XXHighEnd - The Ultra HighEnd Audio Player
April 16, 2024, 11:13:53 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: Mobo main clock replacement...  (Read 84587 times)
0 Members and 1 Guest are viewing this topic.
christoffe
Audio Addict
***
Offline Offline

Posts: 762


View Profile Email
« Reply #45 on: January 12, 2015, 01:45:36 pm »



Very very very  Happy .... and more to come.

Nick.

Hi Nick,

the conclusion will be, that we have to select an appropriate/suitable motherboard (for all of us) and send this to you to implement your sophisticated modifications.

Joachim
Logged
Nick
Audio Addict
***
Offline Offline

Posts: 763


View Profile
« Reply #46 on: January 15, 2015, 11:04:28 pm »

So its looking like various clocks within the PC interact with sound quality and that improving these clocks significantly improves sound.

I thought to run the clock waveform of the high quality clock through a basic mask test on my scope.

What the test does is first record a continuous waveform of the CPU source clock into the oscilloscope's memory for about 1/1000th of a second so in total about 25,000 clock waves are recorded. It is then possible to apply a test "mask" (see the blue area in the trace below which is incidentally a massive deviation for a clock signal)  to all of these waveforms so that the analysis captures each time a wave form shape deviates from unshaded area into the blue mask.

The wave form of a clock should be rock solid and not deviate all. When the clock being tested is captured without the PC turned on, sure enough the clock always stays within the mask and no errors are recorded.

With the PC turned in it is a very different story. The trace shows the capture analysis. The red circle shows one of the wave form errors picked up by the mask. In clock terms the wave form disruption is very large. That is really going to add jitter and phase noise.

Look now at the blue and red dotted lines at the bottom of the trace (see the lines between the white "1" and "2" in the image at the bottom). The trace from 1 to 2 in total represents about 1/1000th of a second in time. Every red dot on that line represents an error wave form that has strayed into the blue mask. This capture which is a typical one shows 444 errors in 1/1000 of a second. This equates to an average error rate just under 2 clock beats in every 100 !!

(Just as a note the trace shape shows reflections, due to the measurement point of the signal the wave form injected to the PC mobo is better.)

So this is driving thinking on further modification to the PC along another direction that I had hoped would not be necessary. I guess it is a case of how far you want to go, the sound quality so far achieved is very very nice, its just when you see this it points towards the possibility of more improvements.


Nick.



* CPU Clock Analysis.jpg (123.36 KB, 800x480 - viewed 1174 times.)
Logged

Audio PC

C621 motherboard, Xeon 40 thread CPU.

 w10 14393 RAM OS => XX V2.10 / adaptive mode / XX buffer 4096 / NOS USB driver v 1.02 buffer 16ms / Q1,2,3,4,5 = 10,-,1,1,1 / xQ1 =15 / unattended / SFS 0.69Mb / 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 35 / Nervous Rate 10 (or15) / 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 28 thread PC.

System power two 3kva balanced tranformers with dedicated earth spur.
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 16837



View Profile Email
« Reply #47 on: January 16, 2015, 11:23:51 am »

Great stuff Nick. And thank you for sharing of course !

There's a word of caution though ...

With the clear notice that you announced part of it yourself already, I would be not so sure that you're looking at a signal decent enough to apply this analysis. This is two-folded :

1. The bandwidth of the analyser looks as if not high enough;
2. The grounding is the probe possibly is not in good fahsion.

Ad 1.
This is how you won't receive a square, which is the signal coming from the clock.

Ad 2.
This is how ground bounce may be injected in your analyser's signal.

In my view it can still be possible to observe as you do if only #1 is in order but #2 requires thorough real-time looking at the way the signal changes (when detected as error).

Of course, the "ground bounce" I'm talking about can be there internally (MoBo) just the same and this is how it requires real-time observation which I can't do (not being next to you I mean). So what it would require is "knowing" what can do it (from the MoBo's point of view) like for example a timer going on. Could be the Clock Resolution as we apply it in the software, but even with that in mind I don't think I would be able to do this because too indirect and complicated.
What I'm saying is that this wouldn't go for one such a "static" analysis, and that it merely requires a few different setups and then draw conclusions out of those. So like you suggested "one clock is better than the other". So, how ? And how does it show in the wave form ? And didn't you see already that by only changing the probes (detach and reconnect) changes the wave form ? I think you did.

