XXHighEnd - The Ultra HighEnd Audio Player
October 16, 2018, 03:24:44 am *
Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
News: August 6, 2017 : Phasure Webshop open ! Go to the Shop
Search current board structure only !!  
   Home   Help Search Login Register  
Pages: 1 2 [3] 4 5 6  All
  Print  
Author Topic: Phasure Mach III Audio PC with Linear Power Supply  (Read 12908 times)
0 Members and 2 Guests are viewing this topic.
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15242



View Profile Email
« Reply #30 on: May 05, 2018, 10:43:34 am »


OK. I am thinking of putting up the original post I had in mind originally (but saved it elsewhere), that post expressing about the sound. But for now it could only be more confusing when doing that prior to this very post I am writing now, because of new data available ...

Yesterday I have been playing with a 10/20 core. Say that this is the one for fmanheck but only in "test" and to see what the 10/20 core would bring.

Now I understand that the CPU is from the gold serie and not the silver serie. So in therms of CPU it is really a step up from the 26-serie.
Peter, am I right? Any thoughts on this decision?

And that was a still pending question ...

Whether Silver or Gold (or Bronze or Platinum) is maybe not of importance (I do not know of inherent technical differences apart from number of cores and processor speed) but anyway this 10/20 is the 4114 and it should represent the good old 2640v4. Well, for the number of cores, that is. But there the comparison stops. And this is all related to the explicit "Server" environment which is radically different. So I need to say it again : the 26xx Xeons were actually consumer oriented and today no such an environment exists any more for the Xeons, unless it would be the new W series but although looking interesting, this is the opposite of what we'd want and is more "laptop" oriented. So we have the "Xeon Scalable" and it is their architecture which makes them interesting for us. Well, says me and with the hopefully present experience (which often is merely about hunches but alas).
It is also about the six lanes of parallel memory processing.

If I first try to summarize everything : this is all about how the power provision was made "super stable" which in itself happened a bit by accident but had its foundation in the fact that no normal cooling mechanisms exists for this. It anticipates 90dB generating blasting fans (ever been in a server room ?) and with that airflow and in an air conditioned room of 15C or so, it will work. But this is all not really us, right ?
So if I did not tell it yet : there's always one very slow fan running now (not PWM controlled which is good) and the result of that is a.o. a stable power draw. It is way more stable than the Mach II and at hooking up a power consumption meter this is easy to observe (Mach II jumps regularly, Mach II is always the same).
Do notice the importance of this *if* we recognize that the minute differences implied by stupidities as XXHighEnd already implies vast sound changes, right ? Thus, seeing the power consumption change by 2-3 Watts all the time, can't be the example of the theoretically best sound (about theories, hunches and real empirical experience).

Back to the 10/20 processor

When I was listening to this setup yesterday, into an hour of it, this text for you all came to me (and I hope you recognize I am always honest) :

This 10/20 processor is for those who do not have the very best system. It now will sound better than the 16/32 (the only other one I have experience with so far) just because the 16/32 demands too much of the system (more elaboration later, if I don't forget).

A few hours later, partner in crime Ciska arrived back home and I know what she sort of complained about the past few weeks, which I will tell about later as well, but which made her say literally this :
"Hmm ... maybe the one processor is better for more inexpensive systems and the other for the more expensive ones". And the sweet question : "how's that actually with the Orelos ?" ... and then I started taking about small chips producing the sound instead of a 1000 components etc. with the question what could be better when. And also about the speaker itself of course and how its speed matches, well, the 16/32 processor's exhibits. So ...
So let's quote myself from the text I did not put up, actually for the reason of too much of theory at the time (which was a small week back) :


Sound ... literally shattering good.

Yes, the keyword is "shattering" here. I wouldn't know a better description, although to you it may come across as a negative. But the contrary is the case.
Let's first try to see that all wht you'll hear from the Stealth III is delineated like crazy (that word again). Each is on its own and with that in mind now think "shattering". Or better : how shattering (like copper instruments) can really shatter. How it creates a crazy amount of layers and fills the whole room with sound, this time not with low frequency only; Supposed you play something of not the most common instruments and you may never have witnessed those instruments live, then such an instrument can/will now sound all over different than you ever imagined. And, as so often, you won't even be able to recognize the track/album. I had an example the other day I now forgot, but think like a most normal Steely Dan of which you have to look twice whether you did not put up a(n unknown) coverband. Com-ple-te-ly different.

Bass is like you ended up in hell.
Deep Thunder is like ... OK the same. But there wasn't deep thunder previously in tyracks accordingly.
The highs are so so idiotly "short" and bright and whatever superlative more we can try to think of, that all else now sounds like dirt. And yes, I have serious problems with listening to the Stealth II now and if I had to go back to it I'd give up the hobby first.
Can't imagine that eh ?


