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

[romaz]

10 hours ago, greenleo said:

@Roy,

 

Have you any tools to tell if an X25E is new or just trust the seller?

 

rickca is right.  Here's a free tool that's popular:

 

https://downloadcenter.intel.com/download/26574/Intel-Solid-State-Drive-Toolbox

[greenleo]

Thank you Rickca and Roy.

[Johnseye]

55 minutes ago, romaz said:

 

Understood, but both the mR and sMS-200 present us with an ideal model to try and emulate.  Good low power X86 processors with large secondary caches exist now.  No need to go to non-X86 CPUs and resort to emulation:

 

https://ark.intel.com/products/95441/Intel-Core-i7-7Y75-Processor-4M-Cache-up-to-3_60-GHz

 

The problem is the lack of suitable motherboards.  This is why I believe you need one of these types of endpoints if your goal is ultimate SQ since no single box server will ever have all the ideal characteristics that audiophiles desire.

 

I think motherboard manufacturers are seeing a market for audiophile quality components and are including some in their onboard sound, but won't make an audiophile specific board.  It's a niche market at the level we're looking to perfect.  We'll likely start to see more appliances with specialized boards and components, but it might be a while longer before someone builds a motherboard or even a prebuilt PC with a purpose designed board where they'll allow installation of any OS.  I hope it happens, but I think there are challenges to making it profitable.

[Silly goose]

9 minutes ago, Johnseye said:

 

I think motherboard manufacturers are seeing a market for audiophile quality components and are including some in their onboard sound, but won't make an audiophile specific board.  It's a niche market at the level we're looking to perfect.  We'll likely start to see more appliances with specialized boards and components, but it might be a while longer before someone builds a motherboard or even a prebuilt PC with a purpose designed board where they'll allow installation of any OS.  I hope it happens, but I think there are challenges to making it profitable.

From what I hear with audio over IP interfaces like Ravenna, you get the same sound from any server regardless of how clean it is. It's not like with USB. So I would just stop tweaking, and enjoy the music for a bit. Then just buy a DAC with Ravenna once they become mainstream after RMAF. Don't waste your money, or time on the server. 

[Johnseye]

1 minute ago, Silly goose said:

From what I hear with audio over IP interfaces like Ravenna, you get the same sound from any server regardless of how clean it is. It's not like with USB. So I would just stop tweaking, and enjoy the music for a bit. Then just buy a DAC with Ravenna once they become mainstream after RMAF. Don't waste your money, or time on the server. 

 

Not MQA?  I guess I'll believe it when I see it.  I think there are great benefits to be had from low, clean, noiseless power as well as efficient components.  I always enjoy the music.

[Silly goose]

2 minutes ago, Johnseye said:

 

Not MQA?  I guess I'll believe it when I see it.  I think there are great benefits to be had from low, clean, noiseless power as well as efficient components.  I always enjoy the music.

Yes when connected with USB maybe, but not when connected via AOIP. If Tidal upgraded the power supplies on their servers in Oslo, would you hear a benefit from your DAC at home? 

[Johnseye]

Just now, Silly goose said:

Yes when connected with USB maybe, but not when connected via AOIP. If Tidal upgraded the power supplies on their servers in Oslo, would you hear a benefit from your DAC at home? 

 

Maybe yes.  I'm sure AOIP has it's challenges.  Admittedly I know nothing about it.  What about AOE?  I look forward to seeing devices that support these protocols.

[Silly goose]

28 minutes ago, Johnseye said:

 

Maybe yes.  I'm sure AOIP has it's challenges.  Admittedly I know nothing about it.  What about AOE?  I look forward to seeing devices that support these protocols.

Yeah biggest challenge was resolution support. But they are there now with up to PCM 384 and DSD 256. Talking to manufacturers who have been testing the latest and greatest Ravenna board ever developed, they are blown away. It also has a fiber optic PHY on it so that eliminates any possibility of any noise getting into the system. Audiophile Ethernet cable guys will be out of luck once these hit the market. Because for audio use, there's no way to improve upon the fiber optic cables. The only way you'll be able to get better sound out of these DAC's is with internal mod's, or firmware updates. 

