A further sound quality update.

When the second high quality clock was fitted to my PC motherboard, I could not find my spare SMB 50 ohm coax connectors so I soldered the coax to the lugs on the back of the clock pcb's SMB plug in order to get going.

Today a proper SMB 50ohm connector was installed to the coax to connect to the clock. Sound quality took another BIG, BIG step forwards. I expected things might improve but this is much more than expected. Every thing describe in posts above about sound quality improved again !! Could not be more pleased.

This is the upgrade that just keeps giving 

I'm happy to provide details of the upgrade to those who would like them, it's just a complete "no brainier" form a sound quality point of view.

Regards,

Nick.

Ps higher spec crystals arrived yesterday, wonder what they might do   

  really looking forwards to hearing them.

Peter hi,


You were right. I've been using a clock ratio of 12, the comments on sound above are based on this. I doubled the ratio to 24 today and the result was a more solid and expansive sound with richer tone. There is a slight loss of dynamic but still miles and miles ahead of where things were without the clock. Not a touch of hash at the faster speed which is what I would have got before at higher ratios. A very nice sound indeed.

Another interesting point. I use 10hz to 20khz sweeps to setup my subs by ear in room. Today when setting up the subs what before I had assumed were a series of very nasty room nodes though out the sweep were just gone. It's now more or less a smooth continuous sweep,  massively different to what was happening before. This I guess ties in with the notes above about the evenness of the sound presentation with equal emphasis across all aspects of the sound.

Kind regards,

Nick.

Mani hi,


I think we have all experienced the frustration of getting settings that are good at many things only to find that there always seem to be some highlight or low light in the sound cannot be overcome. So start tuning again and get a slightly different combination of good and bad points.

Listening here now it feels like there IS a solid floor at the bottom of the rabbit hole after all. I say that with a real sense of relief.

I'm looking forwards to seeing what you and Paul make of my PC over the Christmas break. Also listening to this PC play through Orelinos is genuinely a mouth watering prospect.

Regards,

Nick.

Hi,

I said I would post some more detailed impressions of the sound quality that the mobo clock upgrade produces. First before the sound impressions a couple of quick notes to bring the clock install up to date.

First on clock speed. I don’t have a way (for now) to quantitatively measure clock speed so I cannot say if the set "sweat spot" speed described is the exact nominal frequency of the PCs master oscillator, however trimming the clock speed on the replacement now fitted shows that clock speed is extremely important to sound quality. The magic in performance can be dialled in or out by moving the absolute clock speed just a few 10s or hundreds of cycles per second. The range of sound characteristics that are heard as the sweet spot is approached include booming bass but clear mids, poor dynamics but reasonable balance across the frequency range, generally good but sharp highs, in short a range of “problems” that in my system I have frustratingly suffered at different times. Its very easy to trim speed by ear I really knew when the sweat spot was reached !!

Next point is that with the replacement clock the PC boot cycle can be erratic. The PC may try to start multiple times before the POST process completes. It retries automatically so I just let it get on with it. Once it has booted it’s completely stable and I would swear that things like screen redraws are snappier than before (might be imagining it). If the boot becomes a problem, I think I know what may be happening and have some ideas about how to address it.

So the sound.

First thing to say is that once the sweet spot is found that this in IMHO is an absolutely all upside change in sound quality. Everything improves by very significant step.

Some items that stand out are:

Evenness balance tone and detail across the whole frequency. Ever had that experience where say mids are well projected but highs are slight sharp and bass powerful but slightly boomy  or for that mater many combinations of some parts the music playing well but one or two being recessed or having hash problems and letting the show down ? Well the evenness and balance that come with upgraded and tuned clock result in all frequency ranges work together in harmony without so many incorrectly "highlighted", reassessed, boomy or sharp elements in the balance.

Having said that, what does stand out is that all elements of the music stand out equally, by this I mean each instrument, voice and synthesised effect or spatial reference etc etc are given the right amount prominence and lime light to play in. Everything is writ large like it would be in live music and because things are not getting lost or highlighted the effect is that you hear musicians playing off each other and together. The energy in the music is again more convincing, snap in bass, dynamics and substance of voices, the distribution of energy in the overtones coming from a guitar or piano sound realistic and just easy on the ear to listen to.

Final major point I wanted to mention is rhythm. When the sweet spot is hit (with the clock speed), then the music seems to really speed up. The speed is the effect of the rhythm being portrayed to great effect. Before the mobo clock maybe one or two instruments or a voice would be marking out and really driving rhythm and then not always in an entirely  life like way. I guess I was just used to this and good tracks would bound along and compel my toe to tap. Now all the instrument voices and sounds are equally participating and with the right energy attack and tone, the result is that as you expect from a live performance, everyone is playing their part to drive the music forwards. The results is that the music boles along like the musicians are having bags more fun playing together. This last quality of the sound comes together only with the last fine tweak of clock speed, but you really know when its there !