Back to the bandwidth, I think you can also see how the shape of the individual cycles are (or can be) repetitive. Now this relates to the sampling rate of the analyser and here too, it requires real time observation. So think the analyser running on some clock rate and once in the xxx cycles all is the same again. Sadly the xxx can be millions, so again a tough thing to do (quite impossible, depending on what you see).

An example of what I mean could be the down-going slope in the first cycle in the rad circle, where the up-going peak will be detected as out of bounds. But now watch the remainder of the down-doing slope which is steeper than the others and now also runs out of bounds (near the bottom). And again in the middle of the dip where the next sample is now relatively higher because the previous one was captured too low. Just think that the average signal is represented, but the granularity of the X-axis (= time) is not high enough. How to combine this ? well, if the sampling rate (bandwidth) is not high enough, you'd get exactly this. First this peak is captured and because of that being at a too high level, the next one is additionally low (level). So because of the too low bandwidth the analyser detects an error here (could even be three for this one cycle) and you are happy with that ...

Of course you solve this by making the mask broader, as you btw already have done and now end up with 444 errors for the captured time period.
In the end you could be fooling yourself.
Additionally I'm pretty sure that if you set the mask broad enough, there will be no errors and then what. However, you could hope for a few per million which could be real error.

Lastly, I think I can tell you that no such thing as jitter can be observed from a wave form so deteriorated as this on one hand, and a time scale of 10ns (/div) on the other. Of course, you expressed it by means of error against the mask, but it is the same thing. However, still a bit dangerous to talk like this because such a "clock" can have 10s of ns of peak-peak jitter easily. But will depend on the specs for this (depicted by the MoBo) and you might know. Anyway the headroom you gave it looks to be ~5ns (black areas) and assumed the middle of that to be your target, any error implies 2.5ns deviation. And because it will happen at both sides of the set boundaries (other side for another cycle), it will be 5ns again, and more depending on the excursions.


Enough food for thought I think, and I can tell you that making a few comparisons like I suggested to make some better sense out of it, will require so much time (I think) that nobody is going to ask you to do it. But when you do I think we will be digesting it with much interest.

Regards,
Peter

PS, hints :
a. Always pick the very closest ground point you can find, near the signal (solder a small wire if needed).
b. Make the ground of the probe wire as short as you can. This means : wrap it right around the prope itself until left just enough for making the ground connection.
c. If you can incur for less/more error just because of changing b. somewhat, draw the conclusion that you can never get there (because it can always be better though up till physical constraints).



Logged

For the Stealth III LPS PC :
W10-14393.0 - July 17, 2021 (2.11)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 14/28 cores in BIOS and set to 10/20 cores via Boot Menu) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/0/0/*1*/ Q1Factor = *4* / Dev.Buffer = 4096 / ClockRes = *10ms* / Memory = Straight Contiguous / Include Garbage Collect / SFS = *10.13*  (max 10.13) / 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 = *62* / Nervous Rate = *1* / Cool when Idle = n.a / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / Custom Filtering *Low* (16x) / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^3
A: W-Y-R-G, B: *W-G* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius*^2.5* A:B-G, B:B-G Interlink -> Orelo MKII Active Open Baffle Horn Speakers. ET^2 Ethernet from Mach III to Music Server PC (RDC Control).
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: 763


View Profile
« Reply #48 on: January 16, 2015, 10:35:36 pm »

Peter hi,

Thanks for the comments, they are very helpful.

The bandwidth of the analyser looks as if not high enough;

I think the band width its fine for what the trace is highlighting. The scope used is just just a general 200mhz scope so it cannot get anywhere near the resolution needed to look at clock rise cycle jitter in the range of you super fast analoge sope. What is highlighting though is that the wave form distortions with the PC switched on that I’m seeing are so severe that they are easily picked up even on a scope of this type.

