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

[barrows]

2 hours ago, Confused said:

Over time, this data must be received at exactly the same rate as the output, otherwise the FIFO buffer would empty or overflow.

No, this is not an issue with USB audio transmission.  There will never be empty or overflow conditions in the buffer because the rate of the data being sent from point to point is controlled via bi-directional communication, regardless of clock accuracy.

[barrows]

5 hours ago, mtcs said:

Hi all.

 

I've been involved with computer audio for a couple of years and finally have a "good sound." But, of course, I want  to do better. So, I'd like advice on my next upgrade.

 

I'm considering either adding  a tX-USBultra with Mutec10 ref clock or an AC power regenerator. I'm firmly convinced that better clocks help but I also know that AC power matters. At this point, I haven't done anything for AC beyond high quality power cords and distribution box.

 

My current system is:

 

Nordost Heindall 2 power cords, Nordost QB4 distribution box

SGC/Roon/HQPe/JS-2 => ultraRendu/LPS-1=>ISO Regen/JS-2 =>Auralic Vega dac=>

ARC amp, Wilson speakers

 

What do you think: better clock, better AC or other?

 

Thanks.

If you are not already doing it, here is a free upgrade for you: use ROON to oversample everything to the highest rate of DSD which your Vega accepts.  The Vega uses an ESS 9018 chip, and in my experience the ESS chips do way better with DSD input, and ROON oversampling to DSD is actually pretty damn good.

[barrows]

7 minutes ago, mtcs said:

I agree completely about upsampling. I'm a dedicated fan of HQPlayer. The recently added filters have really made my system sing: more weight and depth with no loss of detail.

Most prefer HQP and you will get no argument from me.  My point was that just by applying ROON's oversampling (to DSD 128/256) there will be a nice gain in SQ without having to purchase anything else (HQP or a more powerful computer needed to run it).  So a free "upgrade"

[barrows]

On 12/31/2018 at 6:39 AM, Tom75 said:

Hi Guys, I have $1000/£1000 for upgrade my system but not quite sure where to put the money in. I have SOtM-200 with Ultracap LPS-1, Chord 2qute with MCRU power supply, Passive Audio PC with linear power supply, PMC 26 with Naim Supernait 2. The cables are all high quality.

 

I am not looking for extra magic or more details, I am more up to the smoother and darker sound. Should I upgrade to Ultra Neo? TX-USB Ultra? Chord Qutest? Or save some money for the later pre and power amp upgrade? 

 

Thanks and happy listening for 2019!

Sell 2Qute and get Qutest, this is a no-brainer upgrade...

[barrows]

On 2/8/2020 at 4:39 AM, FredM said:

Funny that Taiko Audio, Pink Faun and Grimm Audio are all located within a 1,5 hour drive from each other.

 

Just as funny as PS Audio, Ayre Acoustics, and Avalon Acoustics being within 10 minute drive from each other?

[barrows]

5 hours ago, mourip said:

Ghent is great to work with and I love the JSSG360 USB cables he has made for me.

I find it interesting that you like this USB cable, what might you have compared it with?  I ask because it appears Ghent makes this cable with wire not designed for USB, and which does not meet the USB impedance specification.

[barrows]

2 hours ago, RickyV said:

 

The Lush^2 doesn't meet USB specs either but still many people like it. 

I know, that is one of the reasons I am asking.  I owned an original Lush for a brief period (I did not like its sonics).

[barrows]

2 hours ago, MarkusBarkus said:

Any practical experience affixing heatsinks with heat conducting pastes?
 

I have a small aluminum heatsink from Parts-Express "resting" on the eRG, but I could use some goop to further improve the heat transfer. Non-permanent silicon paste.
 

It's sold as a kit, so it includes hardware to screw it down, but I hate to deface the sexy Uptone case with screw holes.

As is, it definitely pulls heat out, based on fin "touch analysis." Helpful to the device? Not sure, but I quite like the stegosaurus look. Old, but new...

