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

[simonklp]

59 minutes ago, austinpop said:

 

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

 

Thank you very much for your reply and your effort in preparation of the index.

 

In fact, I have browsed through quite a few of the posts shown in the index. Through the index, I  have located your post quoted in my above post.

 

But my question was that "is there any post related to experiment specifically using HQPlayer with NAA and found the SQ improvement by replacing the clocks in  the devices at the upstream of the DAC?"

 

Sorry for raising this maybe ignorant question that appears to be repetitive or redundant, because I have no idea how to find the related posts.

 

Thank you again for your kind advice in advance.

[tedwoods]

3 hours ago, simonklp said:

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

You are raising a very valid point.

The fact that we all possess different systems and we also use different formats and software, makes it very difficult to isolate info that directly relates to an audio situation that closely resembles our own.

It would be interesting to see if there's any interaction  between improved clocking and the different software used, HQPlayer NAA in particular, from members who have been able to perform their own listening comparisons between such.

 

[dgarretson]

8 hours ago, zephyr24069 said:

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

One thing I like about the SRS rubidium master clock and the Antelope LiveClock distributor is their ability to accept external 12Vdc power.  I've been using a Hynes SR7 for this.  It does make an improvement relative to the stock SMPSs, and should be a consideration for comparisons at the high end.

[zephyr24069]

2 hours ago, dgarretson said:

One thing I like about the SRS rubidium master clock and the Antelope LiveClock distributor is their ability to accept external 12Vdc power.  I've been using a Hynes SR7 for this.  It does make an improvement relative to the stock SMPSs, and should be a consideration for comparisons at the high end.

Good point,...if you are going to have a unit that takes external 12v power, it is best to use a bespoke audio-grade power supply.

In the case of the Cybershaft, the PS is internal but uses a Schottky Barrier diode-enable power supply in their best units.

[simonklp]

3 hours ago, tedwoods said:

You are raising a very valid point.

The fact that we all possess different systems and we also use different formats and software, makes it very difficult to isolate info that directly relates to an audio situation that closely resembles our own.

It would be interesting to see if there's any interaction  between improved clocking and the different software used, HQPlayer NAA in particular, from members who have been able to perform their own listening comparisons between such.

 

 

Thank you very much for your comment. You have elaborated the point much more clearly. This is exactly what I want to find out. Thanks again.

[austinpop]

10 hours ago, simonklp said:

 

Thank you very much for your reply and your effort in preparation of the index.

 

In fact, I have browsed through quite a few of the posts shown in the index. Through the index, I  have located your post quoted in my above post.

 

But my question was that "is there any post related to experiment specifically using HQPlayer with NAA and found the SQ improvement by replacing the clocks in  the devices at the upstream of the DAC?"

 

Sorry for raising this maybe ignorant question that appears to be repetitive or redundant, because I have no idea how to find the related posts.

 

Thank you again for your kind advice in advance.

 

7 hours ago, tedwoods said:

You are raising a very valid point.

The fact that we all possess different systems and we also use different formats and software, makes it very difficult to isolate info that directly relates to an audio situation that closely resembles our own.

It would be interesting to see if there's any interaction  between improved clocking and the different software used, HQPlayer NAA in particular, from members who have been able to perform their own listening comparisons between such.

 

 

Ah OK I see now. 

 

I don't use HQPlayer, other than trying it early on, and realizing that for my DAC (and I stress this, as it is very DAC-dependent), native resolutions fed to the DAC sounded best.

 

But perhaps others who have deployed sCLK-EX based Ultra chains are using HQPlayer, so they will be the ones to chime in.

 

I suspect the SQ improvements from the clock are applicable across sample rates. In my own library, I have music from Redbook (16/48) to DXD (24/384), and the improvements are as pronounced on either end of the range. It is true that some people use HQPlayer to upsample to DSD512 and beyond, and I cannot speak to that.

[atxkyle]

On 8/13/2017 at 11:12 AM, austinpop said:

 

Yes, indeed. Awesome to have another Austin based CA'er.

 

 

Yes, correct, since it strips down WIndows to its bare bones, it's best suited for a dedicated PC.

 

Look in the index on the first post for some links to folks' experiences with AO, and other tools like Process Lasso (PL).  This isn't a make or break, just something to consider at some point in your journey.

Thanks for the advice. I decided to get a dedicated Roon server (Intel NUC i3) and am now running it bare bones in Windows server 2012 r2 Essentials in minimal server mode with Audiophile Optimizer.  Also upgraded the microRendu to the 1.4 board with the upgraded clock.  These 2 changes definitely have taken my system to next level!

