XXHighEnd Eliminated ?

[PeterSt]

Now I understand your interest in digital filtering lately. We will still need XX for the ArcPrediction algorithm don't we?

Hey Coen,

It is a combination of several things;

As I told (somewhere) I am actually out of ideas for further improvement of SQ in the realm as we know it. Maybe one thing was left and that's the "no device in the PC anywhere" option.

Next of course is the uselessness which popped up very recently just because all influence is eliminated. Careful though because this is for NOS1 owners only while XXHighEnd is general playback software.

Logically, what *can* be improved upon (because it makes so much difference to SQ) is the filtering. And well, if you'd look up the dates of this topic vs. the date of the topic of CharlieMb who started that topic at the very same day I started working on the new filtering, you will see that indeed all is "logic". It is just the one sure means to generally improve further (for example, CharlieMb does not own a NOS1) and thus that is now the work-field.

Whether we still need Arc Prediction ? I am as far as thinking Yes. So no matter you own a NOS1 or not, that is the base.

Side note and maybe off-topic for in here, but yesterday we were playing Love from the Beatles (that sort of remastered version of original tracks - this time very well done) and once into the Sgt Pepper tracks I asked "are we sure these are the original tracks" ?
(notice that only the lead-outs (posed as lead-ins) have changed compared to the originals)
I could not believe it.

I put up Sgt Pepper itself and was the most surprised that indeed it sounded the same. But did it 100% ? "this sounds less stressed" I was told, and it is my idea too (less processed). But still.

So, listening through that first new filter and you'll drop dead only listening to Sgt. Pepper because of the unbelievable detail it suddenly expresses while all of the emphasized highs have gone now (it has always been a strange album regarding that).
So mind you, even an album like that with SUCH all over familiar songs can sound so different that you are sure it is a new take with new performers.
I am not lying ...

So yes it is the way to go (now). I am the most confident about that and it might even bring more than what all happened previously. It will remain true though that having the NOS1 as the base is the very best (because by 100% guarantee it doesn't do a thing itself).

Regards,
Peter

[PeterSt]

Anyway this developement is somewhat of a dream come true. I guess we are now unchained from Windows and their future plans compromising the sound quality of our favourite OSses. No more lusting after expensive usb cables.

Coen,

Theoretically that would be true, yes. And FYI (something new ...) already in 2005 - just prior to the first XXHighEnd development, but pioneering with network streamers which were just new back then - I already streamed peer to peer music from, say, your PC (but MP3 because of the bandwidth). So I should even know a few things about that although the technology has changed. Fact is, the way XXHighEnd works 100% anticipates such means because it has always been in my mind as the future. It is only that SQ would heavily deteriorate but this now is out of the way ...
So, true.

We could have OS-X and Ubuntu NOS1 drivers and AP plugins. Man, I could be streaming music from the cloud directly to my NOS1 or have a NUC do the job. Etc, etc, etc.

This I am not sure about (yet) because it would need an indirect means to get it so some windows environment. I mean, no Linux drivers ...
But if you think a bit further (just be wild) you can see what the next step will be. Or better : now is allowed to be.

Peter

[PeterSt]

As a sort of side topic, what are you considering as the future for XXHE?  Is the way now clear to develop a driver for the NOS1 that includes the digital filters and upsampling so that other playback software can be used (eg. JRiver)?  Maybe even make it Mac compliant (I understand the 384KHz USB limit on macs).

So that.

And FYI : That doesn't require any high output rate from the source (player).

Project code : PhD.


PS: Think where/what actually is the chicken and what is the egg (as base of further solutions).

[PeterSt]

I think Coen is right - the future of XX is hard-core digital filtering, using an ultra high-powered PC feeding a PC-immune NOS1.

Almost correct.

But for now it *is* correct.
And so far ultra-high powered is not in order; I still have my audio PC running at 430MHz. But I guess that needs some step-up soon.

Peter

[PeterSt]

The end of XXHE ? ... It makes the PC dead quiet, helping USB to be more noise free... Minimise OS, unattended... All now for nothing...

?

Yes Alain. If I only look back at that "eliminate devices" sub project you know a few things about ... that is sad. But I am not like that especially because it has been my objective forever.
It even is so that I started to develop D/A converters so I could see how software (PC) influence could be eliminated. Well, it took a while.

Still, and people should no forget that, XXHighEnd was there first and it was not made for any NOS1; it still is not and never explicitly will. It is only that it contains some special features to let the NOS1 excel better. So on the technical side : like automatically shutting off the oscillator not in use. So that is XXHighEnd too and it obviously improves the SQ of the NOS1 (no towards each other oscillating oscillators).
But same thing can be done manually through the NOS1 Driver Control Panel.
Similarly we can manually upsample files (and quite some do I can tell you).
And of course there is more, where the convenience factor plays a role while SQ is the best at the same time. That all is XXHighEnd although part of this vanishes for NOS1 users while other parts are added for exactly the same users (the filtering). That is, as long as I will be able to out-better "others".

Blahblahblah
Peter

[PeterSt]

using an ultra high-powered PC feeding a PC-immune NOS1.

Hi,
that is an item where I'm sitting in the dark (as always).
Peter and/or same Forum members are underclocking the CPU down to 470MHz (under performing). The CPU i7 3930K costs more than € 500,00 and then ......... .

Thanx for an explanation.

Joachim

Hi Joachim,

I hope this is not going to be vague, as I have a clear idea bout this. Look :

All is evolution. Literally. So when this PC emerged, the first thing I did was overclocking it to 4GHz and let remain all cool at the same time (XXHighEnd feature). However, this was from the time that or it was not recognized that lower cpu speeds could sound better or it just was about a nice convenience of loading albums for Unattended in a wimp.

But the mere point was : This was (still is) about a modern 12 core processor fitting a mini-ATX motherboard which is coincidentally very fast. And still nothing else exists.
So I could give the answer "who cares" because it was and still is about other things".

The PC (and mind you, not the home-made ones only carrying the same motherboard and cpu) is about sheer versatility. Can be way overclocked very easily, can be underclocked and either way it stays cool because of how it is setup (but in combination with XXHighEnd). OS drives can be swapped in 5 seconds, playback drives can be added, and well, it is about *that*. And if about that only ... it can also contain a "cheap" 8 core.

Anyway to give an indication :
While I useD it for severe underclocking for the better SQ, I can now usE it for severe overclocking because the new filtering might need it. And all with 12 processing threads in parallel.
So this shows that the "XXHighEnd PC" is only made for anticipating on whatever XXHighEnd's future has in mind. This really has been my objective, not matter it was originally made for 6moons to have a general best audio PC. Ah, "general". So after all eh.
Compare a laptop. Let it run for a minute with turbo boost and it will shut off (throttle down) because of heat. Underclock it and it won't respond what it hardly did to begin with.

Was that something for an answer ?
Regards,
Peter

[AlainGr]

Peter,

I know all that this software had cost you in time to make and improve our music to such a level that it is unbeaten. But if it always had been your aim to eliminate XXHighEnd, then this is a wonderful achievement

I will wait for what's next (the PSU change) to see what are the options... But of course, there are still features that will be necessary to have to bring the right signals that the NOS1 need to work properly. I suppose some of them will now be included in the driver, but anyway, there will always be a need for a software to start from.

Ad this raises questions like the isolation of the USB controller card from the PC chassis (plus the black wire that is detached in the NOS1)... Plus a few other things I am not thinking at the moment...

We can either think of things in a sad way or be happy with the results of your discoveries and hard work.

I prefer to choose the most positive direction

All should be easier to grasp when we will be faced with it in our homes

Alain

[Gerard]

And W7 or W8 does not matter anymore?

 

[christoffe]

using an ultra high-powered PC feeding a PC-immune NOS1.

Hi,
that is an item where I'm sitting in the dark (as always).
Peter and/or same Forum members are underclocking the CPU down to 470MHz (under performing). The CPU i7 3930K costs more than € 500,00 and then ......... .

Thanx for an explanation.

Joachim

Was that something for an answer ?
Regards,
Peter

Thanx Peter,

strategy understood.

Joachim

[PeterSt]

And W7 or W8 does not matter anymore?

How dare you ask the question !
haha

Welllllll ... I am not 100% sure of that. So, when I had this running the first day and tried whether indeed XXHE settings did not matter (they did not) I tried Windows 7 (against Windows . I would swear that I heard a difference, BUT, this was almost without getting used to the new sound. So, not easy to judge.
I tried it a week or so after that and I wasn't sure any more at all.

If there is still a difference this has to be about the OS itself spitting out different "data". Well, like I have been suggesting quite some more often.
But this is a bit similar so setting the SFS too low when really data starts to be lost (and which is not so obvious in 1.186 because of no real ticks occurring) - but what I can measure. So just saying ... for my 430MHz and W8, the minimum SFS I use is 0.32 and that still is not 100% error free (measurement wise). But the lower the better it sounded, so I used that.
Now the difference between W7 and W8 on this can be quite huge so it would be safe only to compare the real higher numbers (think 20 and up).

One thing : Since I use the NOS1a principle there hasn't been a single issue in all the 5 weeks I am using it now that relates to "see, W8 is sh*t". NO-THING. So that is the very first thing what will be happening to you.
Also, when I previously compared the both it was totally clear within the second that both sound different. Now though -as said- "I would swear that" ... etc. or ... "not so sure any more".

Peter

[Gerard]

Well i guess time for saving some money!

 

[manisandher]

Another quick question:

Will XX remain a true 'memory player'? I suppose with the NOS1a, it simply makes no difference, right? (Though for other DACs, of course it's still totally significant.)

Mani.

[boleary]

Also, what about upsampling rates? 8x has been my preferred setting lately. Surely even with the NOS1a the sound will change at different upsampling settings. Oh, one more thought; will a SFS of .4 sound the same as 200? Can't wait to hear this "new" DAC!

[PeterSt]

Brian,

Surely even with the NOS1a the sound will change at different upsampling settings.

Correct (and obviously indeed).

will a SFS of .4 sound the same as 200? Can't wait to hear this "new" DAC!

Yes, but now we must be careful; I think I already said it elsewhere, but when going too low samples are being skipped and the way 1.186 works may make that unnoticable; I think everybody who tried it (find the lower limit) knows that at some stage you start to think "do I hear skips or what" ? At that moment by long the boundary has crossed. So for example, I can see by measurment that 0.32 already does that (can skip) which I do not hear and happily use (still do somehow).
Whether the analyser sees it depends very much on whether it's captured on a place in the full wave cycle that it disturbs. So while a fairly constant skipping should be going on, only once per 10 seconds or so an anomaly is seen.
When I make that (as I recall) 0.70 I see nothing any more, but now the amount may be so low that the chances the analyser sees it have become too small.

Peter