Jump to content
IGNORED

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


Message added by The Computer Audiophile

Important and useful information about this thread

Posting guidelines

History and index of useful posts

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

Recommended Posts

4 minutes ago, austinpop said:

 

John,

 

See my last point in the previous post. I don't own the TLS LPSU. I've always powered it with an LPS-1.2.

 

I don't see the TLS switch in your schematic.  Are you still using it?

 

 

On 2/12/2019 at 6:06 PM, austinpop said:


Yes.

 

 

 

Not belief - empirical observation that it sounds better. This started way back when I first got my sCLK-EX modded switch, and is not represented by the TLS OCXO switch. While I do power this switch with an LPS-1.2, and have JSGT shunting, what seems to really be the SQ lift is the reclocking from the OCXO. In that sense, I consider this switch as an Ethernet regenerator.

 

 

 

Yes. My thinking is as follows:

  • putting a JS-blessed Netgear switch, powered by a basic LPS (I use an el cheapo Breeze) with JSGT shunting, upstream of the server, between it and the router, filters out most of the grunge from the rest of the house.

So, in my experience, it's a quartetta of mechanisms working in concert:

  1. isolation (through magnetics and shunting) in Netgear and TLS switches
  2. bridging - whatever it is that causes this to sound good
  3. reclocking/regeneration - in TLS switch
  4. filtering - in the SOtM dCBL-Cat7 cable from TLS to the NUC endpoint.

Again, this is all from trial and error. Some day this will be better understood. Until then, this is where my experiments have led me.

 

Boy, does it sound good! :) 

 

 

IIRC you were using the TLS switch before the server and SOtM switch before the NUC.

 

A lot has been discussed recently and a lot has changed.

 

Link to comment
8 hours ago, Nenon said:

 

Ordered one to try. Does the quality of the DC power matter on this thing? I wonder if I would hear difference between a cheap chinese LPS and my LPS1.2. I would try and report back if people have not tried yet. Thank you!

Yes, and also consider that using a LPS is not automatically better than a switching supply, which arent always as bad as you think.

Link to comment
53 minutes ago, elan120 said:

In my case, power supply quality does make a difference, but I only compare stock wall wart with what I currently have on hand, which I end up with JS-2 on the enclosure, and LPS-1.2 on the PCIe card.

 

Have you tried different USB cables?  I wonder if they make an audible difference.

 

 

 

Link to comment
6 minutes ago, numlog said:

Yes, and also consider that using a LPS is not automatically better than a switching supply, which arent always as bad as you think. 

 

Thank you. I am not planning to use any SPMS in my system. Even if they are not bad, I don't want to face the chance of contaminating my power line (dedicated circuit) more than it already is. What I was wondering was if people could hear a difference when you power the USB enclosure with let's say a LPS like the HDPlex vs. let's say a top of the line Paul Hynes design. But that's a test that's easy to do myself.

Industry disclosure:
https://chicagohifi.com

Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs

Link to comment
50 minutes ago, austinpop said:

When we do, I can add the experiment to try it without the Ref-10.

 

Thanks Rajiv,

 

Would be greatly appreciated.

 

Might sound daft but adding the TX and one PSU back onto the rack is just about palatable, but adding them a master clock and yet another PSU just doesn't appeal. Too much spaghetti at a point in life where I'm trying to cut down on carbs and clutter! 😉

 

That said I own the sclk-ocx10 still (earmarked for sale soon) so I guess all options are still open.

 

Cheers,

Alan

Synergistic Research Powercell UEF SE > Sonore OpticalModule (LPS-1.2 & DXP-1A5DSC) > EtherRegen (SR4T & DXP-1A5DSC) > (Sablon 2020 LAN) Innuos PhoenixNet > Muon Streaming System > Grimm Audio MU1 server > (Sablon AES) Mola Mola Tambaqui DAC > PS Audio M1200 monoblocks > Salk Sound Supercharged Songtowers

Link to comment
7 hours ago, Nenon said:

 

Thank you. I am not planning to use any SPMS in my system. Even if they are not bad, I don't want to face the chance of contaminating my power line (dedicated circuit) more than it already is. What I was wondering was if people could hear a difference when you power the USB enclosure with let's say a LPS like the HDPlex vs. let's say a top of the line Paul Hynes design. But that's a test that's easy to do myself.

yes, understandable. I meant more if you were already using ATX PSU but being a NUC discussion that was already less likely

Link to comment
2 hours ago, ted_b said:

Yes, we use a startech USB 3.1 card and a Startech USB 3.1 hdd enclosure, separately powered.  The JCAT card is great too. 

https://www.amazon.com/gp/product/B00XLAZEFC/ref=oh_aui_search_asin_title?ie=UTF8&psc=1

 

Although I already ordered one and will try it soon, I am still wondering why connect the hard drive to a USB port rather than using the SATA controller on the motherboard? I kind of want to keep my USB as clean as possible for my digital audio signal, and connecting an external USB drive kind of worries me.