Having a NOS1a is a great privilege; it is so transparent that the improved PC performance is absolutely made available by the NOS1a to listen to. All of the observations above are only possible because the NOS1a is so stupendously good and the description applies to the sound quality of the NOS1a as much as to the PC with clock mod. It may be that in other peoples systems the audio “mileage” from this change may not be so great, but for me there  is absolutely chance at all of going back.

Final thought is that I have read about others on the net replacing mobo master clocks with OCXOs and other fixed speed clocks. My experiment here leads me to believe that using a clock with the ability to trim its speed is very important. By having this facility it possible to improve the odds that you will hit the "magic" spot (speed). Otherwise it’s Clock Audio Roulette as to whether your lucks is in and the speed of the fixed speed replacement clock happens to exactly hit the sweet spot speed for your system.

Happy to provide more information if anyone wants.

Nick.

Hi,

Over the last couple of years its been my privileg to listen to some super systems owned by members of the forum and have been on a long personal journey seeking the best sound quality I can get here. In the many trips back and forth listening to systems and swapping components about I started to have the distinct feeling that near identical components, PCs and DACs did not always sound exactly the same ! In particular the level of harshness, bass performance and presence reproduced seemed to vary annoyingly as my system was often the one with the problem haha. I know there are many many variables with computer audio but this experience just kept happening.

From this and generally playing about far too much with PC settings clocks in the audio chain and researching clock gen and other parts of PCs as an idea has solidified.

I have repeatable experimental evidence that 4 oscillators in the audio chain (PC and DAC) directly impact sound quality.

+ Audio bit clock
+ USB receiver clock
+ USB host clock
+ Mobo master clock

+ Plus there are two more that I’v researched and intended to experiment with which I am almost sure will add to this list.

Its been generally accepted, (I bought into this as well) that if you re-clock at the DAC then upstream timing doesn't matter but this is just not born out by experiment. For each of the clocks listed above supply noise, phase noise, and clock speed, are parameters that impact sound quality. Generally it is not hard to improve on the standard x-tal set ups at these locations (exception being the audio clock Peter has in the NOS1a boy that is good !).
So without going into possible reasons / mechanisms that these clocks interact with sound quality (i have lots of ideas) and relying only on subjective observations relating to each of these clocks my thoughs / conclusions are.

1)   Music quality is absolutely hostage to performance  of the oscillators at these locations
2)   More worryingly, I wonder if normal variation between Mobo, USB and other clocks from card to card  or mobo to mobo (even where they are the same product) explains the variations I mentioned above between what should be similar system components.

If the aim is to sweat the last few percent of performance out of a system (and with the NOS1a this is defiantly an objective which is possible) then a holistic approach to manage these oscillators is needed. Otherwise IMHO there is too much of a luck factor that will determining if the sound will “fly” of have frustrating problems.

It’s like resonance / vibration in a vinyl setup.
Stylus tip => cantilever arm => Cartridge body  => Arm resonance => Arm Board  => platter bearing => desk suspension => equipment support . Resonance in any screws up sound. Maybe we shoud be thinking about replay chain oscillators in a similar way ?

Any thoughts ?


Nick.

Ps As I write sat I’m listening to dance music having just tuned the speed of the master clock on my Mobo (yes speed really maters) and the sound is like being in a really good night club.

Fralippo hi,

It's just a clock replacement nothing to get too worried about if approched correctly. I have test equipment to work out what is needed and all the relevant research done before hand. From my point of view the only risk I was concerned about was if the clock power-on timing sequence was important but it seems not. If the change had failed then the plan would be to revert back to standard or buy any replacement parts needed. Voltages etc are all fine.

I would not recommend doing this if you don't understand what you are doing or are bothered by the cost of replacement kit, but I'm fine with that.

Regards Nick.

Fralippo,

Thanks it's a very interesting post. Some real designer's working there.

I've looked at OCXOs but as you say they are expensive. There are some very good module oscillators about now at low cost. I think as mentioned before that there is a lower cost approach for USB and mob clocks that should give very good results without the cost of the dexas I have used.

kind regards,

Nick.

Peter

My PC  systm is running at a processor speed of 430mhz with max of 1.2gb. I'm in quiet time again so cannot try faster cpu just now, however when Itried high cpu clock ratios a couple of days ago sound was more consistent than before. Will give it a go. I'm expecting that faster cpu may be possible with good sound.

Thought a pic of the PC might be fun. Poor thing has its guts all over but the SOUND haha

Best,

Nick.