Even the basic mask function of the scope is easily triggered. As a note to exceed the mask shown in the trace the amplitude error has to be > +- 200mv or the time axis by > +-3 ns. That's bad repeating momentory external noise distortion on what is a low phase noise clock wave form.

The grounding is the probe possibly is not in good fahsion.

We are thinking the same thing, to be honest I was a bit embarrassed to show the a clock wave form as messy as the attached. I am measuring at the clock’s output before coax to the mobo but at the injection point on the Mobo it’s much better.

I had opened an edit on the post straight after submitting it and started to write an addition to say that there is a possibility that scope’s probe ground set up might influence the observed wave form, but I did not update in the end.  The probes are ground referenced to PE which isn’t good with so much SMPS noise about becuase of the PE loop between the scope and mobo PE wiring. It would be nice to use a differential probe set up for this measurement.  I do have a second basic scope that has a floating probe ground (it can be battery powered) so I’ll take a look with this and also probe at the Mobo.

It’s a shame I cannot show the real time trace being displayed. The scope has digital phosphor persistence to show wave intensity and 50k refresh per second rate. When looking at the trace in real time, the main wave is solid but when the PC is running there is a constant flicker deviation of the wave form momentarily visible traces around the main wave. Looking at the real time trace switch mode noise seems likely, I was sort of expecting to see this but the noise source need to be proven if possible. It would be great if it were then possible addressed it.

Best regards,

Nick.


EDIT

I just looked at the trace and can see now why you mentioned ground bounce. The frame I have selected happens to break the mask just after the wave dropes to zero. I should have mentioned that when paging through the captured frames the mask errors occur througout the clock cycle. At the top the bottom and rise and fall phases of the clock cycle. This is why i'm thinking PC SMPS noise might be the source. The reason to for doing the frame analysis was to try to get a feel for the error frequency and to see if it had a pattern. I could use a decent FFT  Happy 


Logged

Audio PC

C621 motherboard, Xeon 40 thread CPU.

 w10 14393 RAM OS => XX V2.10 / adaptive mode / XX buffer 4096 / NOS USB driver v 1.02 buffer 16ms / Q1,2,3,4,5 = 10,-,1,1,1 / xQ1 =15 / unattended / SFS 0.69Mb / 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 35 / Nervous Rate 10 (or15) / 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 28 thread PC.

System power two 3kva balanced tranformers with dedicated earth spur.
Nick
Audio Addict
***
Offline Offline

Posts: 763


View Profile
« Reply #49 on: January 20, 2015, 10:56:48 pm »

Quote
Looking at the real time trace switch mode noise seems likely, I was sort of expecting to see this but the noise source need to be proven if possible. It would be great if it were then possible addressed it.

I have identified a correlation of the clock distortion with supply rail noise within the PC now. In addition to this another mobo (not the same as mine but with the same ATx supply) I was working on over the weekend showed only small fraction of the clock distortion.
This is becoming very interesting.....

Listening opportunity was very limited on the second PC, but it did sound exceptionally good immediately. There could well be a link here.

Nick.
Logged

Audio PC

C621 motherboard, Xeon 40 thread CPU.

 w10 14393 RAM OS => XX V2.10 / adaptive mode / XX buffer 4096 / NOS USB driver v 1.02 buffer 16ms / Q1,2,3,4,5 = 10,-,1,1,1 / xQ1 =15 / unattended / SFS 0.69Mb / 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 35 / Nervous Rate 10 (or15) / 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 28 thread PC.

System power two 3kva balanced tranformers with dedicated earth spur.
christoffe
Audio Addict
***
Offline Offline

Posts: 762


View Profile Email
« Reply #50 on: January 20, 2015, 11:19:47 pm »


Mani and Paul are over this weekend so it will be interesting to see what they make of the sound..

Very very very  Happy .... and more to come.

Nick.


Hi,

nothing to report from the meeting?

Joachim
Logged
Nick
Audio Addict
***
Offline Offline

Posts: 763


View Profile
« Reply #51 on: January 21, 2015, 12:34:00 am »

Joachim hi,

It was a great meet up, Paul and Mani had chance to listen to what is happening here and I think got a good sense how these changes are moving the PC forwards. The listening included a number of problem tracks which have in the past seen my system really struggle to portray the music enjoyably.  I had spent a lot of last week adding some new components into the PC and tweaking the clock setup but as always seems to be the case, the sound was however significantly better on Sunday (the day after), more on this in a moment.

