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

[hurka]

many outputs just disturbing each other.Im using only i2s.

[ElviaCaprice]

9 minutes ago, hurka said:

many outputs just disturbing each other.Im using only i2s.

If you look back in the thread you will see Hols found the SU-1 i2s to Holo DAC similar as tXusbUltra to usb Holo.  But he did not modify the SU-1 power supply with an LPS-1, so my guess would be that the SU-1 i2s with LPS-1 still trumps a single tXusbUltra via USB.   But, what happens if you add 3 more sclk EX clocks preceding the tXusbUltra in other components?  All bets off and I think the USB chain would surpass the SU-1 i2s in sound quality.

[hurka]

pls don't use su1 with original power ,simple shit.only with lps1.

my plain : add 2 lps1 fed crystek xos in my lks amanero.

And compare with,without tx usb ultra.

if much better with tx usb,then I need think about this, or  sClk options.

 

[romaz]

5 hours ago, limniscate said:

 

At first we replaced the Sonore microRendu with the sMS-200 Ultra and left the Singxer SU-1 in place.  Then we took the Singxer out and just used the sMS-200 Ultra direct to the Yggy via Curious USB.  The sMS-200 Ultra via Curious USB to Yggy was much better than the sMS-200 Ultra via Curious USB to the Singxer via Transparent Reference AES to the Yggy.

 

I may have to try those cables and filter.  I'm really wondering what a system with the SotM sClk-EX would sound like in the whole digital chain including the server and the DAC.  Aren't you testing the clock on the server side now?  Could you please tease us with some preliminary impressions?

 

 

To the ears of most (including mine), the Yggy sounds best via AES and to the extent that you made it sound best via USB says a lot.  Similarly, my DAVE sounds best via USB and this is no secret as even Rob Watts says the same thing.  Even with an Aurender W20 which has an excellent AES output (probably its best sounding output) and a $7k High Fidelity AES cable that I had on loan, the Aurender still sounded better via USB with my DAVE.  Having tested various USB-to-SPDIF adapters in the past with my DAVE including a Berkeley Alpha, modified Audiophilleo with LPS-1, SIngxer F-1, W4S uLink, iFi micro iLink (and a few others), I could never get SPDIF to sound better on my DAVE than USB.  That changed with the dX-USB HD Ultra and today, I have SPDIF sounding better on my DAVE than USB which is really an accomplishment.  It truly is amazing what a clock can almost singlehandedly do.

 

As long as the improvements keep coming, it's hard to stop but at some point, practicality has to set in and my gut tells me that if there are no bad clocks at the beginning of the chain, you won't need so many reclocking devices after the server.  Of course, this means that if you do a lot of internet streaming, you will also need to replace the clocks in your modem/router and so I am seriously considering it.  

 

I do not yet have my reclocked motherboard.  The ASRock board that I bought is in SOtM's possession although I have asked them to halt modifying it.  As you might know, this is a uATX board that cannot be powered solely by my SR7, at least not without using a DC-to-ATX converter and so I purchased one made by HDPlex which has the best specs I could find but still gave me pause for concern because at least on paper, the noise specs are underwhelming and equally important, I have no confidence this is a low impedance device based on my experience with my other HDPlex gear.

 

With the support from a group of friends, we have purchased a few more motherboards to test including a SuperMicro board that I mentioned in a previous post.  This SuperMicro motherboard is a SoC (System on a Chip) board that has no external PCH and so all of its ports including SATA, USB, LAN, PCIe, etc. have a much more direct and low latency path to the CPU.  Moreover, this motherboard has the option of being powered either via an ATX PSU or a 12V DC supply and this ability proved to be invaluable.  While there are probably better ATX PSUs around, we powered this motherboard initially with an EVGA SuperNova G2 850w ATX PSU that we had on hand and without question, it sounded very good and very dynamic.  For $140, this is an excellent ATX PSU and looking at reviews, its ripple values at low load (10-20%) are excellent with all 4 rails (4.6-7.6mV) and handily outperforms the HDPlex DC-ATX converter I bought (at least on paper).  As usual, there are no reported measurements on output impedance as this parameter seems to hold no value for reviewers or most manufacturers.  Anyway, we then compared this to the SR7 and even with all the noisy switching regulators within this motherboard that get in the way, the improvement we heard in terms of improved dynamics and this tonally more vivid and more holographic sound was just stunning.  It was similar to what I heard when I went from my HDPlex to the SR7 with my Mac Mini.  Having now heard this, there was no way I could continue on with my ASRock uATX board and the HDPlex DC-ATX converter.