Here is where I decided not to put that up because I couldn't talk existing Stealth II customers into spending another 4K+ euros only because of something new again, the old for some only months old. But I also seriously started to wonder whether what I perceived over here could work out for others elsewhere. I mean, that "shattering" won't come automatically; it requires speed speed and more speed. Like one small chip processing all of the sound (as how the main amplifiers in the Orelos do it). What would happen when this speed is no available ?
Now, together with Ciska talking about "maybe too cold" (which is an exhibit of "super speed" because of all the highs coming to you) I started to wonder whether I could do something to tame the speed. And I found ...
Q1.
It was a kind of logic (for thought) because we all just started manipulating that one for more speed and after trying a bit, I ended at 30x10 (instead of 30x40). Now there was less of the super speed and more of the necessary warmth.

Well, day before yesterday we both explicitly checked that, and all what I heard over and over was "this is not what I expected (it could do)". So to a large surprise now and the warmth was back, and the speed was there just the same. But her surprise was about S-es remaining normal, while all the highs seem to incur for "overing" that. But not so. Not so at all. And now it requires my original post about the SQ impression, but never mind; I can refer to the "Addictive" I put up for the replacement post. Highs which clearly incur for the worst (like shattering braking windows) while nothing is wrong with S-es. How can it be.

Wait, we were talking about the 10/20. And btw, this is not 4K+ but 3K+. Also I must now revisit the prices because by now I know exactly what goes in there. OK ...
What I could sense of that was the better capability of my system. Yes. Well, mind you, I reason towards matter I hear and this is what I am now saying : my system of course is not the best of all. It just isn't. And it is just the same as reasoning with the 16/32 for a base : the system must be sufficiently good to let that processor (setup) excel. It does here all right, but ...
But it needs just that tad more ?
No wait, with Q1 at 30x10 all is fine. Yyyyyess, but what about more speed sounding worse then ?

I hope I am clear a little, but probably not at all. But the 16/32 summarized : for me this is perfect when I don't dial in the max possible for speed (as how I perecive it now).
The 10/20 is perfect at 30x40 although I can't check beyond that.

The 10/20 sounds totally different again and it is the "woman's" processor. It is more slick (sound wise). It shows all the (crazy) speed of the world more downwards in the spectrum. It is more easy on the ears. It is not as 100% interesting compared to what the 16/32 can do.

And mind you, between the 2640 (10/20) and the 2660 (14/28) I never perceived a real difference "but if anything the 2660 sounds better". Not so when the "Scalable" 10/20 and the 16/32 are compared. The difference is huge again. And btw as huge as the difference is between the 2640 (mind you, for a whole computer system) and any of these two.
See the "dirt" remark more above, from that post I did not put up. It really is so.

So finally the verdict between the 16/32 and the 10/20 through these ears :

- 16/32 exhibits speed that at least lets my system run into limits, as it seems. Not 100% sure yet as more settings exist than Q1. Bass excels, highs excels up to shattering good, mid is just very good but may stay behind a bit caused by the overwhelming highs. Can be tamed by Q1 towards more warm sound, but it seems a waste to do that (the super special may disappear somewhat).

- 10/20 exhibits speed in the mid, is warm up to being in a warm bath. Bass excels possibly even more. Maybe no real shattering highs but easy on the ear always for those who want easy on the ear.

So, done.
nea
The 10/20 was a test. Next up could be the 12/24. I expect this to be "better" again than the 10/20, just because the 10/20 does not let me (with my system) run into limits that I can see (OK, after one day of listening !). Point is that I can't get it at this moment. And, going to the 14/28 is a bit moot because of the price approaching the price of the 16/32. Also, I obviously can't be obtaining processors forever just for testing them. Not at these prices. They should be feasible to sell. And well, the 16/32 for sure is (I probably get one for myself) and the 10/20 for sure is too (I will keep using that if nobody wants it smirk).



Now I hope you don't regard this as the most idiot blabla reasoning ever (about system limits and such) but anyway it is really how I now think.

Regards,
Peter
Logged

For the Stealth III LPS PC :
W10-14393.0 - August 18, 2018 (2.10)
XXHighEnd Mach *III* Stealth LPS PC -> *Xeon Scalable 14/28 core* with Hyperthreading On (*set to 10/20 cores*) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = *140.19*  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = *35* / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> *Lush^2 A: B-W & Y-R, B: B-W* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
Michael_B
Audio Loudspeaker
*
Offline Offline

Posts: 16


View Profile
« Reply #31 on: May 05, 2018, 05:00:44 pm »


To Arvind, Fred and Ramesh, thanks for your warm welcome. smile

Regarding the LEEDH E2 speakers, it's true that they are not very efficient but they are very well matched with the AVM SA8, which has very similar specs than the Primary A32 owned by Fred. I'm kind of allergic to boxy sound speakers. The E2 is totally boxless. It's based on a unique technology using neodimium magnets, pistons made of composite materials enclosed in a totally inert chamber filled with ferrofluids. They are very neutral and their very low mass make them very responsive and quick. In a normal room they can provide sufficient bass but in my case the speakers are located in the living room which is totally open to the whole house. This  makes it difficult for the bass to fill this large volume. That's why I added the LEEDH Sub which is also boxless and has been designed to specifically match the E2. For those curious about these speakers you can go on the LEEDH site http://www.leedh-acoustic.com/Loudspeaker-leedh-e2-uk.php

Fred - You're right NH is not that far away from Montreal. Who knows maybe eventually we could arrange something during a short vacation in your beautiful area. 