Just resting on there will make for good heat transfer, adding thermal paste will probably not help, unless;ess there is some pressure applied (clamping force) to press the sink to the eR's surface.  Thermal paste is designed to be used in an arrangement with clamping force, to insure the paste is just a very, very thin layer.  If you must keep the heatsink removable, perhaps a double sided thermal tape would be the best way to affix the heat sink, which would provide good thermal transfer, and still be removable.

[barrows]

10 minutes ago, MarkusBarkus said:

Thanks Barrows! Howze my PS build going? I have to replace this Shanti. 😉

Shipping today!  Everything has slowed a little bit due to COVID-19.

[barrows]

10 minutes ago, MarkusBarkus said:

...I dunno, gets pretty hot here in summer. 😉

 

Actually, I had no concerns about the case melting. It seems to me if you're pulling heat out of the device, that would be a net-positive. I suppose the caveat to that would be if you were creating inconsistent temps/fluctuations that impacted components that preferred a steady-state operating temp. Then again, I was a humanities major, so...

Of course the heat sink on the case is helping, if it was not drawing away heat, then the sink would remain at room temperature!  I realize that this approach is not the "best" way to cool the internal chips, as these chips are not in direct contact with the case.  But, one has to consider that the air inside the case is transferring heat from the chips, to the case, and hence to the heat sink on top of the case.  

Could the heat dissipation be better with direct contact tot he chips, of course it could, but this does not mean that the sink on the case is not lowering the internal temperature of the eR, it has to be, because the heat sink can be seen to rise above room ambient temperature.  Remember, the case has very little distance internally, from the PCB to its surface, and will conduct heat both by radiation and convection to the case.

 

When I run the microRendu v 1.4 in my bedroom system at DSD 512, it gets very hot, so I added an external heatsink, and this lowers the case temperature by about 5-6 C. this will help with long term reliability.

 

Also, remember it is not just a single chip in the eR which is creating all that heat, because there are multiple linear LDOs in there which drop significant voltage, so just dealing with a single chip is not what one wants to do. Thinking about it, a little ventilation for the case could do wonders as well, especially if it was applied in such a fashion as to encourage convective airflow though the case...  The case is small, so the internal air volume heats up a lot.

[barrows]

2 hours ago, Blackmorec said:

Hey Todd,

You may be interested in what led to this ‘blooming’ of LPSs in my system. 

At the very start, I’d just received a new system after 13 years with an all-tube Balanced Audio Technology - Guarneri Homage system which i loved but hadn’t updated for 6 or 7 years. I’d decided to make a radical change and implement a fully digital amp/DAC  and Server combination for local and remote streaming. Quite a change for a tube-head with a large CD collection. 

 

My system was duly ordered but there was a hold up of the Devialet amps due to the massive Worldwide Core Infinity upgrade which involved new mother boards for all their Expert Pro product line, so my system only arrived 2 months later. Idle hands.....👿

So in the meantime I ordered a Michell Orbe, SME IV and Ortofon Cadenza Black to fill the empty shelf of my rack. 

I also ordered an Innuos Zenith SE with 10m of Synergistic Research Active SE ethernet cable to connect it up (My system uses SR power and speaker cables and I’ve been a ‘loom’ fan for years after my first experience with Shunyata)

 

My amp and speakers duly arrived and I spent a very relaxing afternoon setting up my turntable, arm and cartridge. The reason I ordered the Michell SME unit is that set up is all about calibrated measurements rather than the black magic of trying to get 4 independent springs to bounce uniformly in 3 dimensions. So having got all the set points and measurements exactly to their desired position I fired everything up, using a host of Mint Decca and EMI classical records and a couple of dozen 180g re-issues from HMV. 