 

Unfortunately, the problem with this SuperMicro motherboard is that it uses an embedded CPU that cannot be replaced and the lowest power CPU they incorporate is a dual core Xeon with a TDP of 25w.  The board comes with a passive heatsink but somehow, this thing is a cooker and it gets quite hot, hotter than what I would expect for a TDP of 25w.  Since I insist on my servers to be fanless and since I like to keep my servers on 24/7 so they're ready to go, this is a real dilemma.  Unfortunately, since this isn't a board designed for gamers or for overclocking, the BIOS has no provision for overclocking or underclocking.  All that I can do is turn off turbo which isn't enough.  There is, of course, the option of dethrottling your CPU via software but somehow, I couldn't get temps down low enough and it seems I have to do it every time I reboot the machine and so this whole process just didn't sit well with me and so this motherboard went back.

 

I was convinced, however, that I wanted an SoC motherboard and a motherboard that I could directly power with my SR7 without having to use an ATX converter and so this has led to quite a search.  Most low power motherboards that met this criteria unfortunately use Celerons like the J1900, a 4-core CPU with a TDP of 10w and a measly 2MB of L3 cache.  What's worse, this cache isn't Intel's SmartCache meaning each core only has access to 1MB of cache max and that 1MB is then shared with another core.  Roon has been very clear that with version 1.3, you really should have at least an i3 and preferably an i5 because they have larger caches which are important for smooth operation.  Having bought the original sonicTransporter when it first came out, this unit came with a Celeron and while Roon runs on it, I can attest that the user experience has been subpar, especially when playing DSD files as you get lots of pauses and skips.  This board, however, isn't an SoC board which led me to wonder if that might make a difference.

 

As I reached out to a friend who owns a Fidelizer Nimitra server, I was interested to know his experience since this server utilizes an SoC motherboard with a J1900 Celeron CPU and he told me he couldn't be happier with this server and that his user experience was very good.  With the latest version of Roon, he said everything runs smoothly.

 

A few days ago, @greenleo inquired about the Antipodes DX server.  I didn't respond because I had too much going on but I have heard the Gen 2 version of the DX and it is just wonderful.  If you have no interest in oversampling, are looking for a turnkey solution and you have the funds to afford this server, give it serious consideration.  If you have followed this now gigantic thread from the beginning, then you know that I referenced the Antipodes DX as one of my inspirations for pursuing this "direct connection" during my introductory post.  Mark Jenkins, Antipodes creator, had somehow discovered a way to combine Roon Core and Roon Ready in one chassis and without using the network and based on how good this server sounded, I began to explore what he meant by this.  I believe his solution is based on software and is incorporated into his Linux-based OS but nonetheless, it led to some interesting hardware findings.

 

It turns out the Antipodes DX uses a certain Jetway SoC mini ITX motherboard that incorporates a 4-core Celeron with a TDP of only 7.5w.  Mark also went with an SOtM tX-USBhubIn as his USB output card.  It is unclear if he is using SOtM's newest superclock.  The point is that with this Celeron-based SoC motherboard, this machine runs Roon 1.3 just fine (with the exception of upsampling to DSD) and it sounds wonderful.

