A novel way to massively improve the SQ of computer audio streaming

[austinpop]

On 8/20/2017 at 3:06 AM, AmusedToD said:

Just a quick question for those familiar with digital clocks. Would it be possible to modify an Ethernet switch by installing a BNC word clock input so that the switch could be connected to an external master clock (say something from dCS or Mutec)?

 

I am not a clock expert, but I think the answer is no.

 

Word clock is usually used to refer to clocks that are equal to or (sub-) multiples of the audio data rates of 44.1 or 48 KHz.

 

The sCLK-EX mods actually supply a 25MHz clock to the switch. As I understand it, this is neither a conventional word clock, nor a reference clock. At this point, I only know of SOtM providing this kind of clock improvement.

[AmusedToD]

6 hours ago, austinpop said:

 

I am not a clock expert, but I think the answer is no.

 

Word clock is usually used to refer to clocks that are equal to or (sub-) multiples of the audio data rates of 44.1 or 48 KHz.

 

The sCLK-EX mods actually supply a 25MHz clock to the switch. As I understand it, this is neither a conventional word clock, nor a reference clock. At this point, I only know of SOtM providing this kind of clock improvement.

 

Thanks @austinpop

[JohnSwenson]

On 8/21/2017 at 6:21 PM, austinpop said:

 

I am not a clock expert, but I think the answer is no.

 

Word clock is usually used to refer to clocks that are equal to or (sub-) multiples of the audio data rates of 44.1 or 48 KHz.

 

The sCLK-EX mods actually supply a 25MHz clock to the switch. As I understand it, this is neither a conventional word clock, nor a reference clock. At this point, I only know of SOtM providing this kind of clock improvement.

In this case the 25 MHz IS a reference clock. To my mind a reference clock is one that is used as the reference to a PLL, frequently a PLL used in a frequency synthesizer of some sort. In the case of an Ethernet switch chip the 25MHz feeds a PLL which generates the 125MHz clock which is actually used for the Ethernet data rate. (both 100Mb and 1Gb use 125 Mega Samples per Second on the wire).

 

So Yes it is being used as a reference clock. MANY clocks in digital systems are used this way. USB chips definitely do this (you feed 24MHz in and the data rate is at 480Mb/s), most of the clocks in a motherboard are generated by PLLs from lower frequency "reference clocks".

 

A fair number of DAC chips also have PLLs that generate the internal clocks from an external reference clock.

 

The Sotm clock board has a reference oscillator and a PLL based frequency synthesizer  that can generate 4 independent frequencies. If you want to you can also use it with an external reference clock. 

 

John S.

[austinpop]

1 hour ago, JohnSwenson said:

In this case the 25 MHz IS a reference clock. To my mind a reference clock is one that is used as the reference to a PLL, frequently a PLL used in a frequency synthesizer of some sort. In the case of an Ethernet switch chip the 25MHz feeds a PLL which generates the 125MHz clock which is actually used for the Ethernet data rate. (both 100Mb and 1Gb use 125 Mega Samples per Second on the wire).

 

So Yes it is being used as a reference clock. MANY clocks in digital systems are used this way. USB chips definitely do this (you feed 24MHz in and the data rate is at 480Mb/s), most of the clocks in a motherboard are generated by PLLs from lower frequency "reference clocks".

 

A fair number of DAC chips also have PLLs that generate the internal clocks from an external reference clock.

 

The Sotm clock board has a reference oscillator and a PLL based frequency synthesizer  that can generate 4 independent frequencies. If you want to you can also use it with an external reference clock. 

 

John S.

 

Hi John,

 

Yeah that does make sense. The only reason I was not using the term reference clock for the SOtM sCLK-EX was because of the way the mods are currently done. Currently, the sCLK modified device does not work without the external clock input. So these inputs aren't a reference for the clock - they are the clock. This may simply be an implementation choice. I can imagine it being easier to mod something to replace the internal clock, than to modify the internal clock to accept an external reference.

 

Contrast that with the use case of the so-called reference clocks like the Mutec Ref 10, where the target device functions fine without the reference input, but presumably improves with the application of the reference input.