There was a very long session applying the upgrades to a PC which happened to be based on a mobo that is in current production, which is great. The results were very good indead, possibly better than on the ASRock Extreme 4M that I'm using. There were some measureable differences between the two boards in a key area. Performing the upgrades I spotted possible changes to my configuration regards how the clocks are connected. I tried these changes on Sunday and things improved significantly. I wish Paul and Mani had had the chance to hear this config, but that's the fun of these sessions, by trying things out you learn more.

Paul and Mani may want to post on what they heard here and also what has happened in their own systems since the visit  Happy

Best regards,

Nick.

Logged

Audio PC

C621 motherboard, Xeon 40 thread CPU.

 w10 14393 RAM OS => XX V2.10 / adaptive mode / XX buffer 4096 / NOS USB driver v 1.02 buffer 16ms / Q1,2,3,4,5 = 10,-,1,1,1 / xQ1 =15 / unattended / SFS 0.69Mb / 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 35 / Nervous Rate 10 (or15) / 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 28 thread PC.

System power two 3kva balanced tranformers with dedicated earth spur.
manisandher
Crazy Audiophile
****
Offline Offline

Posts: 2112

from-first-principles.com


View Profile WWW Email
« Reply #52 on: January 21, 2015, 04:05:24 pm »

... nothing to report from the meeting?

I just don't know where to start really...

Well firstly, Nick's system. I'd never heard it before and when he put it on it sounded very good to me. A bit heavy in the upper bass, and not enough top-end sparkle for my liking, but the dynamics were very good. Bass really thumped out, energizing the whole room. Listening position was important, and I ended up standing for most of the time as it sounded better to me. Paul and (obviously) Nick had heard previous iterations of Nick's system and apparently it is now a quantum leap over where it was previously. And most of this down to the NOS1a and Nick's latest mobo mods.

We then proceeded to modify my mobo. We (Nick, actually) replaced two mobo clocks with high-quality clocks (with linear PSUs). Once everything was done, we put the PC back together and hooked it up to Nick's system. And... very, very nice. Even better than Nick's PC, we all thought.

It was very late once I got back to my place, so no real listening done immediately. But I played with things over the weekend and now have a sound that I'm pretty happy with. The main difference that the mobo mods have made is to the 'smoothness' of the sound. It's still incredibly detailed, but right now there is no way anyone would think it's digital. There is an 'organic', 'wooden' (as opposed to metallic) quality to the sound.

I have a record dealer (hopefully) coming to my place on Monday to collect most of my LPs (I bought an Opera collection of about 3500 LPs a few years ago with the aim of getting into it... let's just say it hasn't turned out too well). This dealer ONLY listens to vinyl and hates digital. It'll be interesting to hear what he thinks of my system. I've been toying with the idea of putting a turntable in the room and pretending to play an LP when I'm actually playing a needle drop through the NOS1a... but I think this is a bit too sneaky and underhanded.

Right now I'm so happy with the sound I'm getting that I'd settle for it indefinitely if I had to. There is only one thing that I would change and that is the lack of deep, deep bass. But I'm now pretty convinced that this is due to the room, and nothing else. Who knows, maybe getting rid of 3500 LPs from the back wall with fix this?

Overall, it was a great day with Nick and Paul. My feeling is that Nick is definitely onto something here. But translating it into something feasible for most computer audio users is going to present a LOT of challenges. I mean, there's always the distinct possibility of screwing up the mobo, and even if everything works as planned, shipping mobos around with sensitive clock parts attached to then doesn't seem very practical to me. But I really hope Nick can 'commercialize' this in some way because it definitely works.

I'd like to say is a big thank you to Nick (and Paul) for implementing these mobo mods for me.