Wow....this brought back memories. Firstly the smell of old LP covers, then the vinyl sound. There’s something about dragging a precisely ground diamond through a groove in some plastic  at a precisely regulated speed that has a sound character all of its own.  All the usual attributes were great....wide, deep soundstage, plenty of air, atmosphere and detail and that lovely vinyl warmth to the sound. So I’m Cooking on Gas!!  Loving the new amp and speakers and finding the return to vinyl pretty awesome. 

And then the Innuos Zenith SE arrived. I have 3 branches of large electronics chain stores close by so I could try a few different cables, routers, mesh networks etc  in order to find the best way to stream over a network.  In the end I found that a really clean wi-fi   link gave me the best SQ by a significant margin over 10m Synergistic ethernet cable. No-one more surprised than me, especially given what the SR cable had cost.😏

So with everything connected and run in, the new Zenith SE was sounding magical...better than I’d ever heard and way better than the Michell Orbe, which in comparison sounded ‘old’. The digital was quieter (of course),  but also faster (PRaT), more dynamic generally with much better micro-dynamics due to the silent background. It had better soundstage, better imaging specificity and focus and it was a joy to use ergonomically with its iPad user interface. So the TT and the 180 used records and 20-30 vinyl reissues very quickly became redundant, then idle.  Shame, because it looked very nice on my top shelf and I really enjoyed handling vinyl, but why go to all the trouble with record sleeves, TT clamps, arm queuing etc., when it sounds inferior?

At this point I was still learning the basics about digital hi-fi and I read on this august forum about the benefits of USB reclockers. I was utterly delighted with how my system was sounding.....nothing I didn’t like, but this is a hobby, so I went ahead and ordered an AQVox SE and a SoTM Tx-USBUltra, sPS-500 supply and some high-end USB cables on a 30 day trial basis. The units arrived and were installed on a spare anti-vibration amp base. Sound-wise I could hear some differences, but overall the magic was gone. It was simply not so much fun to listen to. After 400 hours and some minor improvements, but no overall breakthrough I contacted the dealer for a return. I really didn’t understand why I wasn’t getting SQ improvements as practically everyone else was reporting substantial upticks. With the unit removed and replaced by a (very good) Chord C line USB cable, the magic instantly returned.  I discussed this with the dealer and he wondered if the unit could have been faulty so he kindly offered to send a replacement, to be used with its dedicated SMPS wall wart.  The unit arrived, was cabled and plugged in and immediately the magic was gone. 

I wondered if it could be to do with the USB cables I was using so I got some other really good Tellurium units to try.  I got the new units and did a little comparison. The Tellurium USB cables were superior to the Chord...no question. So then I decided to compare music via the Tellurium direct compared to music via the TX-USBUltra. With the TX-USBUltra in circuit, no magic...with the Tellurium cable only....no magic. Hey, what the hell. The Tellurium was working fine before I plugged in the TX-USBUltra and now it wasn’t. But the TC-USBUltra wasn’t in circuit and was getting no data stream as I’d removed its USB input cable. So I pulled the mains plug on the TX-USBUltra and hey-presto...Magic back. 

My conclusion;  my system had a very distinct dislike for switched mode power supplies!  Especially those using the same dedicated mains circuits as my hi-if. Had I got a decent LPS instead of the SoTM sPS-500 it would have probably delivered what others were hearing.  Anyway this got me thinking about the wall wart of the AQVoxSE switch. A leaflet included with the switch made the very bold statement that the SMPS had been fully optimised and even much better and more expensive LPSs would not sound better. Well, a Sean Jacob’s DC proved that advice groundless and a DC3 even more so as improvements in SQ were massive. So that’s how the whole thing got started.

 

I’m currently having to stay home under quarantine laws, so this is a very pleasant way of spending a morning. Thanks for taking the time to read it🙂 

 

 

I applaud you for not just following the crowd, and being willing to test every approach in your own system, and make changes based on your observations despite what other's may say about what sounds "better"!

Thanks for sharing this path...