The comparative description Peter made of the Xeon 10/20 and the 16/32 (the one I'll get) above, is interesting and also intriguing to me. I'll soon have a taste of what he means since the III is now on its way. Happy

Regards - Michael

Logged

Audio PC: Mac mini Mid 2011 / 16GB / 240GB SSD /  Uptone JS-2 power supply / Server 2016 as OS / Audiophile Optimizer / Fidelizer / Process Lasso / Primo RAMDisk / HQPLayer / Euphony (linux)
DAC 1: NOS1a/G3, B75 (2x) / Blaxius / Clairixa and Lush.
DAC 2: Chord Hugo / modified power supply: now runs off 2 x UltraCapacitor power supplies LPS1 from Uptone Audio.
Preamp.: none
AMP: AVM SA8
Speakers: LEEDH E2
Sub: LEEDH 20.1
AC Power: 3 x Dedicated lines for: Amp., PC + DAC and Sub / 2 x Topaz Ultra Isolation transformers 1.8kva (ea) (0.0005pf) for the digital and amp.
Others: Uptone ISO Regen / ICRON USB Extender 2304GE-LAN / TP-Link MC110 Fiber Media Converter
Michael_B
Audio Loudspeaker
*
Offline Offline

Posts: 16


View Profile
« Reply #32 on: May 07, 2018, 01:51:54 pm »

Hi Peter,

The Stealth III is not far from home. There is still customs to clear, so I expect a delivery today or tomorrow.

In an email you sent me you mention that the only way for the III to access music files is via LAN. I will then use my Mac mini as a music server. Technically is the III capable to read music files on a USB stick connected to one of its USB port?

Now, to control the III I want to use RDC with my Surface Pro. This means that I need to use a second Ethernet port on the III to connect it to the Router which has a wifi connection. This will work, right?

Thanks - Michael

Logged

Audio PC: Mac mini Mid 2011 / 16GB / 240GB SSD /  Uptone JS-2 power supply / Server 2016 as OS / Audiophile Optimizer / Fidelizer / Process Lasso / Primo RAMDisk / HQPLayer / Euphony (linux)
DAC 1: NOS1a/G3, B75 (2x) / Blaxius / Clairixa and Lush.
DAC 2: Chord Hugo / modified power supply: now runs off 2 x UltraCapacitor power supplies LPS1 from Uptone Audio.
Preamp.: none
AMP: AVM SA8
Speakers: LEEDH E2
Sub: LEEDH 20.1
AC Power: 3 x Dedicated lines for: Amp., PC + DAC and Sub / 2 x Topaz Ultra Isolation transformers 1.8kva (ea) (0.0005pf) for the digital and amp.
Others: Uptone ISO Regen / ICRON USB Extender 2304GE-LAN / TP-Link MC110 Fiber Media Converter
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15242



View Profile Email
« Reply #33 on: May 07, 2018, 04:56:57 pm »

Hi Michel,

Surface -> (RDC, WiFi) -> Router -> (cable) -> Mac Mini -> (RDC, cable) -> Stealth.

Start with the last part ("Mac Mini").
This is the advice, so let's stick to that. Later you can explore the 1000 more possibilities/combinations, right ?

Kind regards,
Peter

Logged

For the Stealth III LPS PC :
W10-14393.0 - August 18, 2018 (2.10)
XXHighEnd Mach *III* Stealth LPS PC -> *Xeon Scalable 14/28 core* with Hyperthreading On (*set to 10/20 cores*) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = *140.19*  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = *35* / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> *Lush^2 A: B-W & Y-R, B: B-W* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15242



View Profile Email
« Reply #34 on: May 14, 2018, 12:40:02 pm »

Hi all - time for a small update.

The 10/20 processor must be in for 10 days now and I must say that I feel I have it fully "under control" now. Of course it can be so that by now I forgot the sound of the 16/32 processor, but there's now so much more interesting in the music than what I recall from my Mach II "era" that by now I would definitely stay with the 10/20 if I could only choose that one (and my Mach II).

Two main matters happened which made this work really well now :

1. For this processor I changed a few things to the PSU (to be honest it has become a whole new PCB so it better fits the case in the mean time). Will post a photo later, if it means anything to you anyway.
This PSU required some break in of 4 days. At least this is what noticed. This was completed 3 days ago.

2. The somewhat more "grey-ish" sound (say a bit more slow) compared to the crazy 16/32 regarding the highs, could be tamed by putting back the buffer in the NOS1's Driver Control Panel from 16ms to 4ms. So yeah, what experience all can do (didn't I use that 4ms for "years" until the recent long Q1 setting method). So in combination it brings more sparkle for this setup - the sparkle I missed since the 16/32 went out (to Michael).