 

Here is another interesting comment that Mark made.  It turns out he intentionally sought out the Celeron in favor of an i3/i5/i7 because his listening tests showed that the simpler structure of the Celeron "sounded better."  Here is what Mark said when Roon version 1.3 was first released:

 

"Our main concern has been Roon’s stipulation that an i3 CPU, at least, should be used. Unfortunately the complex CPUs, such as Intel’s i3/i5/i7 sound poor when rendering the file to a digital signal so our Integrateds use a high power quad-core version of the Intel Celeron. The issue for Rendering is not so much about power. Lower power CPUs do not sound better and some of them sound worse. The issue is the difference in architecture between the i3/i5/i7 CPUs and the simpler CPUs."

 

I have no way to validate his claim as I have not done this comparison testing but given my great respect for what Mark has accomplished with his DX, I decided to pursue what I hope will finally be THE motherboard.  While SuperMicro makes a mini ITX SoC motherboard that embeds a Celeron J1900 and can be powered directly by my SR7, DFI makes one that is even better suited to my needs.  DFI is based in Taiwan but has offices here in Sacramento and New Jersey.  They have just released a new SoC motherboard called the BW171 and the specific model I chose utilizes Intel's newer Celeron N3710 based on a more efficient 14nm die cast.  Consequently, this 4-core CPU has a TDP of only 6w and runs at a base speed of 1600MHz.  Moreover, it utilizes DDR3L RAM that also runs at 1600MHz.  I have not unequivocally validated this for myself but a respected colleague did some listening tests a few years ago and found that his optimum SQ occurred when his CPU and RAM frequency were identical and indeed, with a previous server build, I found my best SQ when I dethrottled a certain i7 down to 800MHz and dethrottled my RAM down to  800MHz.  Whether this was coincidence or not, I'm not sure but since this DFI SoC motherboard's BIOS does not allow for underclocking of CPU or RAM, it is strangely coincident and possibly fortuitous that both CPU and RAM will be running at the same speed.

 

This DFI mini ITX board also incorporates dual Intel i201AT NICs and a single PCIe 2.0x1 slot that will accept an Adnaco S3B fiber USB PCIe card.   It would be ideal if the NICs were fiber NICs and since DFI is capable of "custom making" me a board, they said they would explore it.  The problem is custom = super duper expensive and so I realistically won't be exploring it.  Having explored this board already, it relies on 2 crystal oscillators (19.2MHz and 33MHz) to provide timing for the integrated SoC hub and all of its subclocks.  The 2 LAN ports each get their timing from a 25MHz clock.  This board is so new that only one exists in the U.S. at the present time although a shipment of 10 boards is due stateside in about 2 weeks.  DFI has agreed to sell me the sole unit that they have.  The cost of the board is $331 which, of course, includes the CPU.

 

This motherboard arrives tomorrow.  My stock Adnaco S3B USB kit has already arrived and will be tested with this new motherboard when it arrives.  A good friend has agreed to loan me his tX-USBUltra which arrives tomorrow.  I will be comparing this "straight USB" solution against my current ethernet-based "triple stack."  Because DAVE's USB implementation is better than its SPDIF implementation, I will try replacing my dX-USB HD Ultra with the tX-USBUltra (meaning switch > sMS-200Ultra > tX-USBUltra > DAVE) to see if this sounds better and I'm confident that it will.  Of course, I will see what a quad stack of Ultras sound like (switch > sMS-200Ultra > tX-USBUltra > dX-USB HD Ultra > DAVE).  

 

Regardless of the outcome, I simply do not envision owning a quad stack.  It's possible I may replace my dX-USB HD Ultra with the tX-USBUltra and SOtM is willing to swap the clocks out for me.  If I had my way, however, I would love for the combination of this DFI motherboard with its 4 clocks replaced along with the Adnaco with its 3 clocks replaced to sound the best as dealing with this simple setup would be ideal and properly powering this type of setup would be easier.  Furthermore, it would be much more pleasant to deal with thin and light fiberoptic cabling than 6m of very thick, heavy and rigid CAT7 Ethernet cabling, especially when my server and endpoints only sit about 6 inches apart.  