[Johnseye]

5 minutes ago, Silly goose said:

Yeah biggest challenge was resolution support. But they are there now with up to PCM 384 and DSD 256. Talking to manufacturers who have been testing the latest and greatest Ravenna board ever developed, they are blown away. It also has a fiber optic PHY on it so that eliminates any possibility of any noise getting into the system. Audiophile Ethernet cable guys will be out of luck once these hit the market. Because for audio use, there's no way to improve upon the fiber optic cables. The only way you'll be able to get better sound out of these DAC's is with internal mod's, or firmware updates. 

 

A long as there's no compression I'm interested.  AoE should work just fine over fiber as well, it's just a medium.

[Silly goose]

16 minutes ago, Johnseye said:

 

A long as there's no compression I'm interested.  AoE should work just fine over fiber as well, it's just a medium.

No the bandwidth of the Ethernet is the only limitation. You can have 1000 channels of DSD 1024 running simultaneously in perfect sync if you have enough network bandwidth. The multichannel version of this board can do 8 channels @ DSD 256 over a standard Gigabit network. That's the boards that will be going into active speakers. For higher resolution support, only a firmware update would be required. But they believe that once you go over DSD 256, higher jitter from the clocks is more detrimental than any gains you get from pushing noise further out of the audible band. But of course in order to determine this, you must have experience with this board, and different clocks. Something they have. Since they developed it, I'd say they are in a pretty good position to make that call. 

[Superdad]

On 4/11/2017 at 4:08 PM, ElviaCaprice said:

 

It's not perfect, no galvanic isolation is perfect, but I have to say, Chords galvanic isolation on it's DAC's is probably the best I've seen (or heard about from others) in the market.  It's that good.  In fact, it has me doubting the benefits of the upcoming ISO Regen, in conjunction with the 2Qute.  I'm getting the best sound I've heard yet with my current setup and that doesn't include any reclockers or additional galvanic isolators.  I think the Topaz (and improved 2Qute power supply) has really brought it together for low impedance throughout my system and even made my poor HDPlex a better power supply.  Instead of adding new expensive reclockers/isolators, I would rather now turn to the server itself and see what kind of additional improvements can be had, whether it's a new power supply, mobo, or clocking.

 

What you say is puzzling in that if you can hear differences at the computer end  it means:

a) such changes are modifying the signal integrity (eye-pattern variables of amplitude, slew, noise, jitter) and leakage current of what is sent to your DAC--as those are the only mechanisms by which bit perfect data can affect the DAC;

b) your DAC is therefore not immune to such vagaries;

so

c) best signal integrity (via regeneration and galvanic isolation) right at your DAC is going to be desired and required in any case.

[Silly goose]

Just now, Superdad said:

 

What you say is puzzling in that if you can hear differences at the computer end  it means:

a) such changes are modifying the signal integrity (eye-pattern variables of amplitude, slew, noise, jitter) and leakage current of what is sent to your DAC--as those are the only mechanisms by which bit perfect data can affect the DAC;

b) your DAC is therefore not immune to such vagaries;

so

c) best signal integrity (via regeneration and galvanic isolation) right at your DAC is going to be desired and required in any case.

With Ravenna, no changes upstream of the Fiber PHY has any impact on the sound. Unless of course it's from DSP. The days of PC and digital signal path hardware tweaks will be long forgotten in the near future. 

 

But for small DAC's like Chord, the FPGA based solution might be a bit too big still. They will likely need to wait for the ASIC in the works. Set for release mid 2018, early 2019, it will finally commoditize  AES67 for the DAC's and active speakers at all price points.  

[Superdad]

On 4/12/2017 at 5:49 PM, Johnseye said:

 

Assuming I went with the HDPLEX 160W DC-ATX would I need Paul to make one of the rails of the SR7 a 19v?  If I go with the Pico it stays 12v but the sound quality could suffer?  Is 12v enough to power a miniATX, or does it depend on the total watts used, or what's being powered by the different voltages?

