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

[Downtheline]

Has anyone used this Startech fiber nic successfully with either audiolinux or gentooplayer in their build?

 

(PEX1000SFP2)

 

[Peter Avgeris]

On 6/16/2020 at 3:29 AM, Superdad said:

 

My eyes certainly popped out when I saw the size and complexity of your phono preamp.

 

 

Thats a LOT of wire and magnetics for just a PHONO PREAMP! 

Very impressive...

Hello!

Thaks for your comments.

The first picture is the phono preamplifier. MM only. For MC cartridges MC step up transformers are required.

The last picture is a power amplifier manufactured eight years ago. EML20A > EML300b mesh > United Electronics 311

All iron stuff is of course calculated, designed and manufactured in house.

[Peter Avgeris]

On 6/16/2020 at 10:03 AM, RickyV said:

 

OMG @Peter Avgeris this must be in the guinness book of records for the biggest phono stage ever. And the first picture is the preamp?? Or is it all phono amp? 

Hello from Athens.

The first pic is the Exotic 3-chassis MM Phono Preamplifier. You are correct.

I have already designed the next level of MC Step Up Transformer. It is one chassis with 4 pairs of SUTs, it is plugged on AC mains and is made out of 24 iron pieces for the stereo set.

[Peter Avgeris]

5 hours ago, John769 said:

 

 

Just wondering if the various net card comparisons you made also included s/w using Windows or Server OS?  In particular with the intel based cards versus the TP-Link, bearing in mind you preferred the latter.

Hello!

The comparisons were made on Euphony Linux. Regarding comparisons under Windows, unfortunately I have no news.

Under Euphony on my own system, all Intels I have tried (82574, i210, i350-T2, i350-T4, i219-V, X540) were clearly inferior with the results coming in less than 10 seconds.

I am now playing with this small adapter and have plans to replace the crystal oscillator with OCXO some time in the future.

[Peter Avgeris]

I have plugged the Buffalo switches that I have received two days ago.

After having them on fire for aprox. 4 hours I decided to make some listening tests.

The Control Point PC was connected on the NAS. In between them I placed the two switches in series. NAS routes the packages to the Control Point PC, so here I did not follow the 'less is more' approach.

No LPS, no special cables, no special equipment. A=nothing, B=2x BS-GS2008

I went A>B and B>A several times.

With the double switch there was always more clear sound and better definition but something was missing. Music had lost some of its thrill & excitement. OK, I thought some sort of burn in would be required.

After 3 hours I decided to try the polarity of the power supplies. Here in Europe, the Schuko power plugs have been designed in a wrong way and the phase/neutral conductors are not explicitly specified. All the good things that were missing are now back, on top of course of all the good aspects.

To sum it up: once you go from A to B, you can't go backwards from B to A.

I do not know if Melco switch does miracles but for the 140 GBP I paid for the pair, I ma more than simply satisfied.

As soon as BS-GS2016 is discontinued, I cannot see a reason not to go for the BS-GS2008.

Just my two cents!

 

[Exocer]

There is a new Startech ASM 3142 based PCI-E card with both USB A and USB C outputs. I wish I had waited a bit longer to get this instead

 

https://www.amazon.com/exec/obidos/ASIN/B0897XXRXT

[afrancois]

 

The single best mod for my system and I wasn't even intended as one. The effect this lamp has on my system is simply amazing. I think it act as some kind of a diffuser. Granted  that I did move the coffee table a bit as well, but from previous try outs this has never had a profound effect. The lamp is dead center between the speakers from my listening position. To give some perspective, the lamp is only 162 cm from the floor and 128 cm from the coffee table. What is most noticeable is that the imaging is even better now, as is the overall sound balance. Both of which I thought there was little margin for improvement without the use of RoomPerfect. Still don't use RoomPerfect yet because some major changes are planned in the living room part you don't see in the picture.

 

[Cooler]

34 minutes ago, afrancois said:

 

 

Oh, I still remember your setup! You have a big rack behind the wall with perfect cable management and I remember you used a lot of supra cables! I want to do similar once I buy a big house with a dedicated room for audio!

 

So if the lamp did so much changes, I would imagine the impact of audio (heavy) curtains on the closest window to the left speaker. It seems this window reflects sound directly to your sitting position.