[romaz]

7 hours ago, ElviaCaprice said:

 

Roy, it will be nice to hear how you find the tX-USBUltra in comparison.  Hopefully you can do a direct USB only, drop the Ethernet, comparison.  I would be curious about just 1 EX against 1 EX , Ethernet vs USB direct.  Which I guess you can't do now that I think about it, with 2EX clocks in the sms-200.

 

 

My friend who is loaning me his tX-USBUltra also owns a standard sMS-200.  Using his BRIX (NUC) server with its standard USB port, he compared BRIX > tX-USBUltra > Chord DAVE against BRIX > sMS-200 > tX-USBUltra > Chord DAVE and in comparison, straight USB sounded flat and dull.  Just like me, he uses the "direct connection" with his sMS-200.  I'm pretty sure I will find the same thing with my setup that he has found and so my greater interest will be to see what happens with the Adnaco in the chain.  Adnaco with stock clocks + tX-USBUltra might give me some sense of how good the Adnaco might sound with its clocks replaced, especially since I believe they are both using the same TI USB chipsets.

[hurka]

10 minutes ago, romaz said:

 

My friend who is loaning me his tX-USBUltra also owns a standard sMS-200.  Using his BRIX (NUC) server with its standard USB port, he compared BRIX > tX-USBUltra > Chord DAVE against BRIX > sMS-200 > tX-USBUltra > Chord DAVE and in comparison, straight USB sounded flat and dull.  Just like me, he uses the "direct connection" with his sMS-200.  I'm pretty sure I will find the same thing with my setup that he has found and so my greater interest will be to see what happens with the Adnaco in the chain.  Adnaco with stock clocks + tX-USBUltra might give me some sense of how good the Adnaco might sound with its clocks replaced, especially since I believe they are both using the same TI USB chipsets.

So the Brix NUC usb out flat and dull.(only one dc input for all components!!!)

But my separately powered(u caps) mobo sound not.

[ElviaCaprice]

4 hours ago, romaz said:

 

My friend who is loaning me his tX-USBUltra also owns a standard sMS-200.  Using his BRIX (NUC) server with its standard USB port, he compared BRIX > tX-USBUltra > Chord DAVE against BRIX > sMS-200 > tX-USBUltra > Chord DAVE and in comparison, straight USB sounded flat and dull.  Just like me, he uses the "direct connection" with his sMS-200.  I'm pretty sure I will find the same thing with my setup that he has found and so my greater interest will be to see what happens with the Adnaco in the chain.  Adnaco with stock clocks + tX-USBUltra might give me some sense of how good the Adnaco might sound with its clocks replaced, especially since I believe they are both using the same TI USB chipsets.

 

Roy, I take it your friend was using quality power supplies on his testing?  I don't understand what flat and dull mean in audio terms?  Sometimes some folks attribute electrical noise as being alive or full of energy. 

I don't have a lot of confidence in the Adnaco S3B in single mode, after reading an old review here, where they compared the Adnaco (powered with an LPS) to an SoTM Usb audio card and couldn't declare a winner. 

Page down about 3/4 to Obee.

 

As always I shall be interested in your findings.

[hurka]

5 minutes ago, ElviaCaprice said:

 

Roy, I take it your friend was using quality power supplies on his testing?  I don't understand what flat and dull mean in audio terms?  Sometimes some folks attribute electrical noise as being alive or full of energy. 

I don't have a lot of confidence in the Adnaco S3B in single mode, after reading an old review here, where they compared the Adnaco (powered with an LPS) to an SoTM Usb audio card and couldn't declare a winner. 

Page down about 3/4 to Obee.

 

As always I shall be interested in your findings.

My older z97 itx mobo usb sound dull and flat.external usb card help ,but new series 270 much much better.

sound galleries don't use any usb cards I think.Linn ,lumin makes only better separately power fed cheap clock.

sotm use special tcxo,ocxo?send picture pls,

master clock input?(1000dollar stuff)rubidium clock required?this is really funny

4 clock output?   the cables need short one ! cant put together with one clock.....

