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

[matthias]

50 minutes ago, guiltyboxswapper said:

TBH I was quite surprised to find I preferred without moat on my main system, without resorting to an extreme.  Ryzen 3900x w/industrial ram running HQP Server on one end, NAA oR on other.   

 

It reminded me of switching an Ethernet port between 100mb and 1gig - sure, less haze as you mention, but you do lose some transient definition in voices & natural instruments especially (which translates into less presence).

 

You must have a very good system.

 

 Similar the findings of  @romaz from another forum:

 

"Now for the caveat. When I first got the ER, like everyone else, my router was feeding the ER via the "A" side (1000Mbit) and the ER was feeding the Extreme via the much vaunted "B" side (100Mbit). To be honest, even with the ER powered by my SR7, I was a bit underwhelmed. Midrange was rendered beautifully and there was better smoothness but transients were soft, the attack was quite dull, and the sound stage felt considerably smaller. Ok for vocals recorded in a studio but not ok for large orchestral music. To be bluntly honest, I was hearing more good things coming from a Sablon Ethernet cable than I was from the ER and I was preferring going direct from my modem/router to the Extreme via the Sablon cable.

When Emile told me his ER was sounding better than my dual SOtM switches that he had in his possession, I was quite surprised but then he also told me he didn't care for the "B" side of the ER at all and much preferred using ONLY the "A" side. He was further connecting to the Extreme with fiber using these new SFP transceivers that he found to be much better than the Startech transceivers. I went ahead and switched to the "A" side and agree fully with Emile on this. At least with the Extreme, the "A" side, even when using purely copper (and not fiber) sounds better. It is a touch less warm and more coarse than the "B" side but transients are more fully and dynamically expressed. The problem that remained is the sound stage still sounded more confined compared against directly connecting the modem/router to the Extreme.  

Before fully giving up on the ER, I decided to give these new SFP transceivers that Emile liked a try and I believe I have found them. This is the link that Emile provided and neither of us were sure these were the actual transceivers he likes because these are quite a bit less expensive at only $30 each.  

https://planetechusa.com/product/mgb-tlx-mini-gbic-lx-module-20km/

I was skeptical that these inexpensive transceivers could rescue the ER but somehow, they do. They are indeed more dynamic. They also sound stage larger than the Startechs and yet, they don't sound mechanical which is how Emile described the Finisar 10G SFPs that I had also considered trying. With these transceivers, I now am preferring having the ER in the chain as opposed to going directly from modem/router to Extreme." 

 

Matt

[guiltyboxswapper]

@matthias I did see this posted, infact, I brought the planet SFPs as soon as Emile mentioned them.  

 

I suspected with the tiny chassis's the oR and oM came with that they too would benefit from lower EMI footprint SFPs.  They did btw, as mentioned certainly sound less mechanical.  I should have posted about that, but apologies writing isn't my forte.

 

One thing I've done that I don't see mentioned by others (or maybe they dont discuss it) is vibration control.  Main system is covered in Daiza's and Towshend seismic podiums for the KEF blades, all other ancillaries using combo of Pulsar Points + mini ISO-Pucks.   

 

Main listening area is bordering padded cell - prob should get professional assistance there at some point (!).

 

But yes, still surprised their findings match my own.

[matthias]

6 minutes ago, guiltyboxswapper said:

@matthias I did see this posted, infact, I brought the planet SFPs as soon as Emile mentioned them.  

 

I suspected with the tiny chassis's the oR and oM came with that they too would benefit from lower EMI footprint SFPs.  They did btw, as mentioned certainly sound less mechanical.  I should have posted about that, but apologies writing isn't my forte.

 

One thing I've done that I don't see mentioned by others (or maybe they dont discuss it) is vibration control.  Main system is covered in Daiza's and Towshend seismic podiums for the KEF blades, all other ancillaries using combo of Pulsar Points + mini ISO-Pucks.   

 

Main listening area is bordering padded cell - prob should get professional assistance there at some point (!).

 

But yes, still surprised their findings match my own.

 

Yes, I am very interested about your experiences with the Daizas.

If you think it is OT here, please send me PM.

Thank you

 

Matt

[austinpop]

I've been biting my tongue wrt all the EtherREGEN discussion, while my review was pending. It's finally ready and should be live within the next couple of days.