 

Just to be a bit more clear about both the picoPSU and the HDPlex cards:

They are both switching regulator adapters.  That is, they use DC-DC switching converters to drop the one input voltage to 5V and 3.3V.  

Unlike the picoPSU (XT150 and XT160 versions) which run from 12V and just pass that--ideally LPS  sourced--12V to the motherboard, the 19V input HDPlex boards have to use fairly large switching converters to drop down to for the high-current 12V feed.

 

So regardless of whose power supply you choose, for "ATX" application I'd recommend you go with the 12V input picoPSU to best preserve the "goodness" of your LPS for the critical main 12V powering of your motherboard.  And if you have another supply or an LPS with an extra rail, then you can consider directly powering some important 5V things such as you OS and/or music drives.

[ElviaCaprice]

39 minutes ago, Superdad said:

 

What you say is puzzling in that if you can hear differences at the computer end  it means:

a) such changes are modifying the signal integrity (eye-pattern variables of amplitude, slew, noise, jitter) and leakage current of what is sent to your DAC--as those are the only mechanisms by which bit perfect data can affect the DAC;

b) your DAC is therefore not immune to such vagaries;

so

c) best signal integrity (via regeneration and galvanic isolation) right at your DAC is going to be desired and required in any case.

I'm just presupposing your examining my signature for my system streaming.  That PPA V2 card is reclocking the audio stream and sending it to the 2Qute which performs it's own galvanic isolation at the DAC.  Not sure what your saying, Alex. 

I'm saying that any changes to the server, be it clocks, software, hard drives, cables or memory are going to be a factor in SQ, regardless of regeneration and galvanic isolation.  I thought this was already well accepted in theory?

  Of course the 2Qute is not immune to such vagaries, but I think one needs to carefully choose the components that can improve on SQ, obviously the server is one of them.

I don't disagree that best signal integrity at the DAC is desirable, the question is, which component/clocking can give the best result?  Would double USB galvanic isolation be a negative to SQ?

[romaz]

1 hour ago, Silly goose said:

From what I hear with audio over IP interfaces like Ravenna, you get the same sound from any server regardless of how clean it is. It's not like with USB. So I would just stop tweaking, and enjoy the music for a bit. Then just buy a DAC with Ravenna once they become mainstream after RMAF. Don't waste your money, or time on the server. 

 

This is a bit of a presumptuous statement.

 

First of all, Ravenna has been out since 2010 and AES67 was introduced in 2014 and all of a sudden, it's going to be mainstream after RMAF this year?  The most important part of a digital front end, in my view, is the DAC and not the server or the transfer protocol and none of the DACs that I prefer (presently a Chord DAVE) have plans to release an Ethernet interface this year, this I know for a fact.  There's no way I'm going to buy just any DAC because it has an Ethernet input.

 

Second, based on my recent observations, the Ethernet is a very dirty place and typical Ethernet devices such as switches and routers largely contribute to this.

 

Third, fiber doesn't guarantee good sound.  Look how jitter prone Toslink is over coaxial copper.  Granted, the low bandwidth of Toslink is partially responsible for this but many of us who have implemented this "direct connection" between our server and our AOIP device (microRendu or sMS-200) have done away with our high bandwidth fiber either because it added nothing or worse, it degraded SQ.  During my recent comparison of fiber vs copper Thunderbolt cables, fiber sounded the worst.  While fiber does some good things, it isn't a panacea solution.

 

Fourth, what is accepted in the professional world doesn't necessarily translate well to the audiophile community.  We have different priorities.  I have pro audio friends who are also audiophiles and the equipment they use at work is often different than the equipment they choose for their homes.

 

When the microRendu was first released, it seemed everyone was proclaiming how the source no longer mattered and how all sources with it sounded equally good.  If you've read Chris Connaker's review of the microRendu, then you know what I mean.  Of course, this is similar to what you are proclaiming with Ravenna.  While I have no good explanation for it, it's interesting that with this "direct connection" where the router is removed from the path between server and endpoint, SQ takes a leap forward but all of a sudden, sources no longer sound the same.  Is it possible that with Ravenna, there is something in the network connection path that is making all sources sound the same but may not be so good for SQ?

