unless you get way more specific into aliasing and (non-)analogue filters and what not. And I have never seen anyone calling aliases "noise" as such.

Noise is a too common term to use in these cases.

 

51 minutes ago, fas42 said:

I don't care what the source is of a sound degrading element in a system

 

Sure you care. Otherwise you can't attack it.

 

Think like this: you won't arrange for reconstruction of Redbook with a ferrite ring around your cables. Similarly you won't get rid of coupled-in oscillator noise by means of upsampling.

[PeterSt]

36 minutes ago, idiot_savant said:

none of this helps make bit-identical files sound different

 

You only think it does not. But it does wildly.

Could you be a tad more neutral on this one (I recall that you told you wanted to know about this all). The first thing you should do is try it out, agreed ?

[PeterSt]

51 minutes ago, idiot_savant said:

Don't forget, the claim here is that two bit identical files played back under identical conditions sound different.

 

Fine and agreed. But your response was to @Summit's post and that had its own context.

[PeterSt]

54 minutes ago, idiot_savant said:

the claim here is that two bit identical files played back under identical conditions sound different. Same playback software, same settings.

 

- same buffer sizes.

 

1 hour ago, idiot_savant said:

OK, so I said I wasn't going to comment on buffers, but...

 

There are numerous types and implementations of buffers, but to be clear we're talking about the memory kind, not the voltage translation kind, right?

 

1 hour ago, idiot_savant said:

 

Now, you also have the kind of buffers @Summit is talking about, which is typically a PC application feeding "audio buffers" - in this case the sound device is pulling audio from a buffer to play it in "the real world",

 

[... etc. ...]

 

1 hour ago, idiot_savant said:

none of this helps make bit-identical files sound different

 

So this sure does not help the conversation to be clear.

[PeterSt]

1 minute ago, idiot_savant said:

@yamamoto2002- the claimant is using an SSD,

 

Is he ?

 

1 minute ago, idiot_savant said:

the player reads the whole file into RAM before starting playback.

 

If you suggest it is played from RAM ... maybe. But you are guessing that. Abd if you are not guessing that, the "claimant" is guessing that (assumed you refer to @manueljenkin here). The author is not claiming that - at least not that I am aware of.

[PeterSt]

8 minutes ago, idiot_savant said:

copies to RAM directly (like playpcmwin) and plays to enjoy the benefits, as long as you don't move or edit the file (moving between different local drives, making a copy and editing would likely nullify the optimization).

 

This will be my last post in this thread for the time being. This seriously is shifting goal posts one time too many for me.

Maybe you can explain why you left our the context ? (see the full quote below).

I like games, but they should not be played with me.

 

 

On 6/1/2021 at 6:29 PM, manueljenkin said:

Since you would have both the unoptimized and optimized files you could easily a/b between the two and compare when playing on the junilabs player. You can use this new file with any music player that copies to RAM directly (like playpcmwin) and plays to enjoy the benefits, as long as you don't move or edit the file (moving between different local drives, making a copy and editing would likely nullify the optimization).

 

 

6 minutes ago, idiot_savant said:

Oh, and junilabs player most definitely loads into RAM before playback. 

 

Sure. Another one out of context. This time your own.

 

15 minutes ago, PeterSt said:

If you suggest it is played from RAM ... maybe. But you are guessing that.

 

I hope you see the difference in my suggestion and your stipulation ?

Have fun with the others !

[PeterSt]

Something else for a change ...

Juniplayer is from France. Why do I recall that the very (very !) same principle was ever back presented by someone from Japan ?

I thought to look into that for more data points, but I really don't know any more who that was.

And if I am right it was from Japan, maybe @yamamoto2002knows ?

[PeterSt]

1 hour ago, idiot_savant said:

are these different? Or can we know what they mean?

 

They all represent the same data. The song of each is the same.

They all look different. The song sounds different for each of them.