 

For what it's worth, I DO find the A<->B crossing to sound notably better. Not sure why the Extreme folk find differently. In general, this space is very system-dependent, so nothing is real, other than what you hear in your own system.

[guiltyboxswapper]

17 minutes ago, austinpop said:

For what it's worth, I DO find the A<->B crossing to sound notably better. Not sure why the Extreme folk find differently. In general, this space is very system-dependent, so nothing is real, other than what you hear in your own system.

 

Out of interest, any fibre in your system's network?

 

Agreed - use as works best for yourself, that's all that matters.

[shahed99]

Connecting Audiophile switches like SoTM or ER with the router - is everyone using SFP port lately as oppose to Ethernet cable?

[mourip]

11 hours ago, shahed99 said:

Connecting Audiophile switches like SoTM or ER with the router - is everyone using SFP port lately as oppose to Ethernet cable?

 

I am using ethernet from my LAN switch to my ER. In the past I have found ethernet sounded better in my system than when I used FMC's to connect my switch directly to my server or from my server to my endpoint.

 

My ER has about 100 hours of music on it and has now moved from being a worthwhile addition to being remarkable.

 

A few of you may remember old analog CRT based TVs with a gross and a fine tuning nob. Adding the ER is like finally getting the tuning just right. All of the sudden you are hearing what was really there all along. I am no longer listening to the system, just engaged in the music. That is about the highest praise I can offer.

[Gavin1977]

Does anyone else have experience they can share in comparing the Sonore opticalRendu with other streamers?  I can find reviews, but not comparisions.

 

I would imagine that the eRED-dock would be similar in performance (but it's limited to uPNP).  I'm interested in filling in the percieved gap with the products other preople have found that lie between the opticalRendu and the ultra highend Pink Faun and Taiko Audio Extreme.  Both of which are way out of my price range 🙂

 

 

[zettelsm]

On 3/7/2020 at 11:43 PM, matthias said:

 

Thanks for sharing.

 

IMO, the most fair comparison between ER as FMC and OM should include:

 

1.) Using the ER as FMC only not as switch (one copper input only).

2.) Using the A-side only of ER.

3.) Using for ER and OM the same HQ PS.

4.) The downstream component must have fiber input like OR or Taiko Extreme.

 

AFAIK from his posts on WBF even @romaz has not yet done this comparison.

 

Thanks

 

Matt

 

 

Hi Matt,

 

I hope to be able to do some ER and OM testing with my Extreme starting later today. I plan to satisfy all the conditions you've postulated.

 

My current set-up is DSL modem > Sablon ethernet jumper > Startech FMC/Planet Tech SFP (-10db attenuator on receiver port ) > fiber > Planet Tech SFP (-10db attenuator on receiver port)/Taiko Extreme

 

I am powering my modem and the Startech FMC from different rails of an Uptone Audio JS-2 linear power supply.

 

I will also be trying the different permutations with and without the fiber attenuators. If time permits I'll attempt to also try the various Startech FMC, ER and OM configurations with the original Startech SFPs Emile recommends as well. I have had a failure of one Startech SFP already, so I'm a little leery of damaging another, though at this point the failure mode of the SFP isn't known.

 

I'll post my results here as well as on WBF.

 

Steve Z

[matthias]

2 hours ago, Gavin1977 said:

Does anyone else have experience they can share in comparing the Sonore opticalRendu with other streamers?  I can find reviews, but not comparisions.

 

I would imagine that the eRED-dock would be similar in performance (but it's limited to uPNP).  I'm interested in filling in the percieved gap with the products other preople have found that lie between the opticalRendu and the ultra highend Pink Faun and Taiko Audio Extreme.  Both of which are way out of my price range 🙂

 

The problem is that PF and Taiko are server/streamer where the OR and the eRED-dock are endpoints only.

For both endpoints you would need an additional server anyway.

 

Matt

[guiltyboxswapper]

3 hours ago, Gavin1977 said:

Does anyone else have experience they can share in comparing the Sonore opticalRendu with other streamers?  I can find reviews, but not comparisions.