Okay, we don't want vibrating parts in our music server chassis, but theoretically would not it be better to connect the HDD to the SATA controller and keep it external (plus external LPS)? Or does the SATA controller tend to generate more noise?

I am either asking questions about an unexplored area or there is research already done that I have missed (sorry if that's the case). Or perhaps it is computer architecture/system dependent and no one knows the answer.

 

Also, I really like the indexing in the original post. But with currently 545 pages it is difficult to extract so much of the useful information here. Have you considered spinning a few sub-threads? Storage could be its own section as well as each of the topics in the original post index.

Industry disclosure:
https://chicagohifi.com

Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs

Link to comment
7 minutes ago, Nenon said:

 

Although I already ordered one and will try it soon, I am still wondering why connect the hard drive to a USB port rather than using the SATA controller on the motherboard? I kind of want to keep my USB as clean as possible for my digital audio signal, and connecting an external USB drive kind of worries me.

Okay, we don't want vibrating parts in our music server chassis, but theoretically would not it be better to connect the HDD to the SATA controller and keep it external (plus external LPS)? Or does the SATA controller tend to generate more noise?

I am either asking questions about an unexplored area or there is research already done that I have missed (sorry if that's the case). Or perhaps it is computer architecture/system dependent and no one knows the answer.

 

Also, I really like the indexing in the original post. But with currently 545 pages it is difficult to extract so much of the useful information here. Have you considered spinning a few sub-threads? Storage could be its own section as well as each of the topics in the original post index.

The Startech USB3.1 Gen2 card has the option to connect an external power supply, and in my setup, I use LPS-1.2 with very good result.

 

Link to comment

The USB adapter still uses a similar type of controller chip, as mentioned its possible to use an external supply with it easily but I would think the internal SATA has a more direct data path.

 If internal SATA supports a lot more inputs than what is needed disabling may be better but I would agree about keeping USB clean... both have potential pros and cons and even if 2 people heard the same thing they might not agree on which sounded best, the only way to really know is to try it out yourself .

8 hours ago, Nenon said:

 

Although I already ordered one and will try it soon, I am still wondering why connect the hard drive to a USB port rather than using the SATA controller on the motherboard? I kind of want to keep my USB as clean as possible for my digital audio signal, and connecting an external USB drive kind of worries me.

Okay, we don't want vibrating parts in our music server chassis, but theoretically would not it be better to connect the HDD to the SATA controller and keep it external (plus external LPS)? Or does the SATA controller tend to generate more noise?

I am either asking questions about an unexplored area or there is research already done that I have missed (sorry if that's the case). Or perhaps it is computer architecture/system dependent and no one knows the answer.

 

Also, I really like the indexing in the original post. But with currently 545 pages it is difficult to extract so much of the useful information here. Have you considered spinning a few sub-threads? Storage could be its own section as well as each of the topics in the original post index.

 

Link to comment
17 minutes ago, lmitche said:

Here we go!

 

SATA Saga returns! 😁

 

Please start a new thread for this.

 

Yep. That sounds good. No more SATA drama from me here :).

 

I feel like it's worth to take the index in the original post and start a new topic for each section and provide a link. So much good information here but all looped into a giant and unmanageable thread (besides the indexed posts). It's probably easy for the people who followed it on regular basis in the past two years, but good luck to new members/readers. But I already said that, didn't I?

Industry disclosure:
https://chicagohifi.com

Dealer for: Taiko Audio, Conrad Johnson, Audio Mirror, and Sean Jacobs

Link to comment
20 minutes ago, Nenon said:

It's probably easy for the people who followed it on regular basis in the past two years, but good luck to new members/readers. 

 

Actually it is no easier for those of us who have been reading it from day one. We are also now getting lots of posts asking questions that have already been covered and folks are posting "new" ideas that have already been put forth.

 

The nice thing about this thread is that it is moderated but it is definitely getting long...


"Don't Believe Everything You Think"

System

Link to comment
26 minutes ago, Nenon said:

 

Yep. That sounds good. No more SATA drama from me here :).

 

I feel like it's worth to take the index in the original post and start a new topic for each section and provide a link. So much good information here but all looped into a giant and unmanageable thread (besides the indexed posts). It's probably easy for the people who followed it on regular basis in the past two years, but good luck to new members/readers. But I already said that, didn't I?

 

Nothing is stopping people breaking off and starting new threads to offload from this thread. In fact, folks like @rickca, @lmitche, and @bobfa have done exactly that for some key topics.

Link to comment
3 hours ago, rickca said:

Am I just having a mental block today?

 

Not sure Rick, but happy to reiterate. The key ideas were:

  • Think of the SOtM switch between server and the endpoint as an Ethernet Regenerator.
  • The 2 ports on the JCAT card on the server also do Ethernet regeneration and bridging. 

To your questions:

Quote

 

@austinpop so internet traffic (like Qobuz streaming) goes to your server without first passing through the switch, right?

 

Yes.