 

Your example better had been:

12345678

12345678

12345678

12345678

12345678

12345678

 

[PeterSt]

Thank you, Yamamoto.

 

So for example:

 

 

On 12/2/2016 at 12:43 AM, YashN said:

Nikhil said:

It appeared to me that the optimization is in using the appropriate SIMD instruction set.

I understood it as using the optimal instruction set to handle the data transfer to the CPU memory.

I'm sure Yokota San has his reasons to use these instructions. Here, his work on BH must have helped him decide upon this.

 

So, if you look at how data is stored on an HDD, you'll find that what we call 'digital' data is actually stored in an analogue manner, with magnetic elements aligned in a certain direction.

 

Under correction from someone like John Swenson who knows this better:

 

Now, to determine if a region on the surface represents a 0 or a 1, there's a circuit in the HDD which will have to work a lot to amplify these signals that are represented on the surface platter as these are tiny signals. So you could have analogue values A1, A2, An, and all these have to be first amplified then compared to a threshold to determine whether they are a '1'.

 

This work done by the circuitry has inherent noise and this same noise can pollute the power planes in the computer, in turn affecting the power lines on the USB used for the DAC, in turn affecting the DAC chip and thus SQ.

 

Presumably, what Yokota found here is that the tiny elements and signals degrade with time, requiring more work from that amplifier circuit to determine the bits.

 

So, this is what Yokota started to try and solve with Rewrite Data: he uses his software to read the data, then rewrites it (in between he now uses some optimisations that he discovered when coding BH but that's not the original issue he wanted to resolve). On rewriting, the analogue levels freshly written on the platter are high enough that the amplifying circuits don't have to work as much, and thus don't create as much noise as before rewriting especially when you're doing async USB streaming of that data to a DAC which is not totally impervious to power planes noise issues (i.e. most of them).

 

Now, it follows that if just rewriting the data on the HDD suffices to reduce the accompanying noise during reading (data which is then transferred to memory before streaming), one should hear a similar effect by just a normal O.S.-level copy of the audio file.

 

This is a proxy test of Rewrite Data without using RD: take an old file you haven't touched in a long time: not moved, and then just do a copy of it. Compare the initial file with the copy.

 

Indeed, I tried this despite all my brain and knowledge telling me there *shouldn't* be any difference, and there is: I find it especially audible on the 'first kick'...

 

Do the actual instructions provide extra here? There could be a good reason for him to do so, so hopefully we get him to answer directly, but as English isn't his mother tongue, it's difficult. I suppose the memory optimisation he found for BH also help by making those analogue signals even better, but without him saying so explicitly it's hard to know for sure.

 

I am fairly sure if he hadn't found any difference, the actual instructions wouldn't matter much.

 

Perhaps Minorisuke on the jplay forum can help being an intermediate.

 

 

 

[PeterSt]

3 minutes ago, kumakuma said:

Have any of you tested the software that the OP recommended

 

I did not.

Should I ?

[PeterSt]

People will now claim that you are talking about acoustical noise. Of course you talked about that, but it is only an exhibit of what is going on electrically. And that is the thing what matters.

[PeterSt]

16 minutes ago, yamamoto2002 said:

His laptop has a CPU fan that becomes loud when CPU is hot. Therefore the CPU fan rotates loud and it continues to rotate for a while even after playback starts.

 

And the fun : the cpu itself (heats and) cools within a fraction of a second. So it is not about that (except when we'd think that the heatsink absorbed the heat and now heats the cpu 😜) ... so it is really about the OS and what it all has to do in order to recover.

I did mention process explorer, right ?

[PeterSt]

6 minutes ago, idiot_savant said:

Now, I know I’m going to regret this, but…

are we now claiming that [...]

 

Are you ?

I am not.

[PeterSt]

2 hours ago, OldHardwareTech said:

They weren't yellow though.

 

I recall they were pink.

 

[PeterSt]

Not only submarines.