[fibonacci1618]

Hello everyone, I have read chunks of this thread with great interest and have recently purchased a sotm 200ultra (9V rated version). I'm deciding now which PS to feed it as I intend to replace the stock wallwart that came with it.

 

I know from reading many of the posts here that the Ultracap LPS-1 comes very highly regarded. However, I have also recently read some posts on the new sotm sps-500. The price difference is about $100 between the two and I do know that the LPS-1 is limited to supplying 1A at 7V, whilst the sps-500 is more versatile (and does up to 3.3A at 19V).

 

My intention (since I am new to experimenting with the PS) is to limit my purchase to a single power supply until I build up some experience and verifying for myself that this is all real vs. the stock wallwart.

 

So, my questions are,

 

(i) will the LPS-1 (a single unit) sufficiently power the 200ultra (9V rated version), and if so -

 

(ii) which one will sound better with the 200ultra? The LPS-1 or the sps-500?

 

I know the word "better" is very vague, and so I hope someone can describe as best as possible, in what ways "better" means. 

 

Thank you all in advance, and I look forward to contributing too sometime soon.

 

Melvin

 

[Confused]

@fibonacci1618  I cannot answer all of your questions, but I can advise that there is one additional piece of information from yourself that would be useful in accurately answering this one.  Can you advise what the sMS-200Ultra will be feeding?  This is an important point because if the USB input of whatever you are using draws current from the 5v USB feed, then this could easily push you over the current limit for the LPS-1.

[fibonacci1618]

2 minutes ago, Confused said:

@fibonacci1618  I cannot answer all of your questions, but I can advise that there is one additional piece of information from yourself that would be useful in accurately answering this one.  Can you advise what the sMS-200Ultra will be feeding?  This is an important point because if the USB input of whatever you are using draws current from the 5v USB feed, then this could easily push you over the current limit for the LPS-1.

 

Yes, thanks for this @Confused. The 200ultra is feeding a Denafrips Terminator R2R DAC via USB, so it should not be drawing any additional current from the 200ultra (I believe).

 

The 200ultra is being fed by a sonicTransporter i5 via a bridged ethernet (actually a USB 3 to ethernet converter) connection.

 

Hope this completes the picture.

 

Thanks!

 

Melvin

[Confused]

There have been one or two recent posts here that have picked up on the question of how a modified switch with clock feed can improve sound quality in a system where Ethernet packet data is effectively buffered prior to being converted to a data stream.  @simonklp questions were specifically related to HQPlayer, which is fair enough, but I believe that in principle this question relates to most common streaming protocols.  (I know there are some exceptions in pro audio)  Of course, it is easy to use basic knowledge of packet data and buffering to simply dismiss everything that may be happening here.  However, I would far prefer to think that the experience and efforts of those reporting the improvements from the modified switch suggest strongly that there may be something very real happening here.  Indeed, I have followed this thread quite closely and it seams to me that that the modified switch has been one of the 'stars of the show', a simple thing that has offered a surprising result.  In fact, 'surprising' might be the key word here.  Most experienced audiophiles will be all too familiar with the problems surrounding expectation bias, confirmation bias and so on, but here we have one thing that perhaps nobody expected to yield an improvement, and yet we have now had multiple results that it has done just that.  So having said that it is easy to dismiss this stuff, I would say the more interesting challenge is to try to understand how this works and why this may be a very real working phenomenon.  I know there has been some speculation, but the real answer could of course be something quite unexpected.  I recall reading that at least one member of this forum has a tentative plan to bench test the 'clock chain' theory, this could be a very interesting enterprise indeed.  If anyone tries this I would like to wish them the very best of luck finding something new and interesting, which would certainly be far more interesting than simple debunks ignoring an increasing body of evidence.

[Lebouwsky]

8 hours ago, atxkyle said:

Thanks for the advice. I decided to get a dedicated Roon server (Intel NUC i3) and am now running it bare bones in Windows server 2012 r2 Essentials in minimal server mode with Audiophile Optimizer.  Also upgraded the microRendu to the 1.4 board with the upgraded clock.  These 2 changes definitely have taken my system to next level!

Are you saying that the server upgrade by itself (feeding a microrendu) was an upgrade in sound quality? Because if it does, it's one more argument that improving the packet data feeding the microrendu influences the sound. It's in conflict with the statement that the server can be as noisy as can be.