 

Thanks for explaining the rationale for the 24 MHz and 25 MHz frequencies, for USB and Ethernet.

[auricgoldfinger]

Clock experts:  I am in the process of ordering a tX-USBultra with a master clock connector.  Is there any reason to choose a 50 ohm connector vs. a 75 ohm connector?

 

[Johnseye]

On 8/20/2017 at 7:30 PM, austinpop said:

This is going to be a long 2-3 weeks.

 

@Bamber's ISO-Regen is on its way back. My Ultra trifecta is on its way to May to be retrofitted with the SMB connectors, and a master reference clock input on the tX-USBultra.

 

In the meantime, I am using "just" this chain:

- Roon core bridged > mR > Lush USB > DAC.

 

This was state of the art just a year ago. Now - wow, how much further we've come!

 

Hope @Superdad hits his Aug 24 target for my own ISO-R that's on order, so I can at least get some of the magic back.

 

Hey there Rajiv, I was just talking with Alex and he recommended I get in touch with you about troubleshooting the new ISO Regen.  I don't have the time at the moment to type it all out, but I'll be back in an hour or so and available all weekend if you're willing to chat.  Hopefully I'm not on your ignore list

 

Thanks in advance

[austinpop]

4 minutes ago, Johnseye said:

Hey there Rajiv, I was just talking with Alex and he recommended I get in touch with you about troubleshooting the new ISO Regen.  I don't have the time at the moment to type it all out, but I'll be back in an hour or so and available all weekend if you're willing to chat.

 

Hmm, not sure how I can help, but I'm happy to chat. It'll have to be later this evening. I'll PM you, and we'll figure something out.

 

4 minutes ago, Johnseye said:

Hopefully I'm not on your ignore list

 

Not anymore! 

[Johnseye]

22 minutes ago, austinpop said:

 

Hmm, not sure how I can help, but I'm happy to chat. It'll have to be later this evening. I'll PM you, and we'll figure something out.

 

 

Not anymore! 

 

Thanks, I appreciate it on all counts.  I'm back home now and will watch for your pm.

[dgarretson]

On 8/23/2017 at 9:46 PM, auricgoldfinger said:

Clock experts:  I am in the process of ordering a tX-USBultra with a master clock connector.  Is there any reason to choose a 50 ohm connector vs. a 75 ohm connector?

 

That depends on whether your future 10mhz master clock has a 50 or 75 ohm output.   75 ohm is common in Pro studios, 50 ohm in lab equipment.  I ordered 75 ohm SOtM units to match the SRS Perf-10 master clock that I was already using to time an Esoteric DAC and an Antelope LiveClock frequency divider into a Tascam DA-3000 DSD recorder.   All have 75 ohm clock connections. 

 

Including the SOtM-modified D-Link switch, that's six devices synchronized to the same master clock.  So when choosing the connector impedance, it's wise to think about where your system is headed.   

[auricgoldfinger]

4 hours ago, dgarretson said:

That depends on whether your future 10mhz master clock has a 50 or 75 ohm output.   75 ohm is common in Pro studios, 50 ohm in lab equipment.  I ordered 75 ohm SOtM units to match the SRS Perf-10 master clock that I was already using to time an Esoteric DAC and an Antelope LiveClock frequency divider into a Tascam DA-3000 DSD recorder.   All have 75 ohm clock connections. 

 

Including the SOtM-modified D-Link switch, that's six devices synchronized to the same master clock.  So when choosing the connector impedance, it's wise to think about where your system is headed.   

 

Thanks for your thoughts.  I've done a little reading, and it seems that 75 ohm is the best approach for high end audio and video.

[austinpop]

6 hours ago, dgarretson said:

That depends on whether your future 10mhz master clock has a 50 or 75 ohm output.   75 ohm is common in Pro studios, 50 ohm in lab equipment.  I ordered 75 ohm SOtM units to match the SRS Perf-10 master clock that I was already using to time an Esoteric DAC and an Antelope LiveClock frequency divider into a Tascam DA-3000 DSD recorder.   All have 75 ohm clock connections. 