 

[tboooe]

1 minute ago, hurka said:

My older z97 itx mobo usb sound dull and flat.external usb card help ,but new series 270 much much better.

sound galleries don't use any usb cards I think.Linn ,lumin makes only better separately power fed cheap clock.

May I ask what z270 mobo you are using?

[hurka]

12 minutes ago, tboooe said:

May I ask what z270 mobo you are using?

gigabyte h270n usb out fantastic.

[hurka]

z series don't know,h series need 2 amper fed ultracapacitors

[hurka]

12volt 20000mah battery sound almost the same,really enjoyable,can run 6 hours without charge,cost around 40 dollar

lampi have battery mobo upgrade,can you decide witch one do you like.

[hurka]

if you don't have money but you have sms 200,buy some  battery 12v ,find some caps more as possible...and go 

[hurka]

uptone audio lps1 best power supply I think.but I need 12pcs .so check taobao about this if you can soldering....

[hurka]

ok romaz, nobody have sms 200 ultra but you have .

hard words coming.......

pay for you sotm  or not.

[spacexpert]

I like this hurka fellow.  Good energy, a lot of enthusiasm. 

[hols]

On 2017/5/11 at 3:48 PM, austinpop said:

Since there is so much interest, i do want to point out that I am still experiencing the random pause problem with my Ultra stack that I mentioned in an earlier post: 

https://www.computeraudiophile.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-the-sms-200-and-microrendu/?do=findComment&comment=659047

 

 

SOtM have asked me try a few different experiments to isolate the problem, and I have sent them the data. They seem very committed to solve the problem for me, so I'm quite hopeful. But I do want to be clear the issue still exists for me.

It's great that SOtM is committed to solve the problem. After using it for 2weeks now my tx-USB Ultra still experience this pause problem but it is a little different from your situation. I only have the pause problem when I use the LPS-1. When I use the stock switching power supply then there is no poblem. I have also tried it on different DACs including Holo Spring and T+A DAC8DSD, in different OS like Windows10 and Linux the result is the same. Just for your reference. 

[ElviaCaprice]

1 minute ago, hols said:

It's great that SOtM is committed to solve the problem. After using it for 2weeks now my tx-USB Ultra still experience this pause problem but it is a little different from your situation. I only have the pause problem when I use the LPS-1. When I use the stock switching power supply then there is no poblem. I have also tried it on different DACs including Holo Spring and T+A DAC8DSD, in different OS like Windows10 and Linux the result is the same. Just for your reference. 

Is the stock switching supply at a higher voltage?  Like 9V?  Since the LPS-1 doesn't shut down and start blinking, it must be supplying enough amperage, but that doesn't mean the SOtM doesn't need more voltage.  I suspect from what your saying it needs more than 7V.  If you had two LPS-1's you could try doing a series and see if that solves the problem.  Like 5V and 3.3V

[austinpop]

7 hours ago, romaz said:

 

To the ears of most (including mine), the Yggy sounds best via AES and to the extent that you made it sound best via USB says a lot.  Similarly, my DAVE sounds best via USB and this is no secret as even Rob Watts says the same thing.  Even with an Aurender W20 which has an excellent AES output (probably its best sounding output) and a $7k High Fidelity AES cable that I had on loan, the Aurender still sounded better via USB with my DAVE.  Having tested various USB-to-SPDIF adapters in the past with my DAVE including a Berkeley Alpha, modified Audiophilleo with LPS-1, SIngxer F-1, W4S uLink, iFi micro iLink (and a few others), I could never get SPDIF to sound better on my DAVE than USB.  That changed with the dX-USB HD Ultra and today, I have SPDIF sounding better on my DAVE than USB which is really an accomplishment.  It truly is amazing what a clock can almost singlehandedly do.

 

Yes, I agree this is a game changer. A lot of "conventional wisdom" regarding input types could be turned on its ear. For one, I think it brings synchronous clocking modes like S/PDIF and AES back into contention. I am still amazed you got S/PDIF sounding better than USB, given that in S/PDIF the clock is embedded in the data stream and has to be extracted!