Yesterday I ended the listening session first with Autobahn (Kraftwerk) and of course everybody is allowed to call me crazy to play this "music" in the first place. But mind you, because it is so blatantly stupid - 20 minutes long, it really has to have something or otherwise indeed it is "nothing" as such. So might you don't know, those (simple) synthesizers together with the simplicity can just spellbound you. But it has to work.
earlier on I told about the mid detail of this 10/20 setup. But what mid-detail I actually mean here ? well, how a lower keyed synthesizer can be sparkling (shattering) in the lower frequencies because of its on/off sound. Yesterday I heard things in that 20 minute track I never ever heard one bit of before. When it happened the first time we both looked at each other and equally said "huh ??!".
It was not even me who said that the PSU must be broken in now, she also referring to a "hey, but the other day this Autobahn did not work much, right ?". Correct, I even shut it off preliminary.

Talking about spellbound ... when Autobahn was finished there was a bit of time left before dinner would be ready and I thought "oh wait, if Autobahn sounds so mighty good suddenly, what would In-a-gadda-da-vida do ?". And yes, nothing worked out better than that if it was for the return to the atmosphere of dark basements with (fish)netting against the ceiling, an LP shimmering while playing this 17 minute track with the xth Iron Butterfly drummer because they all died of the too long drum solo (who ever made up that story for this few minutes of real solo I don't know, but I am sure that I believed everything at the age of 12).


If I feel like it, tonight will be important because I plan to move my USB cable (ending at the input of my NOS1a/G3) to the Stealth II (yes, that is still running from its 8++weeks RAM boot). Part of the plan is that I will try to tell you as accurately as possible the difference between the sound of the (2640v4 - 10/20) Stealth II and the 10/20 Stealth III. I will (honestly) do that too if the Steath II sounds better of some sort. But I would not count on that. evil

More later. Kind regards,
Peter
Logged

For the Stealth III LPS PC :
W10-14393.0 - August 18, 2018 (2.10)
XXHighEnd Mach *III* Stealth LPS PC -> *Xeon Scalable 14/28 core* with Hyperthreading On (*set to 10/20 cores*) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = *140.19*  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = *35* / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> *Lush^2 A: B-W & Y-R, B: B-W* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15242



View Profile Email
« Reply #35 on: May 15, 2018, 05:24:43 am »

I didn't have the proper ears for anything, yesterday.
It's a bit of a psychological thing as well, because while I couldn't find my sound, I noticed that the settings (of XXHighEnd) were not as how I expected them to be, so I probably too soon set them to what I thought they should be, for the worse (relative to day before yesterday).

On a side note, I can't figure out how that happened. So the settings I found yesterday are a combination of settings I definitely used somewhat longer ago, but with having the Audio PC (Stealth III) rebooted regularly (say about daily), the settings always being nicely saved to the SSD, on Sunday (day before yesterday) I rebooted too, but this time in "hard" fashion because I couldn't connect to it because of something I did wrongly on Saturday night. So Sunday afternoon the settings were not saved with for result it got hold of older settings ? I wouldn't know how it did that, but it did.
The settings are auto-saved each day, also to permanent storage once the RAM-OS is normally shut down. So I set myself to the task of finding out when actually these settings we saved in "old" fashion for the first time.

As you can imagine this confuses highly. I don't know about you, but for me such settings are an evolution of relative judgments and findings. And if suddenly this has reverted to something you don't know, then you are, well, confused. Had I been listening to these settings for one day more ? two days ? a week ?? This unknown alone is sufficient to be without ears. And that happened to me yesterday.

Peter

PS: Below we can see how there's a gap between the 9th and the 13th of May (Sunday). So the Mach III has been running for 4 days from of the 9th, until I had to hard reboot it and it couldn't save the settings files to permanent storage of the lacking days. But these are the saved settings, which doesn't readily tell what I used after any boot up. Anyway, the file of May 13 should contain the wrong settings already, which settings are a copy of the real settings used when XXHighEnd was started the first time after the reboot, Sunday afternoon, 1:38pm. It seems logic that the settings of the 9th were used (loaded) at that moment. Btw, it resembles my (previous post) mentioned "4 days" of burn in of the newly built PSU. So after the 4th day suddenly all "snapped in". Yeah, but with different settings than the 3 days prior to that ...


* MachIIIAnomaly01.png (34.81 KB, 616x228 - viewed 359 times.)
Logged

For the Stealth III LPS PC :
W10-14393.0 - August 18, 2018 (2.10)
XXHighEnd Mach *III* Stealth LPS PC -> *Xeon Scalable 14/28 core* with Hyperthreading On (*set to 10/20 cores*) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = *140.19*  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = *35* / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> *Lush^2 A: B-W & Y-R, B: B-W* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
Nick
Audio Addict
***
Offline Offline

Posts: 725


View Profile
« Reply #36 on: May 15, 2018, 12:48:40 pm »

Hi,

I dont want to distract from Peters Mach III thread but I can offer some supporting comments on the sound of a 6130 16/32 core system.

Sound quality is EXCEPTIONALLY good, its an absolutly outstanding device for music replay. My old PC which was an x99 14/28 core with some modifications to its clock and power supply arrangements,  which set a very very high bar for SQ. To have the 6130 play so well in comparision is a major supprise.

The 6130 was not a straight forward build, as Peter says above server boards are certainly a new environemt, and it is expensive. I absolutely agree with what Peter says about sound quality here. Having been through a build, IMHO the Stealth III looks like very good value to me.