 

Quote

Would it not be beneficial to have something like an EtherREGEN in front of your server?

 

Perhaps. TBD.

 

Quote

What do I lose by connecting the switch to the router and the server to the switch? 

 

Empirically, bridging on the server sounds better. Still unclear why. So the shown topology locks in the benefit of bridging AND the reclocking on the switch between the server and the endpoint.

 

Quote

The original bridging idea was to avoid connecting the endpoint to the router and instead connect the server and endpoint directly.  So this seems to be a different kind of bridge?

 

 

Nope. Same old same old.

 

Link to comment

I tried 4 network setups. 

 

1. Windows 10 PC -> Bridged using (ASUS XG-C100C PCIe 10G Network Adapter) -> Pro iDSD -> Soundaware P1 -> Susvara

2. Windows 10 PC -> Bridged using cheap USB 2.0 Insignia  USB to Ethernet Adapter -> Pro iDSD -> Soundaware P1 -> Susvara

3. Pro iDSD is connected to the network via a network switch. 

4. Ultrarendu (USB) -> Pro iDSD

 

Everything is played through Jriver. The first 3 options sound indistinguishable to my ears and they all sound more dynamic than Ultrarendu. The Ultrarendu was fed by Uptone JS2.

 

Here are the things I find interesting:

1. I expected the option 1 to sound the best because ASUS 10G is an excellent, modern NIC so it should have the best clock.

2. I would expect option 2 to be the worse than options 1 and 3 because one it's connected to a USB port which is noisy and it's a $15 adapter. 

3. I totally didnt' expect Ultrarendu to be the worst sounding.  It sounds less dynamic and slightly anemic compared to options 1-3. BTW I also tried Roon Endpoint, Roon -> HQP and using HQP NAA. The results were the same.

Alienware R7 with Paul Pang V2 USB PCIE -> iFi Pro iDSD -> McIntosh MHA100 -> Hifiman Susvara. Keeping it simple!

Link to comment
13 hours ago, BigAlMc said:

I'd also love if yourself and/or @romaz could comment or compare what adds the most of an improvement: The EVOX modded TX-USBUltra or the sNH-10? If you had to choose only one of these which would you go for?

 

Alan, prior to the EVOX cap upgrade, my tX-USBultra (when powered by an SR SR7 at 12V) was sitting unused.  With my NUC powered by my DR SR7 rail, the tX was bringing it down.  Before the EVOX cap upgrade, I would have easily told you the sNH-10G was the more impactful device but to be fair, my sNH-10G came to me with the EVOX cap.  The EVOX cap resulted in a much larger improvement than I was expecting on the tX and so it is back in the chain.  Since I only presently own a single DR SR7 rail, I had been using this on the NUC but surprisingly, I have found better SQ by shifting an SR SR7 rail to the NUC and the DR SR7 rail to the tX.  The tX really scales to the DR rail in the same way that the sNH-10G does.  

 

I would consider the tX-USBultra to have equivalent impact to the sNH-10G at this time but because the sNH-10G also very significantly improves the Tivo/Netflix/Vudu streams that go through my Oppo player, if I was forced to choose one device, I would choose the sNH-10G.  

Link to comment
19 hours ago, chrille said:

Could you possibly sum things up in a "Reader's Digest" version!!!!

 

A summing up of essentials only, where you only mention the exact parts needed  for a solution that would not break the bank but still provide a clear improvement in SQ for a person like me who so far, has not done anything more than exchange the PSU of his Qutest for a battery powered linear one but  who is still running his M Scaler via the supplied PSU.

I am  still out travelling in Asia over the winter as usual, and while "on the road" I am still using my old  2011 i7 17" mbp with my music on  portable harddrives connected via firewire and Audirvana+ as music player playing my music from ram connecting my M Scaler to my mbp via usb.

 

Hi Chris, a Reader's Digest version is not so easy to do.  If you're looking for good sound that is reasonably priced but also portable and if you're committed to staying with your 2011 i7 MBP, it may be best just to stay with A+ since that is a very good platform for the Mac. 

 

Consider powering your Qutest with an LPS-1.2, it's better than a battery.  The M-Scaler is different.  It appears that the FPGA in M-Scaler upsamples at a steady rate and consequently, it consumes current at a steady rate.  In other words, M-Scaler doesn't really benefit from a low impedance PSU, just a low noise PSU and so I have found this inexpensive 16V battery to almost compete toe to toe against a DR SR7:

 

https://www.amazon.com/gp/product/B015OAJFOC/ref=ppx_yo_dt_b_asin_title_o01_s00?ie=UTF8&psc=1

 

Just make sure you don't go beyond 16V on the M-Scaler.

 

Without breaking the bank, because these units are portable and easy to travel with, consider adding an Iso-Regen and LPS-1.2 to your laptop.  I sometimes travel with my Iso-Regen and LPS-1.2.  If you wish to go beyond this, just PM me.

Link to comment

Create an account or sign in to comment

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

Create an account

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

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...