 

Similarly, the simplistic view of asynchronous USB was that since the DAC was the clock master, the clocks on the source devices "shouldn't matter." Yet, a whole cottage industry around regenerators, decrapifiers and their ilk have shown that the source clock matters too.

 

The tX- and dX- boxes really allow a new level of flexibility in optimizing DAC performance.

 

7 hours ago, romaz said:

As I reached out to a friend who owns a Fidelizer Nimitra server, I was interested to know his experience since this server utilizes an SoC motherboard with a J1900 Celeron CPU and he told me he couldn't be happier with this server and that his user experience was very good.  With the latest version of Roon, he said everything runs smoothly.

 

Just a shout out to @mozes who I know is also running USB direct from Nimitra > tX > Brooklyn DAC, using a combo of JS-2 and VR Mini supplies.

 

7 hours ago, romaz said:

As long as the improvements keep coming, it's hard to stop but at some point, practicality has to set in and my gut tells me that if there are no bad clocks at the beginning of the chain, you won't need so many reclocking devices after the server.  Of course, this means that if you do a lot of internet streaming, you will also need to replace the clocks in your modem/router and so I am seriously considering it. 

 

Over on @Cornan's Experimental Tweaks thread, we've discussed this in the context of a simple goal:

• Make Tidal Great Again! Just kidding. Make the same track played back through Tidal sound as good as the same track played back locally.

I can tell you that as of now this is not true for me at all. Local files sound significantly better than streamed from Tidal. It also is where I plan to focus my experimental energy next. 

[romaz]

2 hours ago, ElviaCaprice said:

 

Roy, I take it your friend was using quality power supplies on his testing?  I don't understand what flat and dull mean in audio terms? 

 

1

 

I'm quoting my friend when I wrote "flat and dull" but keep in mind that he said this "relative to" BRIX > sMS-200 > tX-USBUltra > DAC.  In the same way that Rajiv described the soundstage collapsing when he replaced his sMS-200 Ultra with the mR, I'm sure he meant it as a relative description since those of us who own an mR know that its sound stage does not sound small.  You should also know that the dCBL-CAT7s that I received for testing came from this same friend and so with these cables in his BRIX > sMS-200 > tX-USBUltra  chain, SQ is definitely stacked in favor of the Ethernet configuration.  As far as PSU, he is using SOtM's recently discontinued mBPS-d2s.  Regardless, I will make my own determination.

 

2 hours ago, ElviaCaprice said:

I don't have a lot of confidence in the Adnaco S3B in single mode, after reading an old review here, where they compared the Adnaco (powered with an LPS) to an SoTM Usb audio card and couldn't declare a winner. 

Page down about 3/4 to Obee.

 

As always I shall be interested in your findings.

2

 

Yes, I have seen that thread.  There's not much out there on the Adnaco and so this represents only one data point.  However, what you see as a fail, I see as a victory.  The Adnaco S3B was not designed for high-end audio.  It was designed to allow USB transmission over long distances and so it has not been optimized for high-end audio...yet.  Adnaco also makes the JCAT Femto USB card for JPLAY and so they understand what audiophiles want and yet none of the audiophile components that went into the JPLAY card (like the Crystek clock and the linear regulators) went into the Adnaco.  The only thing the Adnaco has going for it is fiber.  What impressed me is that against the SOtM audio card and its much better clock and linear regulators, the Adnaco equalled it  With 3 clock replacements and with the replacement of at least 4 switching regulators (this includes the switching regulators on the PCIe card which I now plan to also replace), this card can only improve further.  Keep in mind that before I sent my cheap $20 switch to SOtM for clock replacement, I also tried it in my chain and powered by my LPS-1, it added nothing positive but with the clock replacement and replacement of its noisy switching regulators, it came back a hero.  As I combine the Adnaco with the tX-USBUltra later today, I am hoping this will give me some idea of just how much improvement I can expect but truth be told, I am actually expecting even greater improvement with an optimized Adnaco since we are talking about 3 clock replacements for the Adnaco vs only 1 clock replacement for the tX-USBUltra.  If I can get Paul Hynes to provide me the regulators that I need, this is where I am hoping this card can really distance itself.

