Jump to content
IGNORED

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


Message added by The Computer Audiophile

Important and useful information about this thread

Posting guidelines

History and index of useful posts

Most important: please realize this thread is about bleeding edge experimentation and discovery. No one has The Answer™. If you are not into tweaking, just know that you can have a musically satisfying system without doing any of the nutty things we do here.

Recommended Posts

My positive experiences with the Ultra stack using a cascade of sCLK reclockings has gotten me to reflect a bit.

 

The three primary factors that I have found to improve the SQ are:

  1. Signal integrity - the Ultra stack being the most recent example, but previously experienced via the W4S RUR (Recovery)
  2. Leakage loop blockage:
    • from the AC mains via power supplies like the LPS-1
    • on the signal path, via isolators like the Baaske on Ethernet, and the Intona on USB
  3. low output impedance on PSUs - again, my direct experience is with the LPS-1, including using Canare star quad DIY DC power cables.

Now - regarding SI (signal integrity) vs. GI (galvanic isolation from leakage loops). The experts at Uptone and SOtM tell us that implementing isolation on the signal path in USB comes at the cost of increased jitter. The ISO-Regen has had a lot of thought and expertise applied to balance these effects. We are all eager to see how this pans out.

 

We don't know whether the same also holds for Ethernet isolators like the EMO EN-70HD, the Etalon, or the iso-CAT6, but it's reasonable to expect so.

 

Having previously had a chain that required an Intona for max SQ, and given the fact that the new Ultra chain degraded when I added in the Intona, it's making me wonder if there is an empirical model emerging from these experiments. Let me be upfront and say - this is just a theory. But here it is in a nutshell:

  1. Optimize the data path for maximum signal integrity, by improving the clock quality as far upstream of the DAC as practical
  2. Avoid isolators on the signal path to minimize jitter
  3. Address the blockage of leakage loops on the power plane, using isolating power supplies - both on the DC side, like the LPS-1, VR Mini, or the Hynes SRn - and on the AC side with Topaz and similar supplies.
  4. Minimize the DC PSU output impedance - again by selecting quality PSUs like the LPS-1, VR Mini, or the Hynes SRn

Having said that, @romaz's positive reports with the dCBL-CAT7 and iso-CAT6 seem to contradict #2 above. However, as I remember it, he placed this chain prior to his sCLK-modded switch, so perhaps two factors could explain it:

  • all the clocks upstream of the isolator are not optimized (yet)
  • the 4 sCLK chain post-isolator sufficiently restores SI, and the iso-CAT6 adds an additional benefit from isolation.


We don't know for sure. 

 

But one of the experiments that will be very interesting will be when Roy adds sCLK mods upstream of the iso-CAT6 in his server. At this point the key comparison will be between the full chain with and without the iso-CAT6 isolator. In other words - if there are superior clocks upstream of the iso-CAT6 isolator, does its presence still improve SQ - or does it degrade SQ?

 

We shall see.

Link to comment

newbie here:confused:

Does the BNC WCLK word clock output on the Singxer SU-1 allow superclock insertion? If you installed a superclock via the SU-1 would this be too close to the DAC for maximum effectiveness? Would this be better and cheaper than going all in on the SOTM solution which is a moving target? I have the SOTM 200/ LPS-1 and would like the keep things simple but on the cusp of buying the SU-1 for its connection options and clocks.

Link to comment

Hereis a series of truly great posts by Paul Hynes, for those who are interested. Extremely interesting and well worth the time to read! :)

 

 

Ariston RD40 turntable with Denon DL103, Mytek Brooklyn DAC/HPA with Rogers MCP100 SUT, Fostex TH900 HP with balanced Forza Audioworks Noir Hybrid HPC Mk2 & Genelec 8030 CRW              

Knowledge speaks but wisdom listen. Jimmy Hendrix

Link to comment
1 hour ago, mozes said:

I also would like to know how Vinnie Rossi Mini compares to Paul Hynes SR5 or SR7

According to Vinni, one Mini will not be able to power the sMS-200 Ultra and the dX-USB Ultra at the same time.  However, one Mini should be able to power one of the devices.

Link to comment
46 minutes ago, limniscate said:

According to Vinni, one Mini will not be able to power the sMS-200 Ultra and the dX-USB Ultra at the same time.  However, one Mini should be able to power one of the devices.

Thanks! I was asking about the sonic performance between the two 

Link to comment
10 hours ago, austinpop said:

Let me be upfront and say - this is just a theory. But here it is in a nutshell:

  1. Optimize the data path for maximum signal integrity, by improving the clock quality as far upstream of the DAC as practical

...

But one of the experiments that will be very interesting will be when Roy adds sCLK mods upstream of the iso-CAT6 in his server. At this point the key comparison will be between the full chain with and without the iso-CAT6 isolator. In other words - if there are superior clocks upstream of the iso-CAT6 isolator, does its presence still improve SQ - or does it degrade SQ?

 

We shall see.