 

Including the SOtM-modified D-Link switch, that's six devices synchronized to the same master clock.  So when choosing the connector impedance, it's wise to think about where your system is headed.   

 

Thanks for that!

 

BTW - I opted to get 50 MHz for mine, because I know I will never be in the rarefied price point of the Mutec's, the SRS'es, and the Antelope's. 

 

I have my eye on the Cybershaft (http://www.cybershaft.jp/a-products/premiumop.html) units @hols mentioned, and of course the upcoming SOtM one. Both are said to support both 50 and 75 ohm, but I guess we will have to see what the price point of the SOtM is.

[dgarretson]

15 minutes ago, austinpop said:

 

Thanks for that!

 

BTW - I opted to get 50 MHz for mine, because I know I will never be in the rarefied price point of the Mutec's, the SRS'es, and the Antelope's. 

 

I have my eye on the Cybershaft (http://www.cybershaft.jp/a-products/premiumop.html) units @hols mentioned, and of course the upcoming SOtM one. Both are said to support both 50 and 75 ohm, but I guess we will have to see what the price point of the SOtM is.

 

Cybershaft seems like a good choice.  There is an Audiogon thread in which a few guys with high end DACs reported excellent performance and good customer service at Cybershaft.  IIRC the Cybershaft seller packages a used Rubidium core with his own power supply and chassis.  I'm not sure about the lifespan of a rubidium core, but they do need to be factory recalibrated once and awhile.

[auricgoldfinger]

1 hour ago, austinpop said:

 

Thanks for that!

 

BTW - I opted to get 50 MHz for mine, because I know I will never be in the rarefied price point of the Mutec's, the SRS'es, and the Antelope's. 

 

I have my eye on the Cybershaft (http://www.cybershaft.jp/a-products/premiumop.html) units @hols mentioned, and of course the upcoming SOtM one. Both are said to support both 50 and 75 ohm, but I guess we will have to see what the price point of the SOtM is.

 

If it's SOtM, you can expect it will be expensive. 

[austinpop]

48 minutes ago, auricgoldfinger said:

 

If it's SOtM, you can expect it will be expensive. 

 

What is expensive to you?

[austinpop]

BTW - SOtM completed my SMB retrofit and master clock addition in record time, so my gear is on its way back already! It should be here mid-week.

 

Only thing is - I have this uninvited guest named Harvey that showed up, and I have no idea when he'll leave. He's a bit of a wet blanket.

 

Hint - he's also a former hurricane.

[auricgoldfinger]

1 minute ago, austinpop said:

 

What is expensive to you?

 

I find the cost of their trifecta with all the necessary power supplies to be expensive. 

 

I'm guessing their clock will cost $3,500 +/-

[zephyr24069]

8 hours ago, auricgoldfinger said:

 

Thanks for your thoughts.  I've done a little reading, and it seems that 75 ohm is the best approach for high end audio and video.

I'm not sure what this is based upon...75-ohm may be the more common approach but 50-ohm is also a very valid approach in high end applications.  Further testament to this is Esoteric's centering this top-end products on having both 50-ohm 10 mHz master clocking capability as well as 75-ohm reclocking based approaches.

 

Apologies for delayed replies; I'm not on the forums all that much. Any specific questions,...ping me via IM and I jump in.

[zephyr24069]

6 hours ago, dgarretson said:

 

Cybershaft seems like a good choice.  There is an Audiogon thread in which a few guys with high end DACs reported excellent performance and good customer service at Cybershaft.  IIRC the Cybershaft seller packages a used Rubidium core with his own power supply and chassis.  I'm not sure about the lifespan of a rubidium core, but they do need to be factory recalibrated once and awhile.

Yes,...definitely; I'm one of them. You can find more details on Agon as well as WBF and Aficionado (AA).    There are quite a few users out there now.  The OCXO side of the dual clock or the OCXO-only units are my favorite for musical playback; they have superior low phase noise and harmonic distortion characteristics (which BTW the former is part of the calc for how much/little jitter a unit introduces...) over Rubidium implementations despite their parts-per-billion precision specs. (IMHO).    I've evaluated SRS, Antelope and other clocks and prefer the Cybershaft for price and performance.  It's phase noise and effective low-jitter specs are superior to others if you dig in far enough.   The only clock I have not tried is the new Esoteric Grandioso G1 and the DCS clocks as I don't have a DCS stack.  FWIW,..I know several Esoteric and DCS system owners that utilize a Cybershaft 10 mHz master clock as a baseline for their Esoteric and DCS clocks themselves with fantastic results....