[romaz]

22 minutes ago, hurka said:

ok romaz, nobody have sms 200 ultra but you have .

hard words coming.......

pay for you sotm  or not.

 

 

I have already detailed the amount of money I spent for my sMS-200 modification in a previous post and it is the exact same amount of money that @austinpop spent for his modifications.  Anyone could have approached SOtM like I did and they would have offered the exact same service.  I have received nothing for free and I have been completely transparent about my process, including my successes and failures.

[austinpop]

6 minutes ago, ElviaCaprice said:

Is the stock switching supply at a higher voltage?  Like 9V?  Since the LPS-1 doesn't shut down and start blinking, it must be supplying enough amperage, but that doesn't mean the SOtM doesn't need more voltage.  I suspect from what your saying it needs more than 7V.  If you had two LPS-1's you could try doing a series and see if that solves the problem.  Like 5V and 3.3V

 

I've tried a bunch of PSU experiments with the tX, ranging from the LPS-1 at 7V/1.1A to stock SMPS at 9V/2A, HDPlex 100 at 9V/5A, to el Cheapo at 7.5V/3.3A.

 

It's not the PSU in my case.

 

10 minutes ago, hols said:

It's great that SOtM is committed to solve the problem. After using it for 2weeks now my tx-USB Ultra still experience this pause problem but it is a little different from your situation. I only have the pause problem when I use the LPS-1. When I use the stock switching power supply then there is no poblem. I have also tried it on different DACs including Holo Spring and T+A DAC8DSD, in different OS like Windows10 and Linux the result is the same. Just for your reference. 

 

Thanks for reminding me to post an update on this.

 

The last experiment May asked me to try was to replace my Ayre Codex DAC with another. Serendipitously, this was the night before my planned listening session at @limniscate's. In over 5 hours of use, briefly with the DAVE, but mostly with the Yggy, I did not have a single pause.

 

So the finger is pointing at an interaction between the tX-USBultra and the Codex.

 

SOtM are in the process of acquiring a loaner Codex from their distributor in Korea to see if they can troubleshoot the problem in-house. I should know more in a few days how that is going. 

 

Yes, I am very satisfied with their level of commitment in resolving this issue.

[austinpop]

2 minutes ago, romaz said:

 

I have already detailed the amount of money I spent for my sMS-200 modification in a previous post and it is the exact same amount of money that @austinpop spent for his modifications.  Anyone could have approached SOtM like I did and they would have offered the exact same service.  I have received nothing for free and I have been completely transparent about my process, including my successes and failures.

 

And before anyone asks - I have received nothing for free, nor did SOtM even know that I would post here with my impressions.

 

Heck - I even had to pay shipping to Korea and back. And that's not cheap. Thanks for that - FedEx!

[Johnseye]

28 minutes ago, austinpop said:

 

Yes, I agree this is a game changer. A lot of "conventional wisdom" regarding input types could be turned on its ear. For one, I think it brings synchronous clocking modes like S/PDIF and AES back into contention. I am still amazed you got S/PDIF sounding better than USB, given that in S/PDIF the clock is embedded in the data stream and has to be extracted!

 

Depending on what your bitrate and frequency/bandwidth requirements are.  There are limitations with those interfaces.  I liked the idea of I2S because it is less restrictive and supposedly less prone to jitter, but there are some clocking considerations to be made as that's handled differently.  I'm seeing some DACs with 2 channel HDMI and that's interesting.

[austinpop]

Of course.

 

I have a large investment, and interest, in DSD content. This has nothing to do with the "DSD vs. PCM" format wars. As a classical music lover, I have accumulated a lot of my music on SACD, which I've then ripped to DSF.

 

Very few DACs accept DSD over anything other than USB, so my primary focus has been on the USB interface.