Mani.
Logged

Main System:
Phasure Mach III (Win 14393.0 on RAM-OS / controlled by RDC, / connected directly to music server / XXHighEnd 2.11 / Minimize OS / Engine#4 Adaptive / DB=4096 / Q1=10 / xQ1=15 / Q3,4,5=1 / SFS=4.00 / XTweaks = 34, 10, 0, 0, 0 / Straight Contiguous / Clock Resolution = 15ms / Scheme 3-5 (low/realtime) / 8x Arc Prediction / switch #5 'up/off' / Unattended) mobo USB3 port -> Lush^3 -> Phasure NOS1a B75 G3 -> 8m Blaxius^2 -> First Watt F5 mono amps -> Tune Audio Anima horn speakers
Office System:
Phasure Stealth II -> Lush^2 -> RME ADI-2 Pro FS R -> Neurochrome 286 mono amps -> Tune Audio Marvel horn speakers
christoffe
Audio Addict
***
Offline Offline

Posts: 762


View Profile Email
« Reply #53 on: January 21, 2015, 08:13:02 pm »

Hi Nick and Mani,

thank you to comment your impressions.
Let's see what is practicable for us non-electricians. (mobo + a kit + manual?)

kind regards

Joachim
Logged
Scroobius
Audio Addict
***
Offline Offline

Posts: 1170


View Profile
« Reply #54 on: January 21, 2015, 10:09:38 pm »

Quote
... nothing to report from the meeting?

I wanted to do more listening before posting but here goes. I will post again after these first impressions.

As Mani says we had a good day on Saturday at Nick's implementing mods to Mani's server. Nick wielded the soldering iron and me the drill and screwdriver. On Sunday back at home I set to on on my server. Not much fun with the lack of room in the Silverstone case. Last night I sat down for the first listen and I have to say this is a very significant improvement. In fact I would say one of the most significant improvements to my system (but I suspect it  would need all the previous mods i.e NOS1a and Blaxius  to work as well as it does). I will keep the best to last but from the very first note (a trumpet) smoother but with more detail - very noticeable.

Strings, wind instruments etc all now have a "bell like" clarity. More detail for sure but just soooo easy to listen to. Ultra clear you just want to keep listening and keep listening.

The "holy grail" for me is decent reproduction of massed strings. Good though they have been via NOS1a and Orelino's I had written off hearing decent strings through any hi-fi. Massed strings are really improved now. And I would say for the first time over the threshold of acceptability (for me anyway). Very easy to listen to (I have to leave some room here because there will be more mods to come ha ha).

But I kept the best to last. The biggest and most satisfying improvement is with poor recordings. I bought an album recently by Kara - "We Are Shining". Dance music with African rhythms. A mighty fine album to those of us that like to "boogie". But the sound is typical poor modern electronic there is a wall of hash / noise with the music. So much so I thought it was just part of the recording. Anyway it was bad enough to prevent me listening to it on the Orelino's (it actually even sounds bad on my study system and even in the car). After the server mods (rather incredibly) the "hash" has pretty much completely disappeared. And wow does it sound good now (relatively to how it sounded before!). I can properly sit down and listen to the album now. Or even better stand up and rock around to mighty fine dance music.

Now on to another bad recording that Peter knows very well. Elvis Costello - Wise Up Ghost. This is also transformed. I can now happily sit down and enjoy it properly. The dreadful SQ on vocals has all but been transformed with the slurring and lisping pretty much removed from Elvis's voice. And so it goes on with bad recording after bad recording turned in to recordings that can be enjoyed. Don't get me wrong bad recordings will never be good recordings but to be able to listen to them and they sound half decent is some result.

That is why this particular improvement is probably the most important I have heard in my system. BUT I reckon you need the "open window" of NOS1a and Orelino's to really hear the difference. That really is a testament to the quality of this system IMHO.

My ears tell me that these mods reduce "noise" not just change it. I have heard the mod in Nicks system with his server, Nick's system with Mani's modified server, my system with my server and my system with Nick's server. The character of the change is always the same, always entirely predictable.

IMHO this is a fundamental improvement not just a change to the SQ.

A word of warning though I was really reluctant to make these changes as it involves soldering the mobo. That is not for the faint hearted - but I am really glad I did!!!

By the way I have started experimenting making changes to the settings in XX (SFS etc) and I have to say it really does sound as though I can hear changes again (I could not before this mod and with NOS1a). I want to do more testing and will report on this in a subsequent post.