 

Maybe you mean as many good clocks all along the way upstream from the DAC? FWIW, Audiostream.com just ran a review on SotM's USB decrapifier by Steven Plaskin suggesting that the best clock closest to the DAC with the best power supply possible yields the best sonic results. (This has been my experience too, although only based on my very limited testing in my systems.)

 

I'm not suggesting you're wrong @austinpop, I'm pointing out Plaskin's point of view as we all explore these relationships further. We're all waiting for the ISO Regen vs SotM USBultra cage match, if only due to the price difference.

 

Sum>Frankenstein: Audirvana, Uptone EtherRegen+LPS-1.2, Rivo Streamer+Uptone JS-2, Schiit Yggdrasil LiM+Shunyata Venom, Linn LP12/Hercules II/Ittok/Denon DL-103R, Audible Illusions  M3A, Pass XA25, Tellurium Black II, Monitor Audio Silver 500 on IsoAcoustics Gaias, Shunyata & Transparent Audio cables/cords, P12 power regenerator, and positive room attributes.

Link to comment
On 5/12/2017 at 11:41 AM, Quadman said:

I can say this; the improvement I have gotten from digital sources in the past 2 years and the $ that I spent to do so are way less than the $ one needs to spend to get equal improvement in the analog chain.

 

Thanks for your post.  I completely agree, there seems to be much more gain for the dollar with digital than analog these days.

Link to comment
10 minutes ago, feelingears said:

Audiostream.com just ran a review on SotM's USB decrapifier by Steven Plaskin suggesting that the best clock closest to the DAC with the best power supply possible yields the best sonic results. (This has been my experience too, although only based on my very limited testing in my systems.)

I think this is now sort of established and there is no argument about it.

 

My understanding of what @austinpop wrote is that there is some trade-off going on between isolation and signal integrity in a USB chain. What is the impact of a better clock when it improves SI relative to the added Jitter as a result of isolation devices like Intona and now the new Iso-Regen which is supposedly much better than the Intona?

Link to comment
On 5/12/2017 at 11:50 AM, limniscate said:

I want to hear @romaz's report on the server side as well.  If that's as big as a jump with the other clocks, could I just get the latest SotM server with the sClk-Ex and forgo the dX-USB Ultra/tX-USB Ultra and sMS-200 Ultra, or would I have to get all of the devices?

 

I'm also probably going to try the cables and LAN filter as you suggested @romaz

 

What you're saying is intriguing and I had also considered it.  What I will say is this...listen to their latest server before you commit.  Several who have bought it have PM'd me and I won't get into details publicly but It apparently has a very unique sound signature that is not to everyone's taste.   Definitely try before you buy.

 

As for their dCBL-CAT7, it is the very best that I have heard. There is nothing remotely close.  In my setup, I found the combo of dCBL-CAT7 > iSO-CAT6> dCBL-CAT7 to be effective before my server but it was MUCH more effective in the "direct path" between server and DAC.  My preference was to have it directly between server and reclocking switch and then from the reclocking switch to the sMS-200Ultra, I used SOtM's black CAT6 cable which is the cable that comes with the iSO-CAT6.

Link to comment
On 5/13/2017 at 7:12 AM, rickca said:

I'm confused by the SOtM nomenclature.  It says the tX-USBultra comes with a sCLK-EX12 board.  That board is listed under the heading of 'single clock frequency for interface system' on the sCLK-EX webpage.  So does the sCLK-EX12 board have 4 clocks or just one?

 

One other question.  Using the sCLK-EX and transplanting some of the clocks from say, a tX-USBultra to a switch or SMS-200, you have wires interconnecting various devices sharing the sCLK-EX.  Don't these wires somewhat degrade the clock compared to each device having its own clock board?

 

I'd appreciate a response from @romaz, @austinpop, @mozes or anyone else who has experience with this approach.

 

The sCLK-EX board is a single board that has 4 independent clock outputs.  Because they are independent, they can be specified to be whatever frequency you want.  I don't know this for a fact but my guess is that they each get their timing from a single master internal clock.  This internal clock has the option of being bypassed by an external master clock.  SOtM offers an option to install connectors (for $200) to allow you the option of connecting an external master clock in the future which is something I have no interest in doing.  

 

When SOtM specifies "sCLK-EX12", what this means is that the clock has been programmed for a timing frequency of 12MHz.  You could easily request a clock board with an sCLK-EX25 or sCLK-EX48, etc.

 

Yes, the wires in theory will degrade clock performance and the longer the wires, the greater the degradation.  SOtM prefers no max out clock wire length at 45cm.  Whether this is based on actual listening or some arbitrary decision, I'm not sure.  Regardless, at 45cm, the improvements I am hearing are remarkable.

Link to comment
On 5/13/2017 at 6:16 AM, AmusedToD said:

Since this is a blind buy, is there a consensus regarding the SQ improvement with the tX-USBultra over the basic sms-200? It's really an expensive device, I just hope it's 3 times better than the (almost) 3 times cheaper ISO regen.