Regards,

Nick.


Logged

Audio PCs
PC No. 1: 621 motherboard Xeon 32 thread CPU.
PC No 2: X99 motherboard Xeon 28 thread CPU - modified motherboard.

 w10 10393 RAM OS => XX V2.09 / adaptive mode / XX buffer 4096 / NOS USB driver v 1.02 buffer 16ms / Q1,2,3,4,5 = 30,-,1,1,0 / xQ1 =40 / unattended / SFS 5.19Mb / memory straight continuous / system clock 15.0ms / Threadprio RealTime / Playerprio Low / CPU scheme 3-5 / 16x Arc Prediction / Peak Extend off / Phase alignment off / Phase off  / XTweaks : Balanced Load 43 / Nervous Rate 50 / Cool when Idle n/a / Provide Stable Power 0 / Utilize Cores always 1 / Time Performance Index = Optimal / Time Stability  On => Lush USB cable => modified NOS1 USB DAC => no pre amp => Orelo active horn loudspeakers with modified bass channel DSPs.

Music server: X99, Xeon 16 thread PC.

System powered from two 3kva balanced mains isolation transformers.
Michael_B
Audio Loudspeaker
*
Offline Offline

Posts: 16


View Profile
« Reply #37 on: May 15, 2018, 01:05:03 pm »

Hi Nick,

So you received a Stealth III? This one would be the 2nd one I believe.

Michael
Logged

Audio PC: Mac mini Mid 2011 / 16GB / 240GB SSD /  Uptone JS-2 power supply / Server 2016 as OS / Audiophile Optimizer / Fidelizer / Process Lasso / Primo RAMDisk / HQPLayer / Euphony (linux)
DAC 1: NOS1a/G3, B75 (2x) / Blaxius / Clairixa and Lush.
DAC 2: Chord Hugo / modified power supply: now runs off 2 x UltraCapacitor power supplies LPS1 from Uptone Audio.
Preamp.: none
AMP: AVM SA8
Speakers: LEEDH E2
Sub: LEEDH 20.1
AC Power: 3 x Dedicated lines for: Amp., PC + DAC and Sub / 2 x Topaz Ultra Isolation transformers 1.8kva (ea) (0.0005pf) for the digital and amp.
Others: Uptone ISO Regen / ICRON USB Extender 2304GE-LAN / TP-Link MC110 Fiber Media Converter
Nick
Audio Addict
***
Offline Offline

Posts: 725


View Profile
« Reply #38 on: May 15, 2018, 01:22:42 pm »

Hi Nick,

So you received a Stealth III? This one would be the 2nd one I believe.

Michael

Michael,

No I dont have a stealth III. But felt it useful to comment on 6130 sound quality.

Also having been through a build and knowing what bespoke componemts and work Peter puts into the Stealth II, (but not of course the Stealth III) I genuinly think that the physical package and most importantly the music quality make the Stealth III great value.

KR,
Nick.
Logged

Audio PCs
PC No. 1: 621 motherboard Xeon 32 thread CPU.
PC No 2: X99 motherboard Xeon 28 thread CPU - modified motherboard.

 w10 10393 RAM OS => XX V2.09 / adaptive mode / XX buffer 4096 / NOS USB driver v 1.02 buffer 16ms / Q1,2,3,4,5 = 30,-,1,1,0 / xQ1 =40 / unattended / SFS 5.19Mb / memory straight continuous / system clock 15.0ms / Threadprio RealTime / Playerprio Low / CPU scheme 3-5 / 16x Arc Prediction / Peak Extend off / Phase alignment off / Phase off  / XTweaks : Balanced Load 43 / Nervous Rate 50 / Cool when Idle n/a / Provide Stable Power 0 / Utilize Cores always 1 / Time Performance Index = Optimal / Time Stability  On => Lush USB cable => modified NOS1 USB DAC => no pre amp => Orelo active horn loudspeakers with modified bass channel DSPs.

Music server: X99, Xeon 16 thread PC.

System powered from two 3kva balanced mains isolation transformers.
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15242



View Profile Email
« Reply #39 on: May 15, 2018, 04:03:35 pm »

I dont want to distract from Peters Mach III thread but I can offer some supporting comments on the sound of a 6130 16/32 core system.

Hi Nick ! what a nice surprise that is !
I was waiting for you to dive into this, but it is a surprise that you are so soon with it. I mean, I felt a little bit "obtrusive" with my posts about the Stealth III, with you at about the same day telling about your latest clock etc. endeavours. And I really did not want to let "snow that under", but what to do.

For Michael and others :
Nick is quite keen on improving SQ with the soldering iron, which often requires a lot of guts, certainly the knowledge, huge time (as we know) and also some spare $. It wouldn't be the first time that Nick motivates me to try out something (including explicit pointers) in order to me next coming up with something "commercial" worked out by me, all free of charge by Nick. And aren't we all here to stimulate each other. This time it apparently worked the other way around to some degree, which is very very nice.

It is certainly not something for anyone else to try for himself as it may cost you a lot of failure and thus money. The environment is really different and I am glad hat Nick confirmed this a little.