 

Having had a RedNet 3 in house for evaluation for 4 weeks, I had a chance to carefully A/B Rednet and its Dante protocol against my standard sMS-200 driven by my SR7 and at the end of the evaluation period, the sMS-200 stayed and the RedNet went back.  Part of the problem with the RedNet is that it sounds flat which I attributed to its horrible PSU and I realize people are now taking steps to rectify this but now with this "direct connection" that I implemented only after I returned the RedNet and with the even better sMS-200 Ultra I have moved to, my feeling is the gap has widened even more.

 

I don't know what the future holds.  Digital is evolving very rapidly and it's unlikely what I have today will be with me in a year or two based on my history.  If it turns out Ravenna wins out, I'll gladly jump onto that bandwagon but not because all sources sound the same, but because Ravenna sounds better.  Of course, this is also contingent on my preferred DAC having an Ethernet input.

[Superdad]

15 minutes ago, Silly goose said:

With Ravenna, no changes upstream of the Fiber PHY has any impact on the sound. Unless of course it's from DSP. The days of PC and digital signal path hardware tweaks will be long forgotten in the near future. 

 

But for small DAC's like Chord, the FPGA based solution might be a bit too big still. They will likely need to wait for the ASIC in the works. Set for release mid 2018, early 2019, it will finally commoditize  AES67 for the DAC's and active speakers at all price points.  

 

Ha ha.  The "Ravenna revolution" has been just a few months away for almost 5 years now.  

I'm not saying it (or something like it) won't ever happen, but:

a) the hardware is not the only obstacle to adoption, and

b) as those here who are using the Dante/Focusrite/Rednet boxes have discovered, it is not 100% immune to all factors (i.e. those guys are "tweaking" too).

 

So wake us when the "revolution" starts and you have some actual board and software solutions to point to. 

[Silly goose]

2 minutes ago, romaz said:

 

This is a bit of a presumptuous statement.

 

First of all, Ravenna has been out since 2010 and AES67 was introduced in 2014 and all of a sudden, it's going to be mainstream after RMAF this year?  The most important part of a digital front end, in my view, is the DAC and not the server or the transfer protocol and none of the DACs that I prefer (presently a Chord DAVE) have plans to release an Ethernet interface this year, this I know for a fact.  There's no way I'm going to buy just any DAC because it has an Ethernet input.

 

Second, based on my recent observations, the Ethernet is a very dirty place and typical Ethernet devices such as switches and routers largely contribute to this.

 

Third, fiber doesn't guarantee good sound.  Look how jitter prone Toslink is over coaxial copper.  Granted, the low bandwidth of Toslink is partially responsible for this but many of us who have implemented this "direct connection" between our server and our AOIP device (microRendu or sMS-200) have done away with our high bandwidth fiber either because it added nothing or worse, it degraded SQ.  During my recent comparison of fiber vs copper Thunderbolt cables, fiber sounded the worst.  While fiber does some good things, it isn't a panacea solution.

 

Fourth, what is accepted in the professional world doesn't necessarily translate well to the audiophile community.  We have different priorities.  I have pro audio friends who are also audiophiles and the equipment they use at work is often different than the equipment they choose for their homes.

 

When the microRendu was first released, it seemed everyone was proclaiming how the source no longer mattered and how all sources with it sounded equally good.  If you've read Chris Connaker's review of the microRendu, then you know what I mean.  Of course, this is similar to what you are proclaiming with Ravenna.  While I have no good explanation for it, it's interesting that with this "direct connection" where the router is removed from the path between server and endpoint, SQ takes a leap forward but all of a sudden, sources no longer sound the same.  Is it possible that with Ravenna, there is something in the network connection path that is making all sources sound the same but may not be so good for SQ?

 

Having had a RedNet 3 in house for evaluation for 4 weeks, I had a chance to carefully A/B Rednet and its Dante protocol against my standard sMS-200 driven by my SR7 and at the end of the evaluation period, the sMS-200 stayed and the RedNet went back.  Part of the problem with the RedNet is that it sounds flat which I attributed to its horrible PSU and I realize people are now taking steps to rectify this but now with this "direct connection" that I implemented only after I returned the RedNet and with the even better sMS-200 Ultra I have moved to, my feeling is the gap has widened even more.