 

On SMPS, yes many of them are a problem and should not be in the audio system.  But it is important to realize that all SMPS are not created equal.  Most SMPS are designed to be as affordable as possible, and only address noise enough to pass FCC regulations (in the US) on interference and not be high quality in terms which woudl affect audio systems.  But, sophisticated well designed SMPS can be low noise and high performance, so we should not just reject them out of hand.  I have been making some power supplies for my computer gear using very high performance medical SMPS modules, followed by some post filtering (capacitors) and then a discrete regulator circuit for the output, and these approaches produce results which are indistinguishable from that are very high performance linear supplies.  We can also note that DACs such as the EMM labs and the Mola Mole Tanbaqui both use SMPS without any problems, of course both of these companies have some pretty advanced engineers designing for them!

 

[barrows]

2 minutes ago, Fourlegs said:

 

Sure but it rather depends on what you mean when you say that your SMPS build are "are indistinguishable from that are very high performance linear supplies." One would need to say which LPS you were thinking of and also whether you are only referring to measurements or to sound quality tests. Personally I have listened to DACs powered by high quality SMPS and the same DAC by very high performance linear supplies (I'm thinking DC4 when I say that) and the latter are soooo much better.

I develop high performance linear power supplies for audio products professionally, and am quite confident that these supplies are the equal of anything available from anywhere.  These supplies are evaluated both through measurements and by listening.  It really is not very difficult to design a state of the art linear power supply as long as the budget for parts is not limited, despite the cult like following for some power supplies out there...

 

If you heard a DAC sound better with a linear, the only reason for that would be that the SMPS used was not as good as it could be.  There is not magic at work here, power supply requirements and design and very well understood things.  Note though, that the SMPS based supply I was talking about would not be any more affordable as a product than a SOTA linear, so there is not "savings" available from designing a power supply this way, the advantages come in terms of size and efficiency.  Also there are some technical advantages to SMPS operating at higher than audio frequencies as the noise produced is well out of the audio band-although it is important to point out that any of the "problems" associated with either linear, or SM power supplies can be solved by good design.

[barrows]

4 minutes ago, ASRMichael said:

Hi you mention you develop linear power supplies. Are they widely available to buy? If so, where? Cheers 

this is OT, I'll send a PM....

[barrows]

On 10/20/2020 at 8:17 AM, charlesphoto said:

 

I've had good success with the iFi AC iPurifiers plugged into my server closet power strip (a Supra with no surge). I use two, one at the beginning of the strip and then one halfway that separates the modem, main switch etc from the Roon server and oM's. I also have one each on my two dedicated lines for the front end of the system, and one on my office setup. I had two on my DAC V1 line for a it and one with the server, but found two on the server and one on the DAC was just a bit better as the server closet is on a really busy office circuit. Simple, inexpensive, easily moveable and testable, peace of mind from surge protection, and not a four figure beast sitting there with expensive unused outlets. 

 

The Uptone charger SMPS I find to be quite good, as good as the first gen Ifi wall warts at least, though limited to 7.5v. I'll have to look into the X version. The Sonore Power Supply that Barrows builds is really good as well, if somebody is looking for a US LPS instead of the PH wait or Sean Jacobs $$ (who built a really nice power cable for my Naim 160 that has a strange bulgin connector). But again limited to either 5 or 7v. I use the 7v with an SR black fuse and Shunyata NR V10 cable with my opticalrendu. 

The Sonore Power supply could be available at different voltages on a custom order basis...  Anybody interested in that could send me a PM to discuss your needs.

[barrows]

5 minutes ago, HumanMedia said:

would swapping the ultraRendu out for an opticalRendu be a sound quality upgrade? Significant?

I would say, yes, definitely.  The oR is an entirely new design with many improvements over the uR besides the optical connection (which is also an advantage).