[afrancois]

19 minutes ago, Cooler said:

 

Oh, I still remember your setup! You have a big rack behind the wall with perfect cable management and I remember you used a lot of supra cables! I want to do similar once I buy a big house with a dedicated room for audio!

 

So if the lamp did so much changes, I would imagine the impact of audio (heavy) curtains on the closest window to the left speaker. It seems this window reflects sound directly to your sitting position.

Hello, I've replaced all Supra cables with outer self-assembled  cables. For critical listening I lower the curtains.  You see the two portions, if you look carefully, white and black. Those are not heavy curtains granted. My guess is that the Focals are made with a what more lively room in mind. When I start damping all the magic quickly disappears.

[sunny_time_99]

Hi,

I would like to share my listening experience with the Buffalo BS-GS2016-EU with you. For me, the Buffalo 2016 has been in continuous operation for more than 21 days. I initially ran the Buffalo with the built-in power supply for a few days and then built it into my chain. Nice deep bass was the first sentence that I wrote down. After a while I converted the Buffalo to my power supply. The sound corresponds exactly to the SR 7 (Dual Voltage: Series No. 28). I had bought it a few days ago from Garvin (here from the forum) and can now say by comparing it that my power supplies, which I have developed for my own use for 6 years, cannot be distinguished acoustically from the SR7. In my entire chain, I now have a good 30 power supplies that correspond in terms of sound to an SR7. In the DAC  its e.g. 11 individual power supplies.

 

 

About my current setup of the downstream route:

Fritzbox Router (SR7 Equal)> Switch –1 (SR7 Equal)> Sonore OM modules (SR7 Equal)> Finisar or Planet SFP> LC Dual single> Finisar or Planet SFP> Sonore OM modules (SR7 Equal)>( Etherregen) > Switch 2> Endpoint: Mini PC (2x SR7 Equal –Clock and computer) or Intel NUC 7i7DNHE (1x SR7 Equal)

 

Software Euphony and Daphile

 

 

I have noticed differences between the Planet SFP with 20km and the Finisar FTRJ-8519-7N-2.5 that are hardly noticeable.

I built a D-Link DGS 108 with a reference clock, clock and 3.3V and 5V of the switch were equipped with the same power supply (SR7 Equal). This switch is my main comparison to the Buffalo. I now have position 1 and position 2, which I can fill in my chain with the Buffalo or the modified D-Link. After more than 14 days of playing time, the sound of Buffalo is still nervous. The bass is good - but the overall sound looks torn in direct comparison. The modified Switch is completely different. D-Link: everything plays conclusively and without any hardship. To solidify this impression, I also operated the Buffalo directly with the SR 7.

In my chain, I can only operate the Buffalo at position Switch 1, but not at Switch 2. When the buffalo sits right in front of the endpoint, resolution, dynamics, musical flair are missing. I had my setup and my observations checked by my craft buddy who has a comparable chain. He compared the Cisco SG200-8 (without p - for PoE) with my power supply unit at the Switch 2 position with the Buffalo. Both can use the same high quality 12V power supply.

My recommendation: the Buffalo is worth the money! It brings an improvement in the downstream. I would use it at a starting point (for example in „switch position 1“) - but not connected to the end point, because it is compared to similarly upgraded (modified) switches such as e.g. the Cisco SG200 does not match their tonal performance.

So there are some alternatives in the area of switches. Next I will provide the Buffalo with a Neutron Star reference clock.

Happy listening

 

 

Sunny

 

p.s. i hope me and Google translate make it able to understand. english is not my native language.

I don't write so much about power supplies to offer them here. You should only have an idea at which level I am listening to Music.

[paulhynes]

Hi Sunny,

 

I am afraid you are many years out of date with Gavin’s old dual rail custom built SR7 serial number 28 (18v 2A and 9v 2A). Gavin purchased this second hand from one of my old customers. If you are comparing this old SR7 to your own power supplies, and finding them equivalent in sound quality, you still have much work to do on your own power supplies.

 

Much has changed since this SR7 serial no 28 was made with the early PR3 regulator boards. The latest boards used in the SR7 (and the SR5) have several circuit improvements, some component changes to improve the audible performance plus increased energy storage capacitance and the new Turbo upgrade (substantial) as well as electronic shutdown in the event of overload or excessive temperature. Not to mention the new SR7 chassis configuration with improved thermal capacity allowing the use of two 12v 16A supply rails.