 

I don't know what the future holds.  Digital is evolving very rapidly and it's unlikely what I have today will be with me in a year or two based on my history.  If it turns out Ravenna wins out, I'll gladly jump onto that bandwagon but not because all sources sound the same, but because Ravenna sounds better.  Of course, this is also contingent on my preferred DAC having an Ethernet input.

Nobody has ever experienced this level of digital interface before. These boards have been in the works for a couple of years now. Finally the are just getting into the hands of the manufacturers. You can't compare fiber Ethernet to toslink. Because toslink carry's an audio clock. And Ethernet doesn't. If jitter was a problem with Ethernet, when streaming Tidal and the 10000 switches it goes through from the main servers to your DAC, the sound would be a horrible mess. The Rednet devices are subpar at best. They are not designed for audiophile purposes, and certainly nowhere near the level of this board I'm talking about. The clocks, power supply and everything is subpar. And then you need to go through subpar AES or Spdif to get to the DAC. Simply nowhere in this league, and not at all comparable. These boards get installed directly into the DAC. 

[Superdad]

2 minutes ago, romaz said:

 

This is a bit of a presumptuous.......

I don't know what the future holds.  Digital is evolving very rapidly and it's unlikely what I have today will be with me in a year or two based on my history.  If it turns out Ravenna wins out, I'll gladly jump onto that bandwagon but not because all sources sound the same, but because Ravenna sounds better.  Of course, this is also contingent on my preferred DAC having an Ethernet input.

 

Geez Roy, how the heck did you write and type that so fast?!  Great points though, and you did not even get into the issue of virtual sound card software, licensing, and if that stuff is really going to be transparent.

 

[But of course you have seen Silly Goose honking about Ravenna religion in a half dozen threads each day this week.]

 

[Silly goose]

15 minutes ago, Superdad said:

 

Ha ha.  The "Ravenna revolution" has been just a few months away for almost 5 years now.  

I'm not saying it (or something like it) won't ever happen, but:

a) the hardware is not the only obstacle to adoption, and

b) as those here who are using the Dante/Focusrite/Rednet boxes have discovered, it is not 100% immune to all factors (i.e. those guys are "tweaking" too).

 

So wake us when the "revolution" starts and you have some actual board and software solutions to point to. 

It's already a done deal. Manufacturers have them in hand. VSC's for Windows, MAC and Linux are complete. The products they are going into are being built. None of those Chinese pro sound grade boxes with SPDIF output are anywhere in this league. None of them have OXCO clocks in them, and ultra low noise power supplies, or fiber optic inputs. And any gains from the Dante boards get mostly lost in the SPDIF conversion after them. These boards are designed for the Audiophile market. And built to the very highest standards. They are specifically made to raise the bar in the worlds finest DAC's and active speakers. Nothing from the past can be remotely compared to them. 

[Silly goose]

8 minutes ago, Superdad said:

 

Geez Roy, how the heck did you write and type that so fast?!  Great points though, and you did not even get into the issue of virtual sound card software, licensing, and if that stuff is really going to be transparent.

 

[But of course you have seen Silly Goose honking about Ravenna religion in a half dozen threads each day this week.]

 

I wouldn't expect you to be on board. After all they will completely kill the sales of your digital interface products. And that's what the DAC manufacturer's want. They are extremely tired of their clients thinking they need gadgets in the path to get good sound from their DAC's. Finally with fiber optic inputs, that market will be killed off. 

[Superdad]

17 minutes ago, ElviaCaprice said:

I'm just presupposing your examining my signature for my system streaming.  That PPA V2 card is reclocking the audio stream and sending it to the 2Qute which performs it's own galvanic isolation at the DAC.  Not sure what your saying, Alex. 

I'm saying that any changes to the server, be it clocks, software, hard drives, cables or memory are going to be a factor in SQ, regardless of regeneration and galvanic isolation.  I thought this was already well accepted in theory?