[fibonacci1618]

1 hour ago, Lebouwsky said:

Are you saying that the server upgrade by itself (feeding a microrendu) was an upgrade in sound quality? Because if it does, it's one more argument that improving the packet data feeding the microrendu influences the sound. It's in conflict with the statement that the server can be as noisy as can be.

 

Not sure if this is going to make sense, but I played with the RPi3 + Kali Reclocker + Piano DAC and set it up with Moode Audio and Volumio. I then set up Roon Bridge on the same hardware over the DietPi Linux distro, and fed it with the sonicTransporter i5 running Roon Core 1.3.

 

The Moode Audio and Volumio sounded very similar to each other, and for the money (<$200), I thought the RPi really sounded quite good. However, when I setup Roon Bridge and fed it through the Roon Core, it sounded much much better, in every way. I really didn't expect the exact same hardware (RPi + Kali Reclocker + Piano DAC) to sound so different and so much better.

 

So, I'm taking the dive to try out everything and hope to be able to take things further.

 

[Bruce Orr]

AustinPop - I'm late to the game here, but thanks for the sharing ur knowledge base and creating the summary/index back in January. Based on what uv said about upstream power supplies, I'm thinking that is an area of opportunity (among others) for me. Being of sound ears, and little mind, I was quickly confused about identifying the model and power needed. What is the best Paul Hynes "model" to power a Netgear Nighthawk router. It requires a maximum of 34 watts. Do models like the S3, 5 or 7 come with socket plugcord AND a connecting cord with appropriate male jack for the router?  Is a separate regulator of some kind needed?

[austinpop]

11 hours ago, fibonacci1618 said:

(i) will the LPS-1 (a single unit) sufficiently power the 200ultra (9V rated version), and if so -

 

As @Confused correctly points out, every configuration is different. Since your view is that your USB device downstream of the sMS-200ultra will not consume much current, then I think the empirical data points are in your favor. In fact, I don't believe anyone has yet reported a situation where an LPS-1 was insufficient.

 

11 hours ago, fibonacci1618 said:

(ii) which one will sound better with the 200ultra? The LPS-1 or the sps-500?

 

I know the word "better" is very vague, and so I hope someone can describe as best as possible, in what ways "better" means. 

 

 

Well, we have now had both @romaz and @Bamber report that, while the sPS-500 sounds very good for an SMPS, they ultimately preferred the SQ of the LPS-1 over the sPS-500.

[austinpop]

10 hours ago, Confused said:

There have been one or two recent posts here that have picked up on the question of how a modified switch with clock feed can improve sound quality in a system where Ethernet packet data is effectively buffered prior to being converted to a data stream.

 

 the modified switch has been one of the 'stars of the show', a simple thing that has offered a surprising result.  In fact, 'surprising' might be the key word here. 

 

So having said that it is easy to dismiss this stuff, I would say the more interesting challenge is to try to understand how this works and why this may be a very real working phenomenon.

 

I agree with your assessment about the impact of the reclocked switch. Whether or not that is surprising will no doubt be a subject of debate, since the concept has clearly been tried before. Perhaps it is the quality of the sCLK-EX that has enabled what we are hearing.

 

That said, I do want to ask that discussion of "how this works" and "why this works" be spawned off on a new thread. In fact, I seem to remember someone doing that early on, for precisely this topic. If you can find that thread, it would be the perfect place for this discussion.

 

This current thread is firmly rooted in empirical findings based on listening impressions. Clearly, it has proven helpful and beneficial to many. I would like to keep it that way. I want people to report what they hear, without being asked to prove it at every step.

 

I know this isn't what you intended, @Confused, so this isn't a knock on you at all. I just used your post to make a larger point. There is an unfortunate trend on CA at the moment of threads getting derailed by "deniers." As OP-designate, I have the right, according to Chris, to declare a topic to be for listening impressions only.

 

I so declare for this thread!

[austinpop]

37 minutes ago, Bruce Orr said:

AustinPop - I'm late to the game here, but thanks for the sharing ur knowledge base and creating the summary/index back in January. Based on what uv said about upstream power supplies, I'm thinking that is an area of opportunity (among others) for me. Being of sound ears, and little mind, I was quickly confused about identifying the model and power needed. What is the best Paul Hynes "model" to power a Netgear Nighthawk router. It requires a maximum of 34 watts. Do models like the S3, 5 or 7 come with socket plugcord AND a connecting cord with appropriate male jack for the router?  Is a separate regulator of some kind needed?

