Building a DIY Music Server

[gererick]

I don’t agree with the formula.

 

Even if I did, I use Euphony and with a single CPU, high-powered server.  There may be nothing better and it is not expensive.  There wouldn’t be much gained, if anything, trying to find something better.

 

This allows me to focus time, money and effort on hardware, wire, network devices such as the modded Pink Faun Buffalo, power supplies, power conditioners, and other system components .  All of which can provide gains whereas trying to find better software likely will not provide gains, with my server.

 

So, for me, the value of the ‘formula’ is zero, and so is debating the weighting factors.

[lmitche]

OK, so thanks to everyone for your thoughtful feedback on the role of software in the sound quality equation. While there seems to be no overall consensus, the level of passion this topic engenders is impressive.

 

Nenon is right, I am interested in the 80 - 20 rule of music server sound quality. It would be sad to see people think that great digital playback can only be achieved with exotic power supplies and music servers that cost as much as a new car.

 

Enjoy your systems and be well,

 

Larry

[gererick]

I believe I purchased my Sean Jacobs 19V DC3 power supply for in the range of $1,500 or $1,600.  Separately, for a PF Buffalo and a Synology NAS, I purchased a DC3 with three rails - 12, 12 and 5 volts - for ~ $2,275.  There are competing options at around these amounts which may offer comparable performance.  The cost of a high-power DIY server is similarly manageable.  In 1985, one could purchase a Yugo for $3,990.  A new Yugo is no longer available.

[ASRMichael]

15 minutes ago, gererick said:

I believe I purchased my Sean Jacobs 19V DC3 power supply for in the range of $1,500 or $1,600.  There are competing options at around this amount which may offer comparable performance.  The cost of a high-power DIY server is similarly manageable.  In 1985, one could purchase a Yugo for $3,990.  A new Yugo is no longer available.

I found adding more rails gave more value. E.g. 6-7 rails powering 4 devices made more economic sense/value.

[ASRMichael]

42 minutes ago, lmitche said:

OK, so thanks to everyone for your thoughtful feedback on the role of software in the sound quality equation. While there seems to be no overall consensus, the level of passion this topic engenders is impressive.

 

Nenon is right, I am interested in the 80 - 20 rule of music server sound quality. It would be sad to see people think that great digital playback can only be achieved with exotic power supplies and music servers that cost as much as a new car.

 

Enjoy your systems and be well,

 

Larry

I believe in the 80/20 rule also. 80% being the power supply that is! 

[RickyV]

Well I just hate to mess around with software so that’s why I bought Euphony so I could concentrate on the hardware.

[austinpop]

2 hours ago, lmitche said:

I am interested in the 80 - 20 rule of music server sound quality. It would be sad to see people think that great digital playback can only be achieved with exotic power supplies and music servers that cost as much as a new car.

 

Larry, I completely hear where you're coming from. In the first few years I was on CA, this was my posture too, and it paid rich dividends, as there was so much improvement to be had with modest expenditure.

 

My trajectory changed — and I must stress this is unique to each person — when I realized 2 things. First, as I scaled further and further along the "exotic" axis, I was not only able to hear and appreciate the impact, I realized this was giving me the emotional connection to the music in a way that I never had before. Second, I'm of a certain age, where it's unclear how many good "listening" years I have left before my listening abilities degrade, so this is why I've allowed my budget and system cost to grow. Not everyone would make this calculation.

 

It's really important for people coming to these thread understand the guiding principles  of people who are posting. Some people consider DIY to be synonymous with the principle of  "optimization at moderate cost," but others view it as a way to go to the bleeding edge and beyond. The approaches are different, and perhaps it behooves us to periodically clarify which principle we're following.

[bobfa]

3 hours ago, lmitche said:

OK, so thanks to everyone for your thoughtful feedback on the role of software in the sound quality equation. While there seems to be no overall consensus, the level of passion this topic engenders is impressive.

 

Nenon is right, I am interested in the 80 - 20 rule of music server sound quality. It would be sad to see people think that great digital playback can only be achieved with exotic power supplies and music servers that cost as much as a new car.

 

Enjoy your systems and be well,

 

Larry

 

41 minutes ago, austinpop said:

It's really important for people coming to these thread understand the guiding principles  of people who are posting. Some people consider DIY to be synonymous with the principle of  "optimization at moderate cost," but others view it as a way to go to the bleeding edge and beyond. The approaches are different, and perhaps it behooves us to periodically clarify which principle we're following.