Nick, congratulations with your sound ! yes, it really is something. grazy
I am sure you located the clock locations already. And I can tell you (seriously / genuinely) : the sound really feels as if your clock mods were applied. How that goes by description only is a bit beyond myself, but I guess this is about "interpretation" of how you write.

Best regards and thank you for sharing !
Peter
Logged

For the Stealth III LPS PC :
W10-14393.0 - August 18, 2018 (2.10)
XXHighEnd Mach *III* Stealth LPS PC -> *Xeon Scalable 14/28 core* with Hyperthreading On (*set to 10/20 cores*) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = *140.19*  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = *35* / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> *Lush^2 A: B-W & Y-R, B: B-W* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
Nick
Audio Addict
***
Offline Offline

Posts: 725


View Profile
« Reply #40 on: May 15, 2018, 07:47:15 pm »

Hi Peter,

This definaly kicked off for me because of your posts  Happy.

At the time you started this thread, my X99 clock project was finished. It took close to 2 years and literally 100s++ of changes, the sound was really something very special but there was not much else I could think of to keep me busy on the x99 board so I was starting to look out for something new to try. Then you posted on the  Intel Scalable Processors and 601 PCH chipset. Our systems are very similar so I got straight away that you were really onto something. A look at the PCH 601 data sheet ( 3000 pages :-) ) really helped decide to do the build. Besides all of the very important for a music server archtecture features you mentioned in this thread, the 601 chipset clock system has changed from the x99. Its like someone in Intel thought "lets give those audio guys a helping hand"  Happy. On the face of it on paper back then and now listening to the built PC I think you are right in that much of the X99 clock benifit can already be heard in the 601 based build...... billiant !

Regards the sound quality, IMHO the 6130 build is very step from the Stealth II (which is quite a suprise given the excelant sound quality of the Stealth II). My clock bought a lot more to the table on an x99 system and now the 601 / 6130 takes a big step forward from there.

I hope to hook up with an x99 with clock PC at my place next week to do some direct comparisons (unfortunatly I used parts from my x99 in the 6130 build so I cannot play them back to back right now). The focus is not so much on might the 6130 be such a big step forwards. It just is. But on comparing the effect the clock has as I think there may be a little goodness to be got with the 6130 system but I want to be sure before I risk wrecking some expensive components  Shocked

Regarding the 601 / 6130 server build, yes its certaily different to a desktop PC, even now I have some instability that I have to sort out. I know how well rounded and complete your stealth builds are and stand by my comment that the Stealth III is going to be good value given its content and the SQ performance.

Definatly enjoying the music just now  grazy
KR,
Nick
Logged

Audio PCs
PC No. 1: 621 motherboard Xeon 32 thread CPU.
PC No 2: X99 motherboard Xeon 28 thread CPU - modified motherboard.

 w10 10393 RAM OS => XX V2.09 / adaptive mode / XX buffer 4096 / NOS USB driver v 1.02 buffer 16ms / Q1,2,3,4,5 = 30,-,1,1,0 / xQ1 =40 / unattended / SFS 5.19Mb / memory straight continuous / system clock 15.0ms / Threadprio RealTime / Playerprio Low / CPU scheme 3-5 / 16x Arc Prediction / Peak Extend off / Phase alignment off / Phase off  / XTweaks : Balanced Load 43 / Nervous Rate 50 / Cool when Idle n/a / Provide Stable Power 0 / Utilize Cores always 1 / Time Performance Index = Optimal / Time Stability  On => Lush USB cable => modified NOS1 USB DAC => no pre amp => Orelo active horn loudspeakers with modified bass channel DSPs.

Music server: X99, Xeon 16 thread PC.

System powered from two 3kva balanced mains isolation transformers.
manisandher
Crazy Audiophile
****
Offline Offline

Posts: 2041

from-first-principles.com


View Profile WWW Email
« Reply #41 on: May 15, 2018, 08:27:11 pm »

I hope to hook up with an x99 with clock PC at my place next week...

I've never been called 'an x99 with clock PC' before Wink

Looking forward to coming over Nick... should be a very interesting day.

Mani.
Logged

Audio PC: Phasure Mach II *with clock upgrade* / Win 14393.0 on RAM-OS / controlled by RDC, no GPU card installed / connected directly to music server
XXHighEnd: 2.10 / Minimize OS / Engine#4 Adaptive / DB=4096 / Q1=30 / xQ1=40 / Q3,4,5=1 / SFS=5.19 / XTweaks = 52, 10, 0, 0, 1 / Straight Contiguous / Clock Resolution = 15ms / Scheme 3-5 (low/realtime) / 8x Arc Prediction (switch #5 'up/off') / Unattended
Main System: mobo USB3 port -> Lush^2 -> Phasure NOS1a B75 G3 -> First Watt F5-cloned mono amps -> Tune Audio Anima horn speakers
Office System: Lush -> Phasure NOS1a G3 -> BD-Design BD30-SPR mono amps -> Tune Audio Marvel horn speakers
Nick
Audio Addict
***
Offline Offline

Posts: 725


View Profile
« Reply #42 on: May 15, 2018, 09:07:35 pm »

I hope to hook up with an x99 with clock PC at my place next week...