Cheers

Paul


Postscript - OMG I have just listened to Doll by Doll - Gypsy Blood. Another fine album but one that I had long ago consigned to the "cr*p SQ bin". It sounds really quite listenable now. Transformed from how it used to sound.

What a result!!!

Oh - I forgot to mention bass watch this space!!!




Logged

621 Xeon 6120 LPS PC  -> Xeon Scalable 16/32 core with Hyperthreading On (all cores active) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.69  (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 - Off / WallPaper Off/ 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-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelino Active Open Baffle Horn Speakers.
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 16837



View Profile Email
« Reply #55 on: January 22, 2015, 08:29:22 am »

When you guys meet up, are you always bringing whiskey for the host ?
Is there a way to join ? subscription maybe ?
Logged

For the Stealth III LPS PC :
W10-14393.0 - July 17, 2021 (2.11)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 14/28 cores in BIOS and set to 10/20 cores via Boot Menu) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/0/0/*1*/ Q1Factor = *4* / Dev.Buffer = 4096 / ClockRes = *10ms* / Memory = Straight Contiguous / Include Garbage Collect / SFS = *10.13*  (max 10.13) / 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 = *62* / Nervous Rate = *1* / Cool when Idle = n.a / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / Custom Filtering *Low* (16x) / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^3
A: W-Y-R-G, B: *W-G* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius*^2.5* A:B-G, B:B-G Interlink -> Orelo MKII Active Open Baffle Horn Speakers. ET^2 Ethernet from Mach III to Music Server PC (RDC Control).
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: 16837



View Profile Email
« Reply #56 on: January 22, 2015, 09:40:16 am »

Quote
The biggest and most satisfying improvement is with poor recordings. I bought an album recently by Kara - "We Are Shining". Dance music with African rhythms. A mighty fine album to those of us that like to "boogie". But the sound is typical poor modern electronic there is a wall of hash / noise with the music. So much so I thought it was just part of the recording. Anyway it was bad enough to prevent me listening to it on the Orelino's (it actually even sounds bad on my study system and even in the car). After the server mods (rather incredibly) the "hash" has pretty much completely disappeared.

What Paul describes here, actually would deserve its own topic, but alas, I think it would die because of no context then. So like often, just let's put it right in tne middle of that context, while actually it could be THE subject of all. But it can't because it's just a resulting phenomenon.

What I like to put forward, or make clear to *everybody* if possible, is what Paul means by this "hash" or "no noise".

Quote
My ears tell me that these mods reduce "noise" not just change it.

Quote
BUT I reckon you need the "open window" of NOS1a and Orelino's to really hear the difference. That really is a testament to the quality of this system IMHO.

With the clear notice "Who the heck is PeterSt then ?!" I like to emphasize that for a first time I see someone write about the real essentials which show to me that "he" has now achieved a system which shoud perform like my own. This sounds snobbish perhaps, but really what I like to put forward is all over praise about Paul, who now can do what I can do myself. Of course, this takes this well behaving system, but it also takes the minute observations which may take years and years to learn. But once you can do that, you can see (or will see) that it is a communication means. So in this case Paul communicates at least with me, but oh boy, how difficult is that for others. "Noise is reduced". Yeah, nice. But I understand and I'd like so much to bring that across to others. Thus my attempt :

It was only 5 weeks or so ago, that I could show this very phenomonon of "no noise" to someone who came over for an audition and who brought his fine Weiss;
Played through that for 45 minutes and random tracks were played. Most I did not know, but we ended with something I recognized, before we switched to the NOS1. Yes, "1" and not "1a".

We planned to play in reverse order and since the person had also brought his own PC, I had to ask him what that last track was. Oh, it was a track from Yello - Touch. Huh ? Ok, I carefully selected the album in question because 3 versions of it exist, so I was sure it was the same one. This already with the idea to be very careful because I would never have guessed it was that Yello album in the first place. Still I gues I have played the album 20-30 times. So I should know it.