Hi,

Well there has been chance to listen to the Mobo clock upgrade today at normal volume. It doesn’t take two bars of music to realise this upgrade is a forklift sized upgrade to sound quality in my system. The synergy of the modified PC with the NOS1a is just absolutely superb.

Let’s not get too hooked on the hows and whys. I have plenty of thoughts about how the clock might be impacting the PC and music reproduction, but this definatly is not one to get too tied up on theory... it just works and how it works !!! 

I will post more details of what has happened to sound quality with the Mobo clock upgrade. For now I can say with no doubts at all that the ASrock Extreme 4M mobo has become a device in a different league for playing music. 

Nick.

ps two more locations in on the Mobo are on the shortlist for treatment 

Peter hi,

I agree with you there, I am also thinking "clock tree" and also about what this tree "drives". The base frequency for instance is multiplied upto 120 times to generate the cpu clock. If the trunk of this tree is not functioning well for audio, then it is sure  that the "branches" are effected at some level. The impact cascades up the tree and is likely to be felt differently in different branches.

What is known at the moment is that there is "influance", I simply report that subjectively there is a very large change in sound quality with each of the three clocks listened to so far. Setting aside for now if the change heard is better or worse, the mere fact that there is a big change shows that this clock through some linkage mechanisum is closley coupled to sound quality. We don't yet understand how and why, but it just is.

The change is so large it should easy to work out if this a "no brainer" or an interesting diversion. I should get to listen at high volume levels tomorrow so I'll feed back more on sound qualty good bad or mixed however it turns out.

Best regards,

Nick.

Hi,

Quick update. The lash up cheap module clock and LifePo4 battery are out of the PC now. It's been replaced by a spare dexa I modified to work with the motherboard.

My little one is in bed again so I've been listening at -60db and it sounds really good, full sound and disconnected from the speakers    defiantly improved on the lash up clock even at these stupid volume levels. Bass so far as I can tell is better but need to turn up the wick to be sure it might be a couple of days before I can do this. Very frustrating.....

I think I may have a  way to produce clocks for USB, and Mobo with -95db phase noise at 10hz and femto second jitter levels, at significantly lower cost than the dexas. Need to buy some parts and have a play.

Fralippo hi, sure will post some pics.

Nick.

I've spent this morning listening to the Mobo clock. It's very interesting listening. Mid range and highs have never sounded so good here in fact for clarity detail and tone beautiful is the word I would use. There is a smoothness (in the sense of real instruments and voices) that was not possible to fine with settings before. It's difficult to imagine the sound is coming from a digital recording.

I think the NOS1a is instrumental in letting these changes shine through there is so little (any ?) digital sound to it, so changes upstream just pass through it. Nice....

So now for the steps backwards. Well this is about energy and slam and definition in the bass. I had managed to get a great sound going after a lot of work on settings and tweaks throughout my system. The mobo clock is holding back a little of this energy and slam and bass resolution. My toe is not tapping quite so much at the moment as before. BUT I've heard this balance in the bass before, it was when using the same type of module oscillator I've lashed up the mobo with when working on USB clocks some time ago. It's the same signature in the lows and there was a straight forward fix for that....

So summary not a 100% hit first time, but this was not expected with the clock module used. However a very very significant improvement in SQ, and enough to make it worth working on the bass, but might have to spend some cash on a better clock or mod a 24.5mz clock this is unused at the moment.

Final observation clock multiplier setting in bios has very little impact on sound now. That's possibly pointing at the general way the improved improved master clock may work in the system.

 
Nick.

Now I have my NOS1a there is no reason to play inside my DAC anymore, Peter has just nailed it haha 

So feeling the need to play about today, I finally got round to changing the main clock on my audio PC's mobo. I did loads of research on this about a year ago and bought parts to make up a higher spec clock but just never got round to applying it. Well it went in today without any problems and the heart of my PC now beats on LifePo4 battery power.

Carnot say much about sound quality yet as my little one was in bed when the system went back together, but I can say that sound quality at XX vol of -45db (normally play at -30db) sounded extremely good. Will post more when the vol can be wound up. Fingers crossed for good things.

Regards,

Nick.

Hi,

I just wanted to close out this thread with a report of the best outcome.

The breakthrough noise that is described at the start of this thread is gone.

 

My system will now play at XX software volume control settings of -144 db and is silent (no distorted faint music plays anymore).

The fix for this long standing problem ? The NOS1a upgrade.

I've posted elsewhere about the musical magic of the NOS1a, in addition to this magic it's the fix for the breakthrough problem as well. I suspect this may not be by chance :-). That's one of the nice things about NOS1 ownership, Peter is just so determined to do things absolutely right. Very nice in a world where development of much equipment is influenced by bottom line return and not a pure focus on sound quality.

Many thanks Peter.

Kind regards,

Nick.