 

Regards

Paul

[sunny_time_99]

 

Hello Paul,

thank you for your message and assessment. In essence, I wasn't interested in a power supply

comparison. The Buffalo was the topic. I am sorry to be off topic again.

Such a comparison would be a little briefly described as "comparable" or "equal". The new power

supplies that I am building today are also better than the ones that I am building built e.g. 4 years

ago. They are also significantly more powerful. A lot has happened and a lot will happen.

 

Since your SR 7 is known as a reference and is rightly regarded and recognized as a high quality power supply - I also own it and therefore do not tell about something that does not exist or that is only known from hearsay. Providing my power supply as a reference for the observation is of little use.

 

With the reference to your power supply (even if older date) others have a benchmark.

So that you can exclude the power supply as the cause of what I noticed about the Buffalo.

With the different power supplies that I tried on the Buffalo, the same observations always occurred

that made the power supply seem unlikely as a possible cause. After the the next Step in Buffalo tuning - the clock-mod. i can tell if that was the thing that makes ist sound like it did in my chain.

 

Happy listening

 

Sunny

 

[rickca]

Can I achieve the same sound quality with a 65W i7-10700 as with a 125W i7-10700K?

 

OK sure, the 65W CPU can't be overclocked, but let's leave overclocking out of this discussion unless it's relevant to sound quality.  Let's also for the moment leave aside whether the 125W CPU would enable more processor intensive HQPlayer filters/modulators than the 65W CPU.

 

Consider a 12V/10A SR7 rail.  10A isn't nearly enough to run either CPU at its maximum potential processing power.  For example, with certain benchmarks the i7-10700K will peak at over 200W.  Granted such workloads are not representative of a music server, even one running HQPlayer.

 

Here's the thing.  The 65W CPU single core turbo is 4.7 GHz vs 5.0 GHz with the 125W CPU.  And the all core turbo is 4.6 GHz on the 65W CPU vs 4.7 GHz on the 125W CPU.  So the peak power demand for both processors is going to be similar.  The key difference is in the base frequency which is 2.9 GHz on the 65W CPU vs 3.8 GHz on the 125W CPU.  Probably the 65W CPU can't sustain turbo for as long as the 125W CPU, but I haven't confirmed that by comparing the PL2 and Tau specs of these two processors.

 

So my question is whether the sound quality (maybe focusing on dynamics and transients) would be just as good with a 65W CPU as with a 125W CPU given a 12V/10A SR7 rail.

 

What does it matter?  Well, the lower power budget of the 65W CPU means it's going to be easier to cool, and certainly easier to accommodate in a fanless case.  And since 10A is a constraint for either processor, is it better to go with the 65W?  Since we are using high power processors at very low utilization, I think it's a legitimate question.

 

I don't know if anyone has a definitive answer, but I would really like some input from users who have some experience building systems.

 

Here are a couple of articles that got me thinking about this question.  

 

https://www.anandtech.com/show/15758/intels-10th-gen-comet-lake-desktop

https://www.anandtech.com/show/15785/the-intel-comet-lake-review-skylake-we-go-again/5

 

 

  

 

 

[StreamFidelity]

8 hours ago, rickca said:

So my question is whether the sound quality (maybe focusing on dynamics and transients) would be just as good with a 65W CPU as with a 125W CPU given a 12V/10A SR7 rail.

 

 

The answer is already in your very good explanation. 👍

 

Depending on the setting of the CPU clock frequency in combination with undervolting, I achieve low power consumption and an acceptable heat development even with a very powerful CPU.

 

The crucial question is, why buy an expensive, high-performance CPU when a smaller CPU is sufficient?

 

My Opinion: Powerful CPUs have to be of a particularly high quality. As far as I know, less powerful CPUs are sometimes just waste products of these high-performance CPUs. I would therefore tend to use the top product, because then the highest processing quality is given. So why not an i9-10900K? 😉

[guiltyboxswapper]

47 minutes ago, StreamFidelity said:

Depending on the setting of the CPU clock frequency in combination with undervolting, I achieve low power consumption and an acceptable heat development even with a very powerful CPU.

 

 

I tried this with a Ryzen 3900x + JCAT XE, running HQPlayer Embedded.  The good news is, yes MASSIVE gains in cooling to be had, and I was still able to run my filter + combo of choice.  Wall draw also reduced by 25w (100w vs 125w).   