[auricgoldfinger]

5 hours ago, austinpop said:

BTW - SOtM completed my SMB retrofit and master clock addition in record time, so my gear is on its way back already! It should be here mid-week.

 

Only thing is - I have this uninvited guest named Harvey that showed up, and I have no idea when he'll leave. He's a bit of a wet blanket.

 

Hint - he's also a former hurricane.

 

I was in Austin earlier this week taking my son to UT and got back to Houston just in time for the storm!  We have another few days of hairy flood risk.

[auricgoldfinger]

5 hours ago, austinpop said:

BTW - SOtM completed my SMB retrofit and master clock addition in record time, so my gear is on its way back already! It should be here mid-week.

 

Only thing is - I have this uninvited guest named Harvey that showed up, and I have no idea when he'll leave. He's a bit of a wet blanket.

 

Hint - he's also a former hurricane.

 

Since the Cybershaft has only 1 output, how would you connect it to your Trifecta?

[austinpop]

43 minutes ago, auricgoldfinger said:

 

I was in Austin earlier this week taking my son to UT and got back to Houston just in time for the storm!  We have another few days of hairy flood risk.

 

Oh hey - a fellow Texas parent! I'll send you a PM. Stay dry, my friend. You guys seem to be in for a deluge there.

 

34 minutes ago, auricgoldfinger said:

 

Since the Cybershaft has only 1 output, how would you connect it to your Trifecta?

 

Well, the output of the Cybeshaft (or other 10MHz reference clock) would go to the master clock input of my tX-USBultra. Remember, only the sCLK-EX board has the master clock input. The 4 outputs of the sCLK-EX just benefit from the reference clock.

[JohnSwenson]

Those of you getting 75 ohm BNC cables just make SURE you get a cable with a REAL 75 ohm BNC plug. A LOT of "75 ohm" BNC cables actually have 50 ohm plugs on the 75 ohm cables. I don't really know why this happens, probably the purchasing department notices that 50 ohm plugs cost a lot less than 75 ohm plugs and swaps them without telling the engineers. I have a hard time believing that designers working at cable companies don't know the difference between 50 and 75 ohm plugs.  PHYSICALLY they are interchangeable, you can plug a 50 ohm into a 75 ohm jack so you can't rely on that. A 75 ohm plug usually has much less insulation around the central pin, so you can usually spot them.

 

And this is NOT just audiophile companies, I found quite a few well known cable manufacturers that cater to pro video and audio companies use the wrong connectors as well.

 

I not going to give a list, I don't remember the names, just be sure you get the right ones.

 

John S.

[justubes]

Guys,  with all this attention and investment into reclocking,  beside the obvious improvements heard at each stage,  is there a reference e.g spinning the same tracks on cd to compare the actual sonics results?

 

The changing of clocks,  power supply and isolation will greatly contribute to improvements as well as differences in the presentation.

 

For my reclocking journey,  i  compare the same tracks on cd,  tidal and hdd. 

 

After my reclocking efforts,  i can not longer  distinguish differences reliably so now i just listen to tidal. 

 

I have then focussed on comparing strwams via tidal and it MQA versions(1st unfold) .  I was impressed with MQA as it sounding exactly what i was getting after each added reclock and heard the sonic characteristics of non MQA v.  MQA version and tweaking to the point that the differences or improvements between them are very similar as if going from a flac to wav file,  MQA being the Wav version.

 

I would be worthwhile have a sonic basis for the comparisons with each successive reclocking or upgrading clock cables. 

 

 

[simonklp]

On 2017/5/9 at 9:03 AM, austinpop said:

Ultra Trifecta Study Part 3: "I'm sorry I can't do that, DAVE"

 

Here are the results of a very interesting afternoon of critical listening, in the company of fellow CA'er Eric, aka @limniscate, who came bearing gifts. He brought his microRendu/LPS-1 and oh, by the way, a loaner Chord DAVE DAC. WHAT!