[barrows]

24 minutes ago, soares said:

Hi Barrows, Tks for your reply. Not sure however if you have seen my earlier post about my specific configuration:

Router< RJ45 >oM <SFP> Buffalo 2016 <RJ45> Zen MKIII <RJ45> oM <SFP> eR <RJ45> uR <USB> iR <USB> DAC

While I am confident that the oR is better than the uR,      I am unsure about the final results with so many upstream efforts aiming at getting the most clean signal before the uR. So I would also appreciate your views regarding this specific scenario. Thank you, Jorge

Point being:  The Renderer is the last device in the (Network) chain, and it is the only upstream device connected to the DAC.  The DAC is the most sensitive component in the chain because it is a mixed signal device-that is the DAC is both a digital AND an analog component.  The Renderer creates the USB signal, and the cleaner and lower noise that USB signal is, the better the DAC will perform.  Hence, the capability of the Renderer is very, very important.  The oR has many improvements over the uR, it is not just about the optical isolation.

[barrows]

3 hours ago, ssh said:

Hi Barrows,

 

What do you think is the more significant improvement: uR to oR vs. oR to Signature Rendu?

Thanks!

There are too many differences to be able to point out a single one.  Firstly, the optical isolation is probably the biggest, but that is conjecture.  There are lot more isolated power domains on the main board, and most of these power domains are powered with LT 3045 regulators now, the uR uses older regulator designs (still very good linear regs, but not as good as the newer LT 3045s).  Additionally the oR has two Femtosecond level clocks, to handle all clocking duties, where the uR has only a single Femto level clock.  There are many other improvements, but I do not want to spell out the all the details to our competitors publicly.  Suffice it to say, there are many, many improvements besides the optical input.  The Signature Rendu SEoptical is even better, but of course that comes at a much higher cost.

[barrows]

1 minute ago, fds said:

Any experiences out there with inserting a Phoenix between OpticalRendu and DAC? 

Here is the thing.  The USB output of the oR is already designed to be as good as it can be, and IMO, adding additional USB fixer upper devices would be redundant.  The USB output of the oR is already clocked by a dedicated Femto level oscillator, and the USB power supply is already ultra low noise with a dedicated LT 3045 regulator and appropriate decoupling right at the USB output jack.  I would advise that one connect the output of the oR directly to the DAC with the very best USB cable one can get their hands on, and leave it at that. And if your DAC uses USB power, make sure your USB cable pays specific attention to the power leg design, and keep the cable as short as is reasonable to keep power delivery low impedance/inductance. 

[barrows]

8 hours ago, fds said:

Using the oR, I believe there is no possibility to switch of the WIFi or LAN connection to the outside world. Or has anyone maybe found a way to get music from the OpticalRendu without being connected to the WWW?

 

Same applies to all users of Euphony/Stylus to my understanding. Or is there someone there who has found a way to use this without being connected to the WWW?

The opticalRendu, and all other Sonore Rendu's for that matter, do not need an Internet connection to be able to play music.  They do need an Internet connection to boot up properly, and for software updates, etc.  but they will play music with no Internet connection once up and running.

I use the Signature Rendu SEoptical here, and I live in the mountains with a rather poor Internet service.  My system plays music happily with no difficulties when the Internet is down, as long as I need not access the (web based) GUI which allows for changing the settings on the Rendu.

[barrows]

6 minutes ago, fds said:

wifi connection to the Mac server was interrupted

The above is not the same as "Internet".  Your WiFi connection to your Mac server has nothing to do with the Rendu, that would be related to what player software you are running and how it is controlled over WiFi. 

[barrows]

1 minute ago, RickyV said:

You could shield the flat cable of the active output.

Clock cables should be impedance controlled SMB or u.fl coax cables, using dedicated SMB or micro BNC connectors when possible.  A couple of CM of wire is likely OK, but anything longer should be proper coax with proper terminations.  It is counterproductive to spend $$$$ on a fancy clock, and then connect it by a single wire flying lead.

[barrows]

I would like to order a 25 MHz variant, but looking at the pages at diyaudio, my mind starts to go numb and I just cannot figure it out with all of the options.