 

 

 

You'd have to compare their Signature Optical range, ideally with tier ii (+ turbo PSU) - which uses the same opticalRendu PCB.  I tried to get ahold of one in the UK, but its a real struggle hence went for the oR and will recase and apply a heatsink to it's SoC.  Lots of reviews, though.

 

Edit: opticalRendu PCB is the same in all Signature SE Optical ranges, its the modifications (fuse, EMI shielding and Turbo PSU) that differentiates it.  These days they simply have 1 offering (thankfully).

[Gavin1977]

14 hours ago, guiltyboxswapper said:

 

You'd have to compare their Signature Optical range, ideally with tier ii (+ turbo PSU) - which uses the same opticalRendu PCB.  I tried to get ahold of one in the UK, but its a real struggle hence went for the oR and will recase and apply a heatsink to it's SoC.  Lots of reviews, though.

 

Edit: opticalRendu PCB is the same in all Signature SE Optical ranges, its the modifications (fuse, EMI shielding and Turbo PSU) that differentiates it.  These days they simply have 1 offering (thankfully).

Thanks - nobody holds stock in the UK unfortunately as they're all custom made to order.  So no means for me to compare.

 

Agreed that PF and Taiko are server/streamer, and I'm looking at endpoints - whilst an endpoint makes most sense for me (as I almost excusively use Tidal) I'm just following the eternal quest of 'what sounds best' for my given needs, budget  and learning on the way 🙂 

[austinpop]

On 3/8/2020 at 11:39 AM, austinpop said:

I've been biting my tongue wrt all the EtherREGEN discussion, while my review was pending. It's finally ready and should be live within the next couple of days.

 

The review is up on the main page.

[shahed99]

On 3/8/2020 at 10:32 AM, guiltyboxswapper said:

I did see this posted, infact, I brought the planet SFPs as soon as Emile mentioned them.

@guiltyboxswapper @matthias I've a favor to ask. Can you guys please share what fiber cable are you using between the planet SFP mentioned here? Do I need to use a attenuator given they are for long distance?

 

I've a Mikrotik router and SoTM switch and planning to use the SFP between them. Thank you!

[bodiebill]

On 3/1/2020 at 2:58 PM, bodiebill said:

Following some posts on an ISO Regen topic here:

I just made this adapter that allows me to split usb power and signal. The thing on the left is a Chinese 5V/2A LT3045 PSU powered by batteries.

 

I made this in preparation for the Denafrips Terminator that I expect to receive in a few days. However, I already tried it on my Lampizator 'Lite/Big 7'. As the Lampi does not draw any current I just removed the little PSU, to bypass the power, and already I heard a slight step up in SQ.

 

 

The Terminator arrived last week and I am now listening with:

    audio endpoint => Lush => iFi micro iUSB3.0 => Lush^2 (data) => Terminator

where the 5VBUS power for the Terminator USB is supplied separately by a battery powered LT3045 LPSU. See pic.

I need more time before I can comment on SQ.

[guiltyboxswapper]

4 hours ago, shahed99 said:

@guiltyboxswapper @matthias I've a favor to ask. Can you guys please share what fiber cable are you using between the planet SFP mentioned here? Do I need to use a attenuator given they are for long distance?

 

I've a Mikrotik router and SoTM switch and planning to use the SFP between them. Thank you!

 

I did use attenuators for the 80km SFPs, but for the 20km Planet variants I've not had to / tried with attenuators. 

 

Attenuators are very cheap if you want to try it yourself. 

 

https://www.amazon.co.uk/gp/product/B01EN33JZM/

[StreamFidelity]

19 hours ago, bodiebill said:

where the 5VBUS power for the Terminator USB is supplied separately by a battery powered LT3045 LPSU.

 

The 5V is only used as a trigger by the Terminator. The real USB power comes from the linear power supply built into the terminator. I think it`s more important to supply the USB controller from endpoint with clean power.

[bodiebill]

29 minutes ago, StreamFidelity said:

The 5V is only used as a trigger by the Terminator. The real USB power comes from the linear power supply built into the terminator. I think it`s more important to supply the USB controller from endpoint with clean power.

 

Ah, so the the Terminator 5V trigger would not require a top notch LPSU?

Actually, my ISO Regen gets very clean power via 9V Studer900 LPSU => ldovr LT3045 7.1V.