Remark in between the lines : This album is one of the toughest to get going well. I am sure nobody really recognizes this as such, but it has something in it that enables the possibility that you wouldn't dare to play it, were it for showing off your system. When you don't know what is going on with it, you can feel that it has to be something about high resolution, never mind it is 16/44.1 (though one of the versions is 24/48). So there's very granular (silk though nicely sharp) highs in it, and I'd say that because this "high resolution" is in there throughout (so also the lower frequencies) something goes wrong. Today I can say "when not rendered well", but back in 2009 (when it was new) I thought that Yello had lost it and the recording was bad. I mean, their other albums are just fine.

And so we played the same track through the NOS1 and right from the start we both said that this couldn't be the same track. I added : but at least now I recognize it !
haha
All what this track is (and I should have remembered the track name) is a very cosy sound produced by synthesizers, mostly in the more lower frequencies and you could say it ain't nothing much (for listening to). Well, through the Weiss the sound was supported by one big hail storm and when I say "supported" this is wrongly put, because the hail storm was supported by some bass lines. Now whether Costello or Kara or (perceived) worse, we all (you all) must think in terms of more heavy rain or hail in the background to understand what I mean - what Paul means. Don't tell yourself the "oh sure", but envision that that rain.

Through the NOS1 that rain had almost vanished but do notice that especially with such a comparison, both versions will come across to you as "belongs to it". ONLY because you hear both versions and can compare, you can start to think about which is the real one. Nice ...
but neither.
So we then compared with the NOS1a (yes, "a" now) and it was to-tal-ly not there. And so it doesn't belong in there at all.

This is what's often described as "black background" but which won't tell you a thing unless you can (accidentally) compare in such extreme fashion. And really, when such HASH is in there, it makes a track completely unrecognizable if you are used to it not being there.

Because we accidentally found these extreme differences, I could play with different filter settings (XXHighEnd) and it really was seconds work (per trial) to perceive the difference and rank it. So say it took 2 minutes all together to find the best filter setting, just by listening to this amount of hash. If that isn't crazy then I don't know what is.


The things I quoted from Paul's fine post is all about how a better system too, can sneak in. So if our latest NOS1a tells me graduately that all sounds better as it seems, it can be all about such a simple phenomenon only. So, I do know that the "a" is way better to this aspect, but then I play "electronic" music for maybe this reason; it shows fairly easily (for better or for worse).

But what's that hash actually ? I think that depends, but let's look at this :

Quote
Strings, wind instruments etc all now have a "bell like" clarity.

This was my description and assumed Paul didn't copy my words (would be difficuilt anyway because of the slightly differerent context -> ) I described the transformation to the Blaxius as "all sounds as bells" because I could not find a better description. The bells are literally there as well, but otherwise I think what we try to describe is the harmonics in all. So mind you, "crystal clear" doesn't carry harmonics. Bells do and very many of them. And at high frequencies. And low. Sometimes our brain is very creative, unconsciously. Anyway :

What I have seen (and described) is that a rush of hash - al so normal to you, especially when you are used to the track for decades, can transform to a simple hit on a bell, when rendered well. In my case ? oh, NOS1a was OK, software is fine, amps are on par and speakers no problem. But the Blaxius wasn't in yet. So hash (or rain and hail and what misery not) turns into a bell. Yes.

When you read this, it testifies that you could still hold on to my Dutch (and thank you for that), but a message could be : when you see these things happening so clearly, you will start to see that we will never be ultimately there. I myself am in the stage of sheer reality, but I also already have heard Beatles tracks with a dozen or so African washboard players (or scratchers or better : dancers) while not any washboard is there in my normal system. We just have NO CLUE what's all in there, and of which recording engineers back at the time "and still today) will have thought "oh well". And I sm sooo serious you know. Just complete new instruments, only because that not-so-black background appears to be something that's an instrument while without the knowledge it only comes across as a "poor recording". But how long and how many times haven't I said it already : poor recordings do hardly exist. But for some it takes so much to reproduce them well. And if I survive it - let's look after ten years from now where we are. And I bet you ...