 

We listened to the trifecta with each component - the switch, the sMS-200 "ultra" and the tX-USBultra powered with its own LPS-1. After some baseline listening with the trifecta, we did the following experiments:

1. Break apart the trifecta, start with "just" the sMS-200 "ultra", and then evaluate the incremental impact of:
   • The tX-USBultra: The tX-USBultra added a level of refinement in the mids and highs, and some real solidity and oomph on the low end
   • The modded switch: This one just blew us away! It only took us 15 secs of listening before we looked at each other in amazement. The switch just opens up the image immensely.
     Overall comments: In terms of impact, the effects of the two are similar in magnitude, but complementary in nature. While we expected the tX to have an impact, we were just astounded by the effect of the switch. I don't know what stroke of genius impelled Roy to try this, but he is really on to something here. 
      
2. Swap out the sMS-200 "ultra" with the mR in the trifecta and compare. Specifically, we kept the trifecta in place, but replaced the sMS-200 and added in the mR
   
   Result: There was a startling drop in quality! This was rather surprising, as in the past we've found the sMS-200 and the mR to be quite close. What was even more disconcerting was the fact that despite the fact that we still had the sCLK-EX switch and tX-USBultra in the chain, the insertion of the mR seemed to collapse the image size, and reduce the focus and resolution. This lends credence to Roy's theory that the greatest benefits accrue when you cascade only the best clocks before the DAC. Clearly in the chain
              modded switch (with sCLK-EX)->mR->tX-USBultra (sCLK-EX), the clock quality is:
              best > good > best, and
   the lower quality clock in the mR seems to break the chain in some way.
   
   This does not bode well for the ISO-Regen in the path as well, but it will be most interesting to see how this pans out with Roy's and @mozes's experiments.
    
3. Compare the DAVE ($12k) with the Codex ($1.8k) - gulp!
   Result: The DAVE is clearly better. I found this to be more evident than Eric did.  We experimented with some settings on the DAVE, settling on DSD+ and HF Filter ON. I felt the DAVE just presented another layer of detail and clarity than the Codex. Eric felt the impact of the switch in the trifecta was at least as significant as the DAVE. We both agreed the effect of the SOtM stack was more dramatic than the the DAC difference between the Codex and the DAVE, which is notable in itself!
   
   Both of us couldn't justify the 6.5x difference in cost, but there is no denying the quality of the DAVE. Oh - to be rich!!

Overall, we were blown away by the additive effects of the Ultra trifecta. I would be hard pressed to pick a subset, but from a cost perspective, I could certainly recommend an sMS-200ultra with a modded switch, although the benefits of the full stack are undeniable and very compelling!

 

Sorry that I probably have missed some of the related messages. But I have doubt that whether this kind of improvement by using a better clock (e.g. sCLK-EX) only applies to cases where the Asynchronous USB Transfer mode is not properly implemented. Because under this transfer mode, the clock of the DAC is the master that the clocks of the upstream devices should not really affect so much.

 

For example, in case of playback using HQPlayer with NAA, the NAA isolates as much as possible at the software side by using a large asynchronous FIFO buffer. Is there really a substantial improvement by replacing the original clocks by better ones at the upstream side of DAC? Was there any experiment on this case? Thank you for your kind advice in advance.

[austinpop]

13 minutes ago, simonklp said:

 

Sorry that I probably have missed some of the related messages. But I have doubt that whether this kind of improvement by using a better clock (e.g. sCLK-EX) only applies to cases where the Asynchronous USB Transfer mode is not properly implemented. Because under this transfer mode, the clock of the DAC is the master that the clocks of the upstream devices should not really affect so much.

 

For example, in case of playback using HQPlayer with NAA, the NAA isolates as much as possible at the software side by using a large asynchronous FIFO buffer. Is there really a substantial improvement by replacing the original clocks by better ones at the upstream side of DAC? Was there any experiment on this case? Thank you for your kind advice in advance.

 

I suggest you go to the index in the first post to understand how we got here.