 

This was achieved by simply setting the CPU to a lower vcore limit (undervolting), which caused the CPU to "budget" power accordingly.  This automatically limited the max clock speeds of cores, but they werent entirely far from top end frequency range either

 

The bad news is - it absolutely slaughtered dynamics, impact, low-level detail.... quite frankly it was like running an ARM streamer again.

 

Has anyone else achieved a "undervolt" on the Ryzen 3000 series without it killing the sound signature?

[lmitche]

1 hour ago, guiltyboxswapper said:

 

I tried this with a Ryzen 3900x + JCAT XE, running HQPlayer Embedded.  The good news is, yes MASSIVE gains in cooling to be had, and I was still able to run my filter + combo of choice.  Wall draw also reduced by 25w (100w vs 125w).   

 

This was achieved by simply setting the CPU to a lower vcore limit (undervolting), which caused the CPU to "budget" power accordingly.  This automatically limited the max clock speeds of cores, but they werent entirely far from top end frequency range either

 

The bad news is - it absolutely slaughtered dynamics, impact, low-level detail.... quite frankly it was like running an ARM streamer again.

 

Has anyone else achieved a "undervolt" on the Ryzen 3000 series without it killing the sound signature?

Rather than undervolt, why not disable "turbo mode" and all the PBO and overclocking stuff. At a 3.8ghz base frequency, power and heat should be much lower without impacting SQ. With 12 cores at 3.8ghz does Hqplayer DSD256 upsampling with the new modulators still play? Or does there need to be an even faster single core at a higher frequency?

[lmitche]

11 hours ago, rickca said:

Can I achieve the same sound quality with a 65W i7-10700 as with a 125W i7-10700K?

 

OK sure, the 65W CPU can't be overclocked, but let's leave overclocking out of this discussion unless it's relevant to sound quality.  Let's also for the moment leave aside whether the 125W CPU would enable more processor intensive HQPlayer filters/modulators than the 65W CPU.

 

Consider a 12V/10A SR7 rail.  10A isn't nearly enough to run either CPU at its maximum potential processing power.  For example, with certain benchmarks the i7-10700K will peak at over 200W.  Granted such workloads are not representative of a music server, even one running HQPlayer.

 

Here's the thing.  The 65W CPU single core turbo is 4.7 GHz vs 5.0 GHz with the 125W CPU.  And the all core turbo is 4.6 GHz on the 65W CPU vs 4.7 GHz on the 125W CPU.  So the peak power demand for both processors is going to be similar.  The key difference is in the base frequency which is 2.9 GHz on the 65W CPU vs 3.8 GHz on the 125W CPU.  Probably the 65W CPU can't sustain turbo for as long as the 125W CPU, but I haven't confirmed that by comparing the PL2 and Tau specs of these two processors.

 

So my question is whether the sound quality (maybe focusing on dynamics and transients) would be just as good with a 65W CPU as with a 125W CPU given a 12V/10A SR7 rail.

 

What does it matter?  Well, the lower power budget of the 65W CPU means it's going to be easier to cool, and certainly easier to accommodate in a fanless case.  And since 10A is a constraint for either processor, is it better to go with the 65W?  Since we are using high power processors at very low utilization, I think it's a legitimate question.

 

I don't know if anyone has a definitive answer, but I would really like some input from users who have some experience building systems.

 

Here are a couple of articles that got me thinking about this question.  

 

https://www.anandtech.com/show/15758/intels-10th-gen-comet-lake-desktop

https://www.anandtech.com/show/15785/the-intel-comet-lake-review-skylake-we-go-again/5

 

 

  

 

 

Hi Rick,

 

Yeah, I read the same article and agree with your assessment. An Intel labeled 65 watt tdp part that scales to almost 4x - 260 watts is definitely misleading. What's the point of the TDP rating? AMD does similar things, but appear to be twice as efficient with the 7nm technology.

 

Do you have any experience as to what happens when CPU demand overwhelms your power supply?

 

LOL,

 

Larry

 

 

[StreamFidelity]

25 minutes ago, lmitche said:

With 12 cores at 3.8ghz does Hqplayer DSD256 upsampling with the new modulators still play?