[StreamFidelity]

3 minutes ago, bodiebill said:

Ah, so the the Terminator 5V trigger would not require a top notch LPSU?

 

Yes

 

3 minutes ago, bodiebill said:

Actually, my ISO Regen gets very clean power via 9V Studer900 LPSU => ldovr LT3045 7.1V.

 

This shoud be enough! 👍

[seeteeyou]

https://www.whatsbestforum.com/threads/taiko-audio-sgm-extreme-the-crème-de-la-crème.27433/page-143#post-629149

Quote

What I am finding with the Extreme is that it is so fast that anything that diminishes this speed diminishes the Extreme's ability to resolve. This is one of the issues I have with the EtherREGEN's "B" side is that it sounds slower.

 

Since @romaz talked about how fast those Xeon Silver 4114 processors inside SGM Extreme were, it's only natural to find out what's even faster out there

 

HPE, AMD sign $600M contract to build the world's fastest supercomputer for U.S. nuclear tests

https://news.yahoo.com/m/6769704c-70ce-38fe-9b0f-5dfb576ad6fe/hpe%2C-amd-sign-%24600m-contract.html

 

AMD Adds EPYC 7662 64 Core & EPYC 7532 32 Core CPUs To Rome Server Family – US Navy Readies 12.8 PFLOP Supercomputer With EPYC & Volta

https://wccftech.com/amd-epyc-rome-cpus-us-navy-supercomputer-google-cloud-new-skus/

 

AMD Next-Gen Zen 3 Based Ryzen 4000 Desktop & EPYC Milan Server CPUs Arriving Late 2020

https://wccftech.com/amd-zen-3-ryzen-4000-desktop-and-epyc-milan-server-cpus-late-2020-launch/

 

---

 

https://www.head-fi.org/threads/review-comparison-of-5-high-end-digital-music-servers-aurender-n10-cad-cat-server-totaldac-d1-server-auralic-aries-audiophile-vortex-box.787020/page-92#post-15497469

Quote

Emile does have a lower cost server in the works with a single CPU and half the RAM that should probably still outperform everything else out there but according to Emile, the 2nd CPU is adding quite a bit to the performance of the Extreme and so it will be a matter of "you get what you pay for."

 

AMD EPYC 7002 Series Rome Delivers a Knockout

https://www.servethehome.com/amd-epyc-7002-series-rome-delivers-a-knockout/

Quote

AMD has roughly twice the performance per socket in mainstream applications as Intel has throughout the range up to Xeon Platinum.

 

AMD EPYC 7002 Series Coverage at STH Road to Rome Hub

https://www.servethehome.com/amd-epyc-7002-series-coverage-sth-road-to-rome-hub/

Quote

In this generation, AMD has five single-socket only parts. These “P” series parts are special because they leverage the platform’s unique I/O capabilities to deliver industry-leading single-socket performance equivalent to Intel’s dual-socket Xeon performance.

 

---

 

Question for @Nenon since you're planning to build something that's somewhat similar to SGM Extreme. IIRC you're having better luck with Asus motherboards than other choices, did you take a look at their KRPA-U16 for AMD EPYC Rome 7002 Series by any chance? Ryzen / EPYC should be good for your Pink Faun I2S Bridge while having 16 DIMM slots could support 8 memory channels instead of 6 with Xeon Silver 4114

 

https://dlcdnets.asus.com/pub/ASUS/mb/SocketSP3(4094)/KRPA-U16/Manual/E16152_KRPA-U16_UM_V2_WEB.pdf#page=100

 

ASUS Announces New World Records on AMD EPYC 7002 Series 1P Server on SPEC.org

https://www.asus.com/News/tscigh2afvlvwae0

 

AMD EPYC™ 7002 LGA 4094 EEB server motherboard with DDR4 3200 MHz, PCIe 4.0, M.2, NVMe, SAS, SATA, OCP 2.0 Mezzanine card, dual LAN and ASUS Control Center

https://www.asus.com/Commercial-Servers-Workstations/KRPA-U16/

 

As always there's a catch with these powerful processors, real servers weren't designed for passive cooling in the first place while a normal server chassis should have case fans installed by default. In other words, these products from Dynatron and Supermicro wouldn't be such a good idea for a completely fanless setup