 

Unfortunately, I have no experience with Paul Hynes power supplies myself. I will leave it to others who do to respond.

[Johnseye]

9 minutes ago, austinpop said:

 

I so declare for this thread!

 

You mean you don't want 20 pages of double blind study debate?  Could this be a new trend towards civil discussion?  Kudos for keeping order.

[austinpop]

5 minutes ago, Johnseye said:

 

You mean you don't want 20 pages of double blind study debate?

 

No thanks! 

[simonklp]

On Monday, August 28, 2017 at 0:19 AM, austinpop said:

Ah OK I see now. 

 

I don't use HQPlayer, other than trying it early on, and realizing that for my DAC (and I stress this, as it is very DAC-dependent), native resolutions fed to the DAC sounded best.

 

But perhaps others who have deployed sCLK-EX based Ultra chains are using HQPlayer, so they will be the ones to chime in.

@austinpopThank you again for your kind reply.

 

Agreed. Let's look forward to hearing others, who are using HQPlayer and have deployed sCLK-EX based Ultra chains or similar, to share their experiences.

 

After all, that's the purpose of this forum for computer audiophile to ask questions and/or share their experiences. Cheers.

[TopQuark]

6 hours ago, Bruce Orr said:

What is the best Paul Hynes "model" to power a Netgear Nighthawk router. It requires a maximum of 34 watts. Do models like the S3, 5 or 7 come with socket plugcord AND a connecting cord with appropriate male jack for the router?  Is a separate regulator of some kind needed?

 

The Nighthawk uses about 15 watts under load although it comes with 34 watt power supply.  The Hynes SR3 may be good enough at 25 watts max.  My SR3 can go up to 10 amps so you should be good.  Better check with Paul to be sure.

 

The unit can be bought with a 2.1 or 2.5 mm DC plug with annealed copper, fine silver, and fine silver with ultra low impedance connectors.  There is no need for another regulator.

 

Paul is coming out with SR4 that is switchable to various voltages that is very similar to the unit below but the updated model. That will be a better unit for flexibility.  Better yet, upgrade to SR5 with 2 rails or SR7 with up to 4 rails.

 

 

 

 

[beautiful music]

I don't know if someone got a chance to compare UltraRendu to sMS-200Ultra

[atxkyle]

23 hours ago, Lebouwsky said:

Are you saying that the server upgrade by itself (feeding a microrendu) was an upgrade in sound quality? Because if it does, it's one more argument that improving the packet data feeding the microrendu influences the sound. It's in conflict with the statement that the server can be as noisy as can be.

Yes, that's exactly what I'm saying, and it's not at all subtle.  The SQ when feeding mR from bare-bones AO optimized server is MUCH better than when I was just using my office PC.  ( given this, my next step at some point is going to be to feed the Intel NUC a better power supply )

[fibonacci1618]

15 hours ago, austinpop said:

Well, we have now had both @romaz and @Bamber report that, while the sPS-500 sounds very good for an SMPS, they ultimately preferred the SQ of the LPS-1 over the sPS-500.

 

Thanks very much @austinpop I managed to locate @romaz

 and @Bamber 's views on this. Looks like I will get the LPS-1.

[phusis]

17 hours ago, austinpop said:

 

As @Confused correctly points out, every configuration is different. Since your view is that your USB device downstream of the sMS-200ultra will not consume much current, then I think the empirical data points are in your favor. In fact, I don't believe anyone has yet reported a situation where an LPS-1 was insufficient.

 

 

 

Well, we have now had both @romaz and @Bamber report that, while the sPS-500 sounds very good for an SMPS, they ultimately preferred the SQ of the LPS-1 over the sPS-500.

 

Steve Plaskin over at audiostream.com favored the sPS-500 (with 9V output) over the LPS-1 in conjunction with the sMS-200ultra. And, with the sPS-500 the sMS-200ultra can also be run in 12V mode (optional unit) - by many the preferred voltage setting here. Bear in mind also that poster @romaz impressions of the sPS-500 were found with that PSU only having few hours under the hood. From experience I know the sPS-500 benefits from rather substantial burn-in, i.e. some 3-400 hours. Poster @Bamber found the LPS-1 "better," yes, but that was the only word used to come about his impressions; surely we need some elaborated views describing (i.e. an actual, descriptive approach than merely "better") the sonics of both fully run-in units before any assessments can be made.