So guys, I wasn't telling that my system can do what these MoBo mods do. Not at all. Only that it is clear to me that when descriptions are like that from Paul, it most clearly shows an improvement which is for very real. We all will have "this" better than you, while you have "that" better than he. If this is all brought together we might get there. Whether 10 years is enough ? probably not. But let's continue and try. Our goal ? when something sounds nasty, there's more behind it and it has to be dug out. I'm telling you !

Great thanks, Nick.
Peter


Logged

For the Stealth III LPS PC :
W10-14393.0 - July 17, 2021 (2.11)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 14/28 cores in BIOS and set to 10/20 cores via Boot Menu) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/0/0/*1*/ Q1Factor = *4* / Dev.Buffer = 4096 / ClockRes = *10ms* / Memory = Straight Contiguous / Include Garbage Collect / SFS = *10.13*  (max 10.13) / 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 = *62* / Nervous Rate = *1* / Cool when Idle = n.a / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / Custom Filtering *Low* (16x) / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^3
A: W-Y-R-G, B: *W-G* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius*^2.5* A:B-G, B:B-G Interlink -> Orelo MKII Active Open Baffle Horn Speakers. ET^2 Ethernet from Mach III to Music Server PC (RDC Control).
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
Scroobius
Audio Addict
***
Offline Offline

Posts: 1170


View Profile
« Reply #57 on: January 22, 2015, 09:41:17 am »

Peter - its by invitation only  Happy  a bunch of flowers or chocolates usually suffices.



Logged

621 Xeon 6120 LPS PC  -> Xeon Scalable 16/32 core with Hyperthreading On (all cores active) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.69  (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 - Off / WallPaper Off/ 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-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelino Active Open Baffle Horn Speakers.
PeterSt
Administrator
High Grade Audiophile
*****
Offline Offline

Posts: 16837



View Profile Email
« Reply #58 on: January 22, 2015, 09:48:03 am »

Yes, but ...
swoon
Logged

For the Stealth III LPS PC :
W10-14393.0 - July 17, 2021 (2.11)
XXHighEnd Mach III Stealth LPS PC -> Xeon Scalable 14/28 core with Hyperthreading On (set to 14/28 cores in BIOS and set to 10/20 cores via Boot Menu) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/0/0/*1*/ Q1Factor = *4* / Dev.Buffer = 4096 / ClockRes = *10ms* / Memory = Straight Contiguous / Include Garbage Collect / SFS = *10.13*  (max 10.13) / 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 = *62* / Nervous Rate = *1* / Cool when Idle = n.a / Provide Stable Power = 1 / Utilize Cores always = 1 / Time Performance Index = Optimal / Time Stability = Stable / Custom Filtering *Low* (16x) / Always Clear Proxy before Playback = On -> USB3 from MoBo -> Lush^3
A: W-Y-R-G, B: *W-G* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> Blaxius*^2.5* A:B-G, B:B-G Interlink -> Orelo MKII Active Open Baffle Horn Speakers. ET^2 Ethernet from Mach III to Music Server PC (RDC Control).
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
Scroobius
Audio Addict
***
Offline Offline

Posts: 1170


View Profile
« Reply #59 on: January 22, 2015, 09:54:38 am »

Great post Peter.

I spent the whole evening last night listening to string quartets why? because I could not stop listening, the more I listen the better it gets. I will post more later when I have some time.

 Nice Nice

Logged

621 Xeon 6120 LPS PC  -> Xeon Scalable 16/32 core with Hyperthreading On (all cores active) @~660MHz, 48GB, Windows 10 Pro 64 bit build 14393.0 from RAM, music on LAN / Engine#4 Adaptive Mode / Q1/-/3/4/5 = 14/-/1/1/1/ Q1Factor = 10 / Dev.Buffer = 4096 / ClockRes = 15ms / Memory = Straight Contiguous / Include Garbage Collect / SFS = 0.69  (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 - Off / WallPaper Off/ 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-R* USB 1m00 -> Phisolator 24/768 Phasure NOS1a/G3 75B (BNC Out) async USB DAC, Driver v1.0.4b (16ms) -> B'ASS Current Amplifier -> *Blaxius^2 A:B-R, B:B-R* Interlink -> Orelino Active Open Baffle Horn Speakers.
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 0.096 seconds with 20 queries.