 

https://www.amazon.com/Dynatron-A25-Passive-Copper-Skived/dp/B076W5P8CF

https://www.amazon.com/Dynatron-A28-Heatsink-Chamber-Stacked/dp/B079YJ6634

https://www.amazon.com/Supermicro-SNK-P0062P-EPYC-7000-Brown/dp/B07F2M2CGC

https://www.amazon.com/Supermicro-Heatsink-SNK-P0063P-Socket-Brown/dp/B078HH24Y7

http://developer.amd.com/wordpress/media/2013/12/ThermalGuideforSP3.pdf#page=12

 

In theory we're getting 16 cores / 32 threads from a single EPYC 7282 while passive cooling could be somewhat tricky

 

https://en.wikichip.org/wiki/amd/epyc#7002_Series_.28Zen_2.29

https://www.servethehome.com/wp-content/uploads/2019/08/AMD-EPYC-7002-SKU-List-and-Value-Comparison-Full.jpg

 

 

---

 

For the actual power consumption of an entire system with 12 RAM sticks plus dual Xeon 4114 @ 85W TDP, it's hard to tell how that would be compared to a single EPYC 7282 @ 120W TDP

 

https://www.whatsbestforum.com/threads/taiko-audio-sgm-extreme-the-crème-de-la-crème.27433/page-97#post-615725

Quote

While the Extreme consumes only about 60 watts during music playback, Emile tells me that the Extreme can consume up to 300 watts during bootup. This would fry any of the bespoke PSUs I have on hand.

 

https://www.head-fi.org/threads/review-comparison-of-5-high-end-digital-music-servers-aurender-n10-cad-cat-server-totaldac-d1-server-auralic-aries-audiophile-vortex-box.787020/page-92#post-15497469

Quote

The Extreme consumes only 60 watts during playback and while idle.

 

---

 

Now how do we keep the temperature low while having a lower load for the sake of better SQ?

 

https://www.whatsbestforum.com/threads/taiko-audio-sgm-extreme-the-crème-de-la-crème.27433/page-100#post-616762

https://audiophilestyle.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/page/616/?tab=comments#comment-1014096

On 12/16/2019 at 7:37 AM, romaz said:

But you do need a power supply which is very comfortable supplying the load. You really want the least possible variation in load, and higher cpu power / bandwidth systems are better at that with very low load music playback processes."

 

https://service.chelsio.com/beta/drivers/ChelsioUwire-3.9.1.0/Chelsio-GPUDirectRDMA-Linux-UserGuide.pdf#page=6

Quote

NVIDIA Tesla K series adapter should be installed in the machine (required by NVIDIA installer).

 

NVIDIA TESLA KEPLER K10 GPU ACCELERATOR

https://www.ebay.com/itm/NVIDIA-TESLA-KEPLER-K10-GPU-ACCELERATOR/283637156734

 

NVIDIA Tesla K20 GPU 5GB GDDR5 PCIe 2.0 x16 705MHz 900-22081-6310-001 Graphic

https://www.ebay.com/itm/NVIDIA-Tesla-K20-GPU-5GB-GDDR5-PCIe-2-0-x16-705MHz-900-22081-6310-001-Graphic/372935506664

 

NVIDIA Tesla K20X 6GB GDDR5 PCI-E 3.0 x16 Server GPU Accelerator Card CUDA

https://www.ebay.com/itm/NVIDIA-Tesla-K20X-6GB-GDDR5-PCI-E-3-0-x16-Server-GPU-Accelerator-Card-CUDA/352755592029

 

A fast GPU memory copy library based on NVIDIA GPUDirect RDMA technology

https://github.com/NVIDIA/gdrcopy

Quote

GPUDirect RDMA requires NVIDIA Tesla or Quadro class GPUs based on Kepler, Pascal, Volta, or Turing, see GPUDirect RDMA. For more technical informations, please refer to the official GPUDirect RDMA design document.

 

Frequently Asked Questions : GPUDirect RDMA

https://community.mellanox.com/s/article/frequently-asked-questions---gpudirect-rdma

Quote

This provides a significant decrease in GPU-GPU communication latency and completely offloads the CPU, removing it from all GPU-GPU communications across the network.