  Of course the 2Qute is not immune to such vagaries, but I think one needs to carefully choose the components that can improve on SQ, obviously the server is one of them.

I don't disagree that best signal integrity at the DAC is desirable, the question is, which component/clocking can give the best result?  Would double USB galvanic isolation be a negative to SQ?

 

Hi:

Sorry, I actually did not look at your signature.  Of course you already have a lot of things optimized. 

 

As for "double galvanic isolation":  Remember that any DAC with digital isolators places them AFTER the USB PHY and processor--on the I2S lines headed into the DAC chip.  Thus anything you are hearing from upstream chamges (USB cables, cards, etc.) just proves that the DAC input is not immune.  So an extra stage of GI that blocks ground noise and leakage currents from ever entering shows only be a good thing.

Of course we'll all find out soon enogh if our theory (and hard work) pays off. 

[romaz]

10 minutes ago, Silly goose said:

Nobody has ever experienced this level of digital interface before. These boards have been in the works for a couple of years now. Finally the are just getting into the hands of the manufacturers. You can't compare fiber Ethernet to toslink. Because toslink carry's an audio clock. And Ethernet doesn't. If jitter was a problem with Ethernet, when streaming Tidal and the 10000 switches it goes through from the main servers to your DAC, the sound would be a horrible mess. The Rednet devices are subpar at best. They are not designed for audiophile purposes, and certainly nowhere near the level of this board I'm talking about. The clocks, power supply and everything is subpar. And then you need to go through subpar AES or Spdif to get to the DAC. Simply nowhere in this league, and not at all comparable. These boards get installed directly into the DAC. 

 

I'm open to anything that sounds better but similar to the hype of MQA, right now Ravenna is just that, more bark than actual bite.

 

As for the RMAF 2017 time frame, color me skeptical.  Until then, I'll continue on with wasting my time and money on servers.  It's just what I do.

[Silly goose]

Just now, romaz said:

 

I'm open to anything that sounds better but similar to the hype of MQA, right now Ravenna is just that, more bark than actual bite.

 

As for the RMAF 2017 time frame, color me skeptical.  Until then, I'll continue on with wasting my time and money on servers.  It's just what I do.

Well I'm getting all my info from guys with hands on experience with them. And these guys have extensive experience with all other digital interface protocols as well. So I have complete confidence in this. The R&D costs to build these were in the millions. This is why small potatoes like Uptone will never be able to compete. They will need to wait for the ASIC when it's for sale at Digikey if they ever want to use this technology. 

[ElviaCaprice]

13 minutes ago, Silly goose said:

Well I'm getting all my info from guys with hands on experience with them. And these guys have extensive experience with all other digital interface protocols as well. So I have complete confidence in this. The R&D costs to build these were in the millions. This is why small potatoes like Uptone will never be able to compete. They will need to wait for the ASIC when it's for sale at Digikey if they ever want to use this technology. 

 

Uptone doesn't need to compete with that which doesn't exist yet, if ever??  Uptone can switch gears much faster than a bigger company can.  So meanwhile can you start a different thread with this?  Until it becomes a reality, it really doesn't pertain to our discussion here.  But thanks for letting us know about it, definitely be looking for it, in time.

[romaz]

1 minute ago, Silly goose said:

Well I'm getting all my info from guys with hands on experience with them. And these guys have extensive experience with all other digital interface protocols as well. So I have complete confidence in this. The R&D costs to build these were in the millions. This is why small potatoes like Uptone will never be able to compete. They will need to wait for the ASIC when it's for sale at Digikey if they ever want to use this technology. 

 

It would be more credible if you weren't relaying 2nd hand information, which is hardly worth anything.  It would be more credible if you just came straight out and admitted you are in this industry and have vested interest in its success.  That way, I can at least understand why you would evangelize so heavily for a product that you supposedly have no experience with.

 

As I stated, if your product has merit, let it speak for itself.  As for your disrespectful comments toward Uptone, not cool, bro.  You won't win yourself any converts that way.