 

And finally, since Uptone is (boldly) claiming  their new $35 USPSB adapter is better than any USB cable, if I decide to buy one should I put it between the SMS-200 and the tX-USBultra, or between the tX-USBultra and my DAC?

 

If you are expecting the tX-USBultra to be 3x better than the ISO Regen or your sMS-200, be prepared to be disappointed.  It is likely that nothing at this high level of audio will ever be 300% better than its competition.  High end audio is a fool's game and as Ted Brady puts it, we are all bozos on this bus.

Link to comment
30 minutes ago, mozes said:

I think this is now sort of established and there is no argument about it.

 

My understanding of what @austinpop wrote is that there is some trade-off going on between isolation and signal integrity in a USB chain. What is the impact of a better clock when it improves SI relative to the added Jitter as a result of isolation devices like Intona and now the new Iso-Regen which is supposedly much better than the Intona?

That was not our (@austinpop and I) experience.  When we just had the switch in place, it made just as big of a difference as when we had just the tX-USB Ultra in place.

Link to comment

@romaz

Thanks for the reply.

Do you know where I could try a SOtM server?  Also, I saw an old post where you said the iSO-CAT6 had very little effect on sound quality; however that post was when you had the dCBL-CAT6.  Are you saying that the dCBL-CAT7 has some sort of greater synergy with the iSO-CAT6?  

 

Have you had the dCBL-CAT& and iSO-CAT6 combo between the server and switch and the switch and end point simultaneously, i.e., four cables and two iSO-CAT6?  That's what SOtM is recommending to me, but I'm wondering if it's just a money grab, lol.

Link to comment
On 5/13/2017 at 8:51 AM, amir57bs said:

how can you power an ATX socket with SR7? it seems you should buy 3 or 4 SR7! is it right?

 

You can't, not without a DC-ATX converter.

 

The ATX spec calls for a +12V, -12V (for legacy ports like RS-232), +5V and +3.3V rails along with a 5VSB rail (to provide power to the circuitry while the computer is in standby mode).  To build an SR7 with this many rails would be prohibitively expensive. The best that I can hope for is for Paul Hynes to design and build his own DC-ATX converter.  This is why I am purposely avoiding any motherboard that can't be powered by a 12V DC supply.

Link to comment
On 5/13/2017 at 8:16 AM, AmusedToD said:

Hi everyone,

 

I have the sms-200 which is powered by the Sbooster BOTW P&P Eco 12v, and the combo works great. In my chain after the sms-200 I have an Ifi nano iUsb3.0 which further lowers the noise floor and improves the signal coming out of the sms-200. After reading the glowing reviews of the tX-USBultra, I decided to bite the bullet and order one. 

 

Since this is a blind buy, is there a consensus regarding the SQ improvement with the tX-USBultra over the basic sms-200? It's really an expensive device, I just hope it's 3 times better than the (almost) 3 times cheaper ISO regen.

 

And finally, since Uptone is (boldly) claiming  their new $35 USPSB adapter is better than any USB cable, if I decide to buy one should I put it between the SMS-200 and the tX-USBultra, or between the tX-USBultra and my DAC?

I don't know how I would quantify improvement of the tX-USB Ultra over the stock sMS-200 because I haven't heard it in a while, but both @austinpop and I thought that the tX-USB Ultra and his modded sMS-200 (Ultra-like) yielded equal improvements to sound quality.

Link to comment
On 5/13/2017 at 1:22 PM, AmusedToD said:

 

Is there a reasonably priced SLC SSD? Which one did you try? Would there be any benefit in using an external SLC SSD as opposed to regular external (MLC) SSD drive for music storage?

 

Right now I am using a very noisy and slow external 7200 HDD. Sometimes I connect it either to my PC (so that Roon can read the files) or directly to my SMS200, but I would love to try the SLC SSD variant.

 

I bought my Intel X25-E from Ebay for only $80 from an IT supply store.  It was brand new and unopened.

Link to comment
On 5/13/2017 at 4:10 PM, limniscate said:

@romaz

Do you need two or three dCBL-CAT7 cables? It seems like you would need three: one from server NIC to iSO-CAT6, one from iSO-CAT6 to switch, and one from switch to sMS-200 Ultra. 

 

 

 

If you can afford it, put a dCBL-CAT everywhere.  I found plenty of wow factor from a pair of dCBL-CAT7s with an iSO-CAT6 in between when placed between server and reclocking switch.

Link to comment
On 5/14/2017 at 9:22 AM, ElviaCaprice said:

Yes, I know what we could learn.  Go smaller and use better power supplies. 

I'm getting ready to pull the trigger on my long time talked about Jetway Pico-ITX.  Plan on going for this one,

 http://jetwaycomputer.com/JBC400P93.html

 

The 2807 dual core, 4.3W, 12V DC, can be powered with 2-LPS-1's in a series.  Waiting on one minor detail before ordering.  If all works well, I'll send it off to SOTM for clocking upgrade.

 

 

empc753-3.jpg

 

Looks like a great way to go!  Make sure you test with LPS-1s.  On boot up, I suspect it will draw more than 1.1A.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×
×
  • Create New...