 

GPUDirect Storage: A Direct Path Between Storage and GPU Memory

https://devblogs.nvidia.com/gpudirect-storage/

Quote

Both GPUDirect RDMA and GPUDirect Storage avoid extra copies through a bounce buffer in the CPU’s memory and enable a direct memory access (DMA) engine near the NIC or storage to move data on a direct path into or out of GPU memory – all without burdening the CPU or GPU.

 

---

 

Basically not that many applications (e.g. CyberLink PowerDVD and Signalyst HQPlayer etc.) support CUDA offloading but some audiophiles in Taiwan were using NVIDIA Nsight HUD Launcher to take advantage of their Tesla cards with just about anything

 

http://www.myav.com.tw/bbs/showthread.php?threadid=20475707

http://www.myav.com.tw/bbs/showthread.php?threadid=20480094

[Nenon]

6 minutes ago, seeteeyou said:

Question for @Nenon since you're planning to build something that's somewhat similar to SGM Extreme. IIRC you're having better luck with Asus motherboards than other choices, did you take a look at their KRPA-U16 for AMD EPYC Rome 7002 Series by any chance? Ryzen / EPYC should be good for your Pink Faun I2S Bridge while having 16 DIMM slots could support 8 memory channels instead of 6 with Xeon Silver 4114

 

No, I did not. And the main reason for that is that processing power is not the only factor. 

An Intel i9-9900K in the ASRock miniITX gaming motherboard sounds different than the same CPU in the ASUS ROG miniITX motherboard, which sounds different than the AMD Ryzen 7 3700X in a ASUS ROG miniTX motherboard, which sound different that the same CPU in the AMD ASUS ROG ATX motherboard. There are almost infinite combinations one can try. Okay, perhaps not an infinite, but large enough number of combinations to spend years of R&D. 

I wanted to start from somewhere, and felt like the motherboard Emile picked for the Taiko Extreme motherboard is a good start. He said he designed his own custom motherboard that did not sound better than this one. Sounds like a good bet to me.  I would probably go with the Xeon Silver 4214 instead of the Xeon Silver 4114 - more cores and more cache. 

 

As far as the I2S bridge with ultraOCXO clock - I am actually planning to use USB with this future build and have been working on modifying the USB interface on my DAC to improve the USB. 

 

I have said that before, but the interface between the music streamer and DAC is one of the most important places that needs special attention. I am a purist and don't necessarily like the idea of generating USB signal from the computer to convert it to I2S inside the DAC. Every conversion like that has the potential to impact the sound quality negatively. Hence, I have been using I2S.

I2S has a lot of problems too and needs an extreme level of attention to details to sound good (that's a very long topic). But having control over my server and my DAC I managed to optimize both to work together really nicely. And as of right now, I2S sounds better than USB in my system. But my DAC is better optimized for I2S input... that's why it sounds better.

I have learned a lot of new things lately - some stuff from the Cerat Kassandra II Ref DAC, others from power supply design and computer design. And I have a very optimistic plan to try to make USB to sound better. Time will show. 

[dminches]

Have you picked a case that will fit the ASUS sage mobo?

 

[Johnseye]

2 hours ago, dminches said:

Have you picked a case that will fit the ASUS sage mobo?

 

 

I believe it will fit in the HDPlex H5 but @Nenon can say for sure.

[Nenon]

2 hours ago, dminches said:

Have you picked a case that will fit the ASUS sage mobo?

 

 

5 minutes ago, Johnseye said:

 

I believe it will fit in the HDPlex H5 but @Nenon can say for sure.

 

Yes, that's correct. I am pretty sure it would fit, although I haven't tried it yet. 

[Dev]

2 hours ago, Nenon said:

 

 

Yes, that's correct. I am pretty sure it would fit, although I haven't tried it yet. 

 

WS C621E Sage is EEB Form Factor - 12 inch x 13 inch

 

For comparison:

ATX Form Factor - 12.0 inch x 9.29 inch

Extended ATX Form Factor - 12 inch x 10.9 inch

 

Also Sage is a dual socket mobo. If you populate both, you need to figure out the cooling and routing heat pipes on the H5 case.