 

I linked Jussi's answer. 😉

On 4/29/2020 at 1:01 PM, Miska said:

Rough figure for Intel CPUs is around 4 GHz now. But it needs only two cores running at such clocks, other cores can usually run at lower clocks. So Turbo Boost can usually do it's job even on T series low-TDP models or likely on some newest laptop CPU models too (H series).

[guiltyboxswapper]

50 minutes ago, lmitche said:

Rather than undervolt, why not disable "turbo mode" and all the PBO and overclocking stuff. At a 3.8ghz base frequency, power and heat should be much lower without impacting SQ. With 12 cores at 3.8ghz does Hqplayer DSD256 upsampling with the new modulators still play? Or does there need to be an even faster single core at a higher frequency?

 

I can give it a try.  I'm using poly-sinc-xtr at DSD 512 rates, which according the turbostat (or whatever util i use) means all 12 cores are actually running a workload of 1.9ghz~ quite evenly.  Yes, if i was running EC i'd be in trouble as noted the Ryzen cores like to use PBO to achieve high single core speeds.

[rickca]

52 minutes ago, lmitche said:

Do you have any experience as to what happens when CPU demand overwhelms your power supply?

You know I do.  It depends on the design of the power supply.  My HDPLEX 200W fried the rectifier and went up in a puff of smoke.  The new 300W unit is supposed to fix that.

[lmitche]

5 minutes ago, guiltyboxswapper said:

 

I can give it a try.  I'm using poly-sinc-xtr at DSD 512 rates, which according the turbostat (or whatever util i use) means all 12 cores are actually running a workload of 1.9ghz~ quite evenly.  Yes, if i was running EC i'd be in trouble as noted the Ryzen cores like to use PBO to achieve high single core speeds.

That's right, your T+A dac is perfectly paired with poly-sinc-xtr at DSD 512.

 

Thanks, it would be great to know if poly-sinc-xtr at DSD 512 will run at the base frequency. Also, if there is a way to easily try the new EC modulators at DSD256 that would be very helpful.

 

Many thanks,

 

Larry

[lmitche]

3 minutes ago, rickca said:

You know I do.  It depends on the design of the power supply.  My HDPLEX 200W fried the rectifier and went up in a puff of smoke.  The new 300W unit is supposed to fix that.

Yes, it was a rhetorical question. Hence the LOL!

 

I think of you and this incident whenever sizing a power supply and making bios settings. I use a kill-a-watt when experimenting to ensure I don't hit the danger zone for any length of time. In a split power configuration with a Hdplex DC 400 ATX powering the big ATX power connector and a separate supply on the CPU, the DC-ATX never gets into the danger zone unless a GPU is involved.. The CPU is another story. Obviously one of the advantages of split power is that power demand can be spread across more supplies.

 

And yes, the ATX power can be spread across more supplies as well but SQ not power demand is the driver for this.

[guiltyboxswapper]

3 minutes ago, lmitche said:

That's right, your T+A dac is perfectly paired with poly-sinc-xtr at DSD 512.

 

Thanks, it would be great to know if poly-sinc-xtr at DSD 512 will run at the base frequency. Also, if there is a way to easily try the new EC modulators at DSD256 that would be very helpful.

 

I tried the EC modulator with a 2s filter lately, and found it played for about 5s at a time before a blip. 

 

Do bare in mind though that I have my cooling settings in my BIOS set to "silent" which means the CPU is easily getting to 70c when operating in EC mode.  I may try again with "normal" cooling mode on.  

 

But yeah, T+A clearly works best at 512 so .... no big deal.

[rickca]

12 hours ago, rickca said:

Consider a 12V/10A SR7 rail. 

Correction.  The non-custom SR7 does 12V/12A.  Apparently Paul has a new configuration that does 12V/16A.

https://audiophilestyle.com/forums/topic/30376-a-novel-way-to-massively-improve-the-sq-of-computer-audio-streaming/?do=findComment&comment=1061805

 

May I reiterate ... let's also for the moment leave aside whether the 125W CPU would enable more processor intensive HQPlayer filters/modulators than the 65W CPU.

[lmitche]

3 hours ago, rickca said:

May I reiterate ... let's also for the moment leave aside whether the 125W CPU would enable more processor intensive HQPlayer filters/modulators than the 65W CPU.

Rick,

 

Are you asking if all other things are the same, will the 65 watt cpu sound the same as the 125 watt cpu? And if not, which will sound better?