 

If I want a single output, 25 MHz, clock board, what exactly do i need to order and where?  I just do not have hours to spend on figuring this all out.  I am able to solder and DIY fine with the parts sizes I see in the pictures.

[barrows]

I am not sure I am that patient, even with Optane memory as a cache, waiting for a 90-120 minute DSD 512 album file to fully load before playback seems a bit bothersome, as does taking all that time to oversample and store all music files!  But it is no more trying than dealing with LPs I guess.  Although I personally prefer digital playback and I like the convenience of nearly instant playback.

[barrows]

3 hours ago, HumanMedia said:

Go on @cool_chris - take him up on it, what is one more to compare and this might be even better. You are the best person to compare them. Wont you need two, one at each end to match receiver sensitivity?

I am using a pair of the 1318s here and I like the sound, but I do not find large differences between good quality SMF SFPs, but perhaps these are also system dependent, I use the Sonore Signature Rendu SEoptical as my Renderer/NAA and it isolates the different sections of the internal circuitry pretty well, and also uses very sophisticated all linear power supplies and internal regulation stages. 

[barrows]

4 hours ago, manueljenkin said:

Thank you very much! It feels analogous to directly heated tubes.

 

I have a few doubts. Clocks as I understand are mainly generated as square waves (or ideally that's what we want for digital logic, and possibly also dacs if they use these clock to generate the output, like AND of signal with clock). And crystal oscillators as I understand general something sinusoidal or a distorted sinusoid depending upon the piezoelectric non linearities and mechanical restoring force non linearities (if I'm wrong please feel free to correct). How is this converted to squares? Do we just generate a high amplitude one and send through a logic that clips this (I think something like this + a high pass filter is used to create higher clock speeds)? Or is it done by some other way?

 

Do these oscillators exist mostly as is or are they supported by a transistor amplifier (or oscillator) for better signal strength or as a buffer? If so what are considerations for the amplifier to be better for high fidelity audio applications?

 

8.20 in this video. 

if such tiny amount of material can affect the vibration pattern of the oscillator, what about the impact of physical vibrations (due to mild ground shake, or speaker sound vibrating the PCB), how do they affect the performance of an oscillator? Anything about impact of rf noise, or PCB design choices, solder/mounting choices?

 

I hope I'm not trying to ask too much but one last question. What considerations should go into the design of power supply to the crystal oscillator (or in other words what is the electrical load character of the crystal oscillator) - in terms of output impedance, stability, noise, topology matching etc! I remember something called voltage controlled oscillator, are they ever used for audio? (If they are, then pretty sure a stable voltage source is a must 😅).

For the most part, high end manufacturers use oscillators which output a square wave (they have the necessary circuitry internally to filter and convert the rough sine wave output from he crystal into a square wave).  Yes, all clocks are sensitive to vibrations, and components which address this in how they implement their clocks will have an advantage.  Also, paying attention to vibration control of the entire component (good racks, aftermarket footers, etc) can also improve performance.

Note also that OCXO are not necessary for the best performance in audio components.  The oven actually does not help much with performance for audio, the oven actually helps much more with the long term accuracy of the clock-think over hours, or days, rather than over the µS that are important for audio.  Consider, we do not listen to a single note over a period of hours.  For audio performance, low close in phase noise (phase noise at low frequencies) is what is important for audio, so the popular belief that OCXO are best for audio is partly in error.  But, some of the best clocks for low close in phase noise (and not all, this is important, some OCXO are worse than the better non-ovenised oscillators) also happen to be OCXO, so they may be used.  So, be aware that just because a clock is an OCXO does not indicate, necessarily, that it is a really good clock.

For example, in order to get a really good clock for their DAC, Ayre contracted with OCXO maker Morion in Russia, to have them make a really good clock, with an SC cut crystal, but without the oven  This way they got a clock with very low close in phase noise, but without the additional expense of the oven circuitry.