I had done a lot on the DIY route up until later this summer.  I got lost in the testing and fussing.  What I have been doing is simplifying.  Self Powered Speakers is one example.  I was trying to use my AURALiC Altair G1 as a single box, is another.  I “need” Roon in my equation so I need a server.  While my software and hardware are reasonably optimized, I know it could be better.   I am to the point that I need to just get a Roon Server built by @lmitche and get back to listening to more music.  Things move to fast and I am slowing down.  @austinpop suggestion about limited time has a ring to it,  I know where my body is failing me, and hearing is slowing being compromised.

 

I think I will tinker with Ham Radio and just Listen and Enjoy the music!

 

Bob

 

[MarcelNL]

IMO it's always good to have a strategy when tinkering, it's too easy to tinker with too many variables at a time and getting lost.

 

I try to isolate a couple of key parameters at a time, focus on getting to the bottom of each and then tweaking them before embarking on another bout of tinkering.

 

 

[jean-michel6]

13 hours ago, MarcelNL said:

Just out of curiosity, did you use that size for all wires and are those single leads? (meaning; do you have more leads running parallel for the same voltage)?

 

I just used single wire . 
The tdp for this cpu is 45w .... therefore I do not think it will ever draw 12A of current . Typically I measured 1 to 3 A .

[MarcelNL]

3 minutes ago, jean-michel6 said:

I just used single wire . 
The tdp for this cpu is 45w .... therefore I do not think it will ever draw 12A of current . Typically I measured 1 to 3 A .

makes sense, I was not aware of the CPU having that low TDP! Ever since my Le Monstre Amp build (decades ago) my approach to anything PSU is 'think big' ;-)

[MarcelNL]

I just bought a used Ryzen 5 1600 (6 core) CPU on a GigaByte B450 DS3H and copious amounts of fast RAM (64Gb, enough for a considerable RAMDisc) that I plan to test drive in a passively cooled case, either HDPlex or Monsterlabo using a likely external LPS as my PSU builds tend to end up being LARGE. It will at least allow me to compare against a NUC Pentium, and it should be compatible with the PF I2S bridge!

 

 

[MarcelNL]

so, finally I have Audiolinux going, slightly counterintuitive setup afaik....support was great though!

 

Comparing the same setup, both at 0db reduction, swapping an SSD, Audiolinux vs Daphile makes me think that Daphile (beta RT kernel) is a step ahead in SQ.

I have not tinkered with audiolinux, no ramroot, hqplayer settings to as close to plain PCM as I can as I want to have my R2R shine.

 

AL sounds like an 80ies CD player, dynamic but flat as a pancake and no overtones, decay, original acoustics. Daphile is like listening to a very good Turntable, involving rythmical, organic, large sound stage, refined highs

 

Will do the same comparison later on the AMD board using the I2S output, am I missing anything essential with AL or are there others with similar observations?

 

 

[gererick]

1 hour ago, MarcelNL said:

so, finally I have Audiolinux going, slightly counterintuitive setup afaik....support was great though!

 

Comparing the same setup, both at 0db reduction, swapping an SSD, Audiolinux vs Daphile makes me think that Daphile (beta RT kernel) is a step ahead in SQ.

I have not tinkered with audiolinux, no ramroot, hqplayer settings to as close to plain PCM as I can as I want to have my R2R shine.

 

AL sounds like an 80ies CD player, dynamic but flat as a pancake and no overtones, decay, original acoustics. Daphile is like listening to a very good Turntable, involving rythmical, organic, large sound stage, refined highs

 

Will do the same comparison later on the AMD board using the I2S output, am I missing anything essential with AL or are there others with similar observations?

 

 

Interesting info, AudioLinux vs Daphile, and interesting too that in your setup Daphile sounds quite good.  

 

I suspect that there are not a material number of followers of this thread that use Daphile, but we will see.

 

Daphile is both an app and an AL operating system, correct?

 

When you listen via AL, are you using an app?

[MarcelNL]

Listening to AL is done by HQPlayer, as I'm probably to stupid to work efficiently with Linux ; setting up things takes me forever and so far I only got the desktop version of HQplayer to work, not the embedded version. HQplayer is controlled using an app on my phone. 

 

Daphile is an all in one package, using LMS. Also not intuitive to use but IMO far easier to work with than AL (for me, I'm not passing any judgement here).

 

I have to add that switching to the latest beta RT kernel in Daphile was quite a jump in SQ, I had to switch as something changed that make Qobuz files skip, that is fixed in the beta...together with something else that changed I dare say.

 

Even the Daphile production RT kernel sounds more organic than what I so far have heard from AL, but the tinkering only just started.  Daphile sounded even better when the Jcat USB still worked, I'm hoping for another step using the PF I2S output w OCXO, but am a bit hesitant about the AMD setup...new linear PSU's needed, not an easy comparison with the NUC on a linear PSU so I probably end up comparing AL and Daphile without all the bells and wistles added except for the Linear PSU for the PF I2S and an external PSU for the SSD.

 

I tried to make the comparison as valid as possible, one key difference is the SSD, though I auditioned the new one with Daphile prior to installing AL and could not detect a major difference between the two. (the Daphile OS SSD also contains most of the music files and is EMI protected, the AL SSD also contained the music but is not yet EMI protected)

And as always, this is in my system, my ears, YMMV...

 

I thought about being biased, Daphile is free AL plus HQplayer runs at approx 300e, after a couple of swaps I'm sure it's not bias what I'm hearing.

[MarcelNL]

Halleluja, I almost needed a younger cousin to get this beast going Daphile....I'm getting old, or at least rusty at making computers do what i want

Finally Daphile is installed, now for the rest.....tomorrow, for a lack of jolt cola

 

 

[bobfa]

On 12/8/2020 at 7:16 PM, lmitche said:

Nenon is right, this a game of the sum of the parts. Based on recent experience, I'd put software at over 50% of the contribution to SQ with hardware/cabling and power supplies at 25% each.

 

SQ = .5(SW) + .25(HW/Cabling) + .25(LPS)

 

What do others think about this formula?

 

Larry

I agree with @Nenon and @lmitche that the work is segmented.  But like Apple has done with their M1SOC putting all of the hardware together and then close coupling the software creates something that is larger than the individual parts.  in the DIY world each of us has our own segment of skills.  I am pretty good at picking the pieces to put together.  I am not one to figure out what needs to be removed/tweaked in the Kernel or how to stop Roon from writing log files until the cows come home.  I grok a lot of the hardware issues, electrical noise, and some of the crazy stuff that happens in electrical cables of all kinds.  And of course none of this fixes a bad sounding room!  Or the need for vibration damping at multiple levels.

 

The stuff I have built on the computer side has gotten better and better, but nowhere near what others have seemed to achieve.  One lesson I am pushing hard on is making things simpler.  Less cables, less devices, less of everything.  So far that has gotten me a lot further along.  

 

Looking at the single computer/server model of systems design has been where I have put a lot of energy this year.  I have made pretty good improvements.    I have reached the end of my knowledge to go deeper.  I will see what unfolds here soon.  

 

There is a lot underlying what @lmitche has done over time, along with many others.  This is a very big playground.  We can all contribute!

 

 

 

Bob

 

 

 

 

[gererick]

On 12/10/2020 at 10:35 AM, lmitche said:

OK, so thanks to everyone for your thoughtful feedback on the role of software in the sound quality equation. While there seems to be no overall consensus, the level of passion this topic engenders is impressive.

 

Nenon is right, I am interested in the 80 - 20 rule of music server sound quality. It would be sad to see people think that great digital playback can only be achieved with exotic power supplies and music servers that cost as much as a new car.

 

Enjoy your systems and be well,

 

Larry

 

 

_{As there has been some discussion as to cost of a DIY high-power server, below is an indication of the parts costs for my build.  Again, this is parts only.}

 

_{Motherboard – ASRock Z390 Gaming ITX – $243 with tax}

 

_{CPU – Intel Core i9-9900 – $507}

 

_{Optane Card – Intel MEMPEK1J032GA01 – $96}

 

_{PinkFaun USB Bridge with ultraOCXO clock, $1,860}

 

_{JCAT USB Femto Network Card, $503}

 

_{H5 HDPlex Chassis, $340 (versus the Streacom in my build)}

 

_{4 rail power supply, ~$2,700}

 

_{Wire – too wide of a range – assume Ghent Neotech JSSG 360 DC cables, $900 (instead of Mundorf silver gold throughout)}

 

_{Total - $7,149}

 

The USB bridge and network card are optional, or lesser models / less expensive models are available.  If you instead use the ports on the motherboard, you will lose some SQ but reduce the cost by $2,363 for a revised total of $4,786. 

 

One can always buy those later if cost-constrained.  Or buy less costly models.  JCAT’s best USB card is 800 euros, and it may be competitive with the more expensive PF USB card with the ultra OCXO clock.  Pink Faun’s USB cards are 325 euros (no clock), 875 euros (OCXO clock), and 1,575 euros (ultra OCXO clock).  One can also spend less on DC cables, and upgrade those later if desired.

 

A Euphony license is not listed and adds $299 (less expensive than Roon’s $700, and the OS is part of the Euphony license).  Daphile is free, so one can avoid or delay software costs altogether, if wished.

 

There are now likely better motherboards than mine, including motherboards which have drop-in slots for the USB card and the network card.  This would cut down on DIY work also.  If anyone has a suggestion for the best current motherboard / CPU combination, please let us know.  Versus this build from January 2020, things have changed.

 

In any event, this is likely superior sound quality to a low-power audiophile-grade NUC with AudioLinux and Roon.  It was for me.  Obviously at less than the cost of a new car (2020 Chevy Spark Hatchback, $14,395, seems to be the cheapest), in further response to a recent post as quoted above.

 

Below is an interesting chart for people to see where the money for such builds is going.

[bobfa]

4 minutes ago, gererick said:

Below is an interesting chart for people to see where the money for such builds is going.

 

From your chart you are missing an important variable.  The cost of time, building, tuning, adjusting.  That could double the cost.  Many in the hobby discount that because they feel that their labor is at no cost.  I disagree rather vehemently. Do some rough numbers since you built recently.  How much time in Research, development and acquisition and finally production?    

[Nenon]

1 hour ago, bobfa said:

From your chart you are missing an important variable.  The cost of time, building, tuning, adjusting.  That could double the cost.  Many in the hobby discount that because they feel that their labor is at no cost.  I disagree rather vehemently. Do some rough numbers since you built recently.  How much time in Research, development and acquisition and finally production?    

 

Here is an updated chart then.

 

 

I charged @gererick a nominal fee of $500 for this build. It was a cool experiment and a way to document it and share with the community. Part of my hobby.

Disclaimer #1: I also sold my DIY modules from Sean Jacobs to @gererick as part of this build, but that was at a lower price I paid for them, so I did not make money out of them. That was way before I got affiliated with Sean. 

Disclaimer #2: I don't build computers for people anymore. But of course I am happy to share my findings as always. 

 

Yes, you can spend days, weeks, months, even years tweaking AudioLinux to perfection. That's why we picked Euphony. For $299 and less than a handful number of settings I have shared earlier in this thread, the results are astonishing. At that point it's the hardware that matters, and you can use your budget for better sounding components.

 

We chat from time to time, and as far as I know, @gererick has not spent any time tweaking, tuning, adjusting things since he got this built. Neither have I changed any of his settings since day one. All his time was spent for music listening! 

 

On the other hand, my chart would look like this, in the best case scenario :)

 

 

This is what this thread is all about - share our DIY findings, discover what works and what doesn't, stay at the bleeding edge of the technology. Ideally, people should be able to read a few pages, and build something really good if they are into DIY. Then some of them may just decide to enjoy the music, while others will continue to push the limits and share.

 

I encourage everyone to share! Share what works. Share what does not work. No one benefits from keeping secrets except those who want to monetize from them. 

 

I have been quiet for a while with sharing experiments, but I have not stopped experimenting. My affiliation with Sean has helped me sponsor some experiments I wanted to do. Too early to share yet, but I expect that in early 2021 we would have some crazy good developments in DIY that would push DIY way ahead of most commercial solutions! 

[bobfa]

@NenonI have a pretty good idea about my R&D costs as everything I have done is R&D. I have not produced a product that I am willing to call ready.  That is mostly because I never had a fixed target.  Equipment and software costs at around $35000.  Labor is near to $100,000.  I have. Built/assembled / purchased eight systems.  Each would cost around $17,000.    The last build is in operation now.  That is the cost of R&D and a lot of fun/pain!  I am pretty sure that I am done with DIY.  It was a chase I maybe should not have done.  

 

The current build will be replaced with a different system soon.  (Commercial-built-to-order, Cost not evaluated yet) That system will use some of the existing hardware, LPSs, and a few other bits.

 

Two of the prototypes will be in production soon, repurposed!  One as a Windows Gaming machine and the other as a Linux server running Docker containers in my network..  One of the NUCs is up for sale and the other NUC is a spare Windows Machine. I sold two sonic transporters and an Antipodes CX/EX and some of the other bits and bobs.  One of the machines is scrap, one motherboard fried, one 8TB HD fried.   My accounting has not been completed for this year, but there is a net loss in operating/ R&D budget.  The company has incurred some debt in the process and the employee, ME has not be properly paid!

 

 

[gererick]

1 hour ago, bobfa said:

From your chart you are missing an important variable.  The cost of time, building, tuning, adjusting.  That could double the cost.  Many in the hobby discount that because