I've never been called 'an x99 with clock PC' before Wink

Looking forward to coming over Nick... should be a very interesting day.

Mani.

Haha, yes looking forwards to seeing you  Happy
Logged

Audio PCs
PC No. 1: 621 motherboard Xeon 32 thread CPU.
PC No 2: X99 motherboard Xeon 28 thread CPU - modified motherboard.

 w10 10393 RAM OS => XX V2.09 / adaptive mode / XX buffer 4096 / NOS USB driver v 1.02 buffer 16ms / Q1,2,3,4,5 = 30,-,1,1,0 / xQ1 =40 / unattended / SFS 5.19Mb / memory straight continuous / system clock 15.0ms / Threadprio RealTime / Playerprio Low / CPU scheme 3-5 / 16x Arc Prediction / Peak Extend off / Phase alignment off / Phase off  / XTweaks : Balanced Load 43 / Nervous Rate 50 / Cool when Idle n/a / Provide Stable Power 0 / Utilize Cores always 1 / Time Performance Index = Optimal / Time Stability  On => Lush USB cable => modified NOS1 USB DAC => no pre amp => Orelo active horn loudspeakers with modified bass channel DSPs.

Music server: X99, Xeon 16 thread PC.

System powered from two 3kva balanced mains isolation transformers.
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15242



View Profile Email
« Reply #43 on: May 17, 2018, 08:25:41 pm »


OK, I am listening for over 90 minutes now to a 14/28. There is just no comparison with the 10/20. It is just ultra-ultra fast and during the first seconds I right away heard back the sound of the 16/32.
There is no tweaking of any parameter needed to let it excel the music in inordinate fashion from any angle. I say it again :

Crazy

Let's keep in mind : the speed is in the higher regions here. The 10/20 has its speed in the mid. That is as interesting (but different) and a lot cheaper ...

Peter
Logged

For the Stealth III LPS PC :
W10-14393.0 - August 18, 2018 (2.10)
XXHighEnd Mach *III* Stealth LPS PC -> *Xeon Scalable 14/28 core* with Hyperthreading On (*set to 10/20 cores*) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = *140.19*  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = *35* / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> *Lush^2 A: B-W & Y-R, B: B-W* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 15242



View Profile Email
« Reply #44 on: June 01, 2018, 12:14:35 pm »

Hi all,

Yesterday I finally went back to the Mach II and could compare it with the Mach III - I am by now very well acquainted with.
Actually one term suffices : Mach II is a rough machine.

In comparison the Mach II exhibits rough sound and nothing of the finesse the Mach III happily shows. Speed ? there is no speed. Interest in the music playing ? not existent. It just is not interesting. OK, if you like a babbling crook and like to doze off during listening, don't get yourself the Mach III. But otherwise ? ...

All right. It is more interesting than what I just said because there's also an intriguing part. Maybe read this first.

The largest part of the discussion we had last night is why Ciska has more difficulty with the sound of the Mach III. Well, the conclusion is, it's a brain thing and perhaps the not being interested really in the "super speed" the Mach III exhibits. So yesterday we had one example of a known album - Yello's Toy, and it's plain flat in comparison. Mind you, flat in the sense of flat "tones" without variation in them. Say as you know it from Yello because you wouldn't know otherwise. But now come over here and have a listen ...

The difference is for example in the voice. Know Mark Knopfler ? good, then you know nothing much. But well, you are entitled to know that Dieter Meier (Yello) surpasses Mark Knopfler on the deep and crackling voice (imagine the combination) by a mile or two. Inordinate (I could call it "crazy" once again). But this is just a voice and while you out there are used to Mark Knopfler and don't care about Dieter Meier being worse on it, you surely would care if Mark Knopfler started to sound as Chris Isaak. Right ?
So now you know how the Mach III sounds. No wait, what it can do.

Only two days ago I was fighting with the Crack Detection in XXHighEnd because one album kept on tripping on it (in each of its 4 tracks at some point). Might you be interested : 4T Thieves - The Digital Lagoon.
So what's up with this one ? well, according to the trip-setting of Crack Detect, more than 50 (100%) subsequent level (voltage) changes from close to the max of minus and the max of plus. Thus imagine your sound level in the room, and that about the loudest sound exhibits from deep silence to the maximum and that more than 50 times in a row at a sampling speed of 705600 time per second and this thus happening within 1 / 705600 x 50 = 0.00007 second. Thus, within this 0.00007 second the volume changes from min to max 50 times and ... you will be able to perceive it. Or at least it does something to the sound which is audible (to a trained ear ?) and a kind of "obviously" it was made like that (it is just in the music).
Btw, combine the title of the album with this ("The Digital Lagoon").
Anyway, when I was working on the weakening of the Crack Detect Tripping (which was set to 16 of these subsequent samples and is now at 100 so this album can be played), I really was focusing on what I could hear of it, knowing that it is in there. I envisioned (that I heard) this blasting fast on/off sound (generally by me referred to as crackle but better to call it sizzle at this frequency).

And that's how Yello is completely flat or maybe dead via the Stealth II in comparison. Try to envision the more complex sounds of a synthesizer (most modern synths can't even produce pure sines) and how the individual build up of such a sound now is audible. All crackles - all sizzles. All rattles when it is more about the lower frequencies. There's a tonne of more interest because of this in everything.

But now Ciska ...

What we tried to figure out yesterday is whether it could be possible that one (or at least she) can't cope with this;
Let's say that all what's happening also has to be processed by the brain. Let's say that Mark Knopfler could come across as tiring somewhat, so what if this exhibit of a voise gets even worse ? Oh, I like it the best, but maybe if you like to doze off you can't because of this. Or, maybe if you want to read a book at the same time, you can't. Too much processing required.

Also interesting is that at least for me it is some other way around. I think I said in this topic somewhere that never mind all what's going on, it does not happen in a way that it distracts. The contrary, it is all more natural and therefore requires less macro attention. Otoh at the micro level all happens and if it gets constipated there the whole thing blows ?

High Resolution link Phasure Mach III Audio PC 15



What was and remains interesting is how it could happen that with the "dumb" idea of making a power supply (the one for the Stealth III) super fast, the whole audio representation indeed becomes that fast. I mean, it actually can not be so that something of the audio itself is processed faster, like we can say this of the mains vs battery (a battery not being fast at all) as a supply to the power amp. Nothing of that order plays a role here. However, and as I told earlier in the topic, I created a super stable supply (as my measure of "speed" - it responds instantly) and now it looks like that this does a few things in the realm of "control" of the clock lines; I forgot the subject myself by now, but a MotherBoard and the fast CPU's of these days, anticipate skewing of the clock lines and take measure on it, while a super stable PSU (including the draw on it !) may incur for those measures not being tripped. And next is : less noise. And from noise by now we sufficiently know what its implications can be, especially in the digital domain(s).

In the next days I will try to focus on how Ciska reacts to the level (volume) and how loud I actually play. So, it is also good to acknowledge that I indeed play way louder with the Stealth III just because the average SPL is lower. How does this work ? well, actually you can hear it by means of the spades of air now being present. Air = literally no sound. It is the gaps between two "cycles" so to speak. No smear - nothing. Now if these "cycles" connect, the general level becomes higher and the SPL (which is a representation of "average") is thus higher. However, the spikes themselves (the maxing of the "cycles") is way louder. This too is audible as this is the sparkle and such and how the sound crackles through the room in s(c/q)uary format. So what's louder ? one canon shot in your room, of 1000 gun shots which add up to the same average SPL ? I know ... it's the gun shots.
So I could be playing too loud.

Blablabla ?
yes


Peter
Logged

For the Stealth III LPS PC :
W10-14393.0 - August 18, 2018 (2.10)
XXHighEnd Mach *III* Stealth LPS PC -> *Xeon Scalable 14/28 core* with Hyperthreading On (*set to 10/20 cores*) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 30/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = *140.19*  (max 140.19) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / Stop Desktop, Remaining, WASAPI and W10 services / Use Remote Desktop / Keep LAN - Not Persist / WallPaper On / OSD Off (!) / Running Time Off / Minimize OS / XTweaks : Balanced Load = *35* / Nervous Rate = 10 / Cool when Idle = n.a / Provide Stable Power = 0 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / *Arc Prediction Filtering (16x)* / Always Clear Proxy before Playback = On -> USB3 from MoBo -> *Lush^2 A: B-W & Y-R, B: B-W* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius Interlink -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere (also from the PC).

For a general PC :
W10-10586.0 - May 2016 (2.05+)
*XXHighEnd PC -> I7 3930k with Hyperthreading On (12 cores)* @~500MHz, 16GB, Windows 10 Pro 64 bit build 10586.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1 / Q1Factor = 1 / Dev.Buffer = 4096 / ClockRes = 1ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.10  (max 60) / not Invert / Phase Alignment Off / Playerprio = Low / ThreadPrio = Realtime / Scheme = Core 3-5 / Not Switch Processors during Playback = Off/ Playback Drive none (see OS from RAM) / UnAttended (Just Start) / Always Copy to XX Drive (see OS from RAM) / All Services Off / Keep LAN - Not Persist / WallPaper On / OSD On / Running Time Off / Minimize OS / XTweaks : Balanced Load = *43* / Nervous Rate = 1 / Cool when Idle = 1 / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = *Optimal* / Time Stability = *Stable* / Custom Filter *Low* 705600 / -> USB3 *from MoBo* -> Clairixa USB 15cm -> Intona Isolator -> Clairixa USB 1m80 -> 24/768 Phasure NOS1a 75B (BNC Out) async USB DAC, Driver v1.0.4b (4ms) -> Blaxius BNC interlink *-> B'ASS Current Amplifier /w Level4 -> Blaxius Interlink* -> Orelo MKII Active Open Baffle Horn Speakers.
Removed Switching Supplies from everywhere.

Global Moderator
Pages: 1 2 [3] 4 5 6  All
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1 RC2 | SMF © 2001-2005, Lewis Media Valid XHTML 1.0! Valid CSS!
Page created in 0.07 seconds with 20 queries.