2.10 sound quality

[Rmalits]

After reading these post I did some tries with much higher SFS values for my Stealh Mach II and reduced Q1x significantly.

SFS = 140.19 doesn't work with the Mach II, it's pretty spurious
SFS = 70.19 brings amazingly good sound stage, but SQ? It sounds too loud, due to distortion I guess.
SFS = 35.19 works nicely when Q1x is down to 2 or so. That's what I am trying now.

SFS seems to be limited somewhere there with my Mach II. But trying such high SFS gave me kind of an idea what it could sound like, holographic, spooky, ....

So, Peter, I need a Mach III. There is no way around I think, althouh this would be my third Stealth already within 2.5 years 
What will be the delivery time when I order one now?

Kind regards
Richard

[PeterSt]

Hi Richard,

What will be the delivery time when I order one now?

When you go for the 14/28 processor (which is really advised by now), say two weeks if I can get a Motherboard from somewhere. They are sparsely available and I just put one to an RMA procedure this afternoon because it wasn't right. Still waiting for another one for a customer (for 3-4 weeks by now), so actually it is going sloooowwww.

By the time you ask me to make an invoice for you, I will start looking elsewhere as well and let you know the estimated lead time.

Thank you Richard,
Peter

[PeterSt]

Richard, in order to keep this topic clean somewhat, I moved your last reply regarding this to here : Re: Phasure Mach III Audio PC with Linear Power Supply.

Regards,
Peter

[Robert]

As a result of a recent Bios update and resetting, disabling Bios settings, more than I have ever been brave to try, the subsequent sound with Peters current XX settings is amazing. I also prefer Disable Cracks off sound wise.

Lush 2 is on the horizon perhaps a Xmas present. 

Robert

[briefremarks]

Robert,

Looking at you signature I see you are using Scheme 3-4; Peter has 3-5.

Ramesh

[Robert]

 Yes only have 4 cores, 3/4 sounds the best. I was meaning SFS and balanced settings primarily. Although I should retry the core settings.

Bios settings have a large effect on the sound something I'd forgotten. I have been running Bios update default settings for 3 weeks and didn't notice till Peter's Disable Crack comment. Its interesting how one can get use to a sound quite easily and not realise some things are not optimised.

Yes I'm miles away from the new Stealth MKIII still better than a laptop.

Robert

[numlog]

Being unfamiliar with software would it be likely for current version of XXHe to work on Server 2019 or will that require some work?
and working might not mean sounding good but at least so far impressions of 2019 have been very positive.

btw would OS changes require reactivation of XXHE, so that experimenting with different OS isnt really an option?

[PeterSt]

btw would OS changes require reactivation of XXHE, so that experimenting with different OS isnt really an option?

It would, but you will be refunded right away because, well, that is the policy.

It might require work but what that is can not predicted at this moment. The least what should happen is that it technically differentiates from the others. From there it can be tweaked. Small problem : not my me ...
So FYI : at this moment there is hardly interest in WS2019. This is because the desktop OS is "minimized" to less than (core) server to begin with. This has been so in the past for all server versions ...

Peter

[PeterSt]

Hey All,

I think it is fair to say that I have chosen a new SFS setting(s combination) : 10.19. This, together with Q1 x xQ1 of 30 and 20, a ClockRes of 15 and Q3,4,5 at 1,1,1. For those with the NOS1 : The buffer size in there is 16ms.

The .19 fraction in the SFS of 10.19 is "as usual" (lately). The "new" about this should be the combination of the somewhat lower SFS with the otherwise higher settings.

Watch out : I am posting this in the general "Sound Quality" section because I have to post it somewhere. However, people should realize that a larger bunch of you is not coming along with all what is new as of late, and therefore what I have so say will be moot to you / that bunch. Of course, everybody must obtain what they deem fit and what the $ (AKA wife) allows for, but in any event the below text will be quite useless to those who did not come along. Still, because yet another group of people sure does come along (and stays on par), it still seems justified that I post about it anyway. OK ?

The specificity about my "hardware" setup what most of you don't possess comprises of :

- Stealth Mach III Audio PC (10/20 Core);
- Lush^2 USB cable (100cm);
- Blaxius^2 Interlink (325cm).

Each of these elements contribute vastly (and I mean vastly) to the sound, thought impossible only 6 months ago. The leap is more than a quantum leap. So much so that I start to feel alone in this universe. And I hear ya : No, it is impossible to improve over what I currently have.
Sure.

What the whole lot today is about is speed, speed and even more speed than you will ever be able to comprehend. This is so much so that "original" completely bland sounds (like from a synthesizer) now fold apart into their individual frequency oscillator on-off sounds. Think like a prior sine (sinus) sound becoming the not-sine at all because of it going on/off in high frequency. I have talked about this more often in a far past and for those who remember : think like the Silverstone USB card's expression and how that too could be dedicated to on/off sound, although in a relatively very low frequency. Think 20Hz or so. And mind you, the normal "sine" frequency can be anything, e.g. 2000Hz, but if 20 times per second that sound briefly goes completely dead, you have a kind of modulation which could be "interesting". Of what I recall, this worked very well on the cymbals, but, it was fake. It brought a flavor. And we only noticed when environmental things improved (I think it was W10-14393.0 which killed the usage of the Silverstone card ? - or maybe it was (the combination with) the Clairixa USB cable).
What I am talking about today is the inherently present on/off sound many instruments produce. Thus not only synthesizes (when set to do it) but also a violin, a trumpet, a saxophone, a snare drum, a cymbal with nails in it (sizzle cymbal) and in the end also a human voice. Etc. etc.

Today it is relatively easy to observe that an almost infinite on/off frequency is still normally audible (say up to 200Hz) or detectable otherwise (above 200Hz). It adds a dimension to even artists as a whole, out of this universe. I mean, look at Jean-Michel Jarre's latest - Equinoxe Infinity. I was encouraged to get and play it while I initially refused. I deem the music too simple and too much of "synthesizer". But oh wow. His "production(s ?)" are completely transferred into something which takes you away with open-mouth interest, just because of what "sound" can do, that ending up in sheer music. And all what actually happens is the bzzzzzzz sound. And yes, it is your turn to transfer this bzzzzzzz into an on/off sound. Try it (with your tongue/mouth) - you will manage, no matter you thought the zzz was completely concatenated. It isn't. It goes on and off at a very high frequency. And when it does not, it won't be your bzzzzzzz but merely a sine of some frequency.

When I now first start out telling about the previous setting of SFS=140.19 (all else the same) and noticing how the Blaxius^2 was (is) more and more breaking in, I found the clear sound to become too profound. Not tiring or the like, but just overly clear. But what I also noticed - and this will be the speed of the speakers - is that some sounds seemed impossible to cope. This is hard to explain without listening, and even with listening it would be tough to discover tracks with it unless I noted them somewhere which I did not, but you could observe a kind of oscillating sound some times which just occurred 1 time too many per day in order to believe it was a natural sound. IOW, after quite some weeks now with the same (hardware) setup, I started to believe that something just could not cope.
Btw, easy to think it is incomplete reconstruction filtering, but I tested that and it doesn't make a difference for these observations.

Making a side step (again, sorry), I am 100% sure it is the Blaxius^2 which in some abnormal way preserves all which otherwise "escaped". I think of "frequency escaping from the cable" which is almost literal. That is, if we think about how shielding could preserve higher frequency (and real on/off is infinitely high frequency - though bandlimited by the sampling rate) and less/no radiation with it, then you can sense how I think these days.

And so I have been working on this and eventually came to the logic" of the system being able to track the higher frequency on/off sounds better when it would respond faster. This now is about the computer only, which, again inherently already is blazingly fast itself (the Mach III with its fast PSU, explicitly made for this (speed) job). But there's also the software ...

Others have noted it too, but when all is so extraordinaire(ly) fast and with that the most critical to distortion, it becomes a must that the PC does not run for too long. Mine normally runs 24/7 for weeks and weeks until something coincidentally requires a reboot, but these days I'd say this is not allowed any more. In the third day you will hear it. Whether this implies a daily reboot I don't know yet, but it could be good behavior.

Now on to the change, may you want to try it yourself, regardless whether you are "on par" :

On a new day and prior to your listening session, reboot the Audio PC (or boot it when it was not on in the first place - haha).
From that point on, set the SFS to 10.19. Thus, do NOT do that after listening to a higher setting like the current (for many) of 140.19 on that same day. If you do that you won't be able to cope hence like it. I am serious and this is a brain thing. Also, the reboot is necessary to not let it fail because of distortion in the first place.

Know that 99% most probably you may need a whopping 40 minutes of playback (all active, hence playing without sound will not help) before you are even allowed to judge. Do notice that I myself will not solve this by means of starting playback and have a shower to avoid the sheer distortion of that period (all way too cold) because, well, I like to observe what happens. No matter it is day in day out like this, it is just my behavior and I have no problems with it. But merely : you may notice a few stages of clicking in and at some final stage your adrenaline boosts, your jaw drops and from there you won't know what is happening to you. Or, what the system now is capable of.


Ah, I was going to talk about the SFS of 10.19. All right ...

What I sense of this, is foremost the strange combination of again more detail but with an unsurpassed silkiness. The too clear sound I talked about in the beginning, now becomes more natural. It could even be less digital, if you want. Less clinical. But mind you, this is not the explicit observation; this is merely about a disturbance eliminated. The disturbance as such is not about bad sound, but a deep down annoyance of the thought "can this be right ?". At least that it how I myself listen. The SFS of 10.19 in my mind responds faster to required changes (or possibly to "frequency"  as such) and now I write this I see that I maybe must look into the Nervous Rate as well, because that would *really* be responding to required changes.

Side note : if you'd change the SFS of 140.19 to 10.19 in an A-B comparison, you will be missing the clarity and will switch back in no-time. This is why you must start your today's session with the 10.19. IMHO etc.

What I find interesting is that a perceived hardware-limit (like something can't cope - speakers or amplifiers) can be solved by more "response speed" in the software. I think we all have the experience with the lower SFS'es (well, maybe not the 10 range so much but way lower) and how that increased crispiness. Am I right ?
So how is it possible that today it goes the exact other way around. Today - and I am quite definite about it - the higher genuine frequency implied fills holes which otherwise are too square and imply false higher frequency. Oh, that is crisp all right, but is it happening in reality ? And, if response is more slow the holes close to a bland(er) sound and all merely become sines. In addition I think that the higher frequency real (mere) sines of, say, 2000Hz, become too much forward because in reality they were no sines but contained small (on/off) interruptions. This softens. Or better : all what is more real will be for the better to beging with. This (too) is how our brain operates.

Right. As you can see I am at inordinate levels of judging now, which only happens because it just can be observed. Key of course is how an observability can be turned into something for the better again. This post was about that. And it took me two weeks or so of trying (getting rid of the deep-down small annoyance). So it is not easy.
But man, is it rewarding.

Peter

[Robert]

What's the max SFS set to with 10.19?

Robert

[PeterSt]

What's the max SFS set to with 10.19?

Robert - What the Max SFS would do is reserve memory for the higher SFS settings; when you didn't make this reservation, the problem could be that by the time you'd need that higher SFS (like 140 again), the contigous memory has been eaten and it will require a reboot first.

From this follows that you can apply the Max SFS as a tweak. Notice though that I don't have the real experience on this (but since you ask ...);
The more memory you reserve, the less the OS will be capable of using it. This is because *you* are using it (albeit empty except for the small SFS amount you really use). This could force OS tasks not to run - and you don't want to run anything anyway (except for your audio).
It also would prevent the Garbage Collector to be active regulalrly because what wasn't used (because you use it permanently) doesn't need to be cleaned up either.
Etc.

Contrary to the above, one could argue that you make it the OS more difficult by taking out memory it needs to use. So now more activity takes place.
Again, etc.

Peter

[Robert]

Are you running the Max SFS 10.19 at 10.19 max?

I've tried this and there is a clarity/smoothness over 140.19.

Mine normally runs 24/7 for weeks and weeks until something coincidentally requires a reboot, but these days I'd say this is not allowed any more. In the third day you will hear it. Whether this implies a daily reboot I don't know yet, but it could be good behavior.

Yes this is something I discovered last week after leaving on 24/7 for a week. I thought gosh music has become dull or is it my ears. I rebooted and music came back to life. If one plays music everyday I think fresh reboot is required everyday but as you suggest no longer than 3 is ok. I don't bother removing HD that often anymore.

Robert

[acg]

Yep, I've gotta reboot my AudioPC every day or the sound gets flat...it's very noticeable.

[PeterSt]

What's the max SFS set to with 10.19?

Robert, aoplogies. Somehow yesterday I was not able to interpret this easy question. I read it as "what is the relation with".
Now the answer :

I have been careless and after looking just now I found it was set to 140.19. This is because at first the SFS itself was at that. Better would be to activate the "Maximum the same as SFS". That is, assumed one wants to reserve as least memory as possible. But now you won't be able to test e.g. 20 after a while (memory eaten).

And hmm ... yesterday I changed my Nervous Rate from 10 to 3, did not find it to be a real difference, but now I see that it also requires an Apply (so it was atill active at 10). Maybe it was Black Friday ? Or the placebo just didn't work.
I now clicked Apply. Next we can bet that tonight all sounds lousy and I can't find why. Maybe someone reminds me ...

Regards,
Peter

[Robert]

Yes I find some changes are easily heard others subtle. I havn't changed to Lush 2 yet kind of waiting for things to settle but now its moved to other cables. I do wonder if Nordost have already discovered what you have with these recent advances with screens. The more I think about it they can only do so much with materials. Mind you when you can afford a 24/7 design team its highly likely you discover things.

I really see the benefits of the Stealth III the source has always been critical in audio.

Xxhighend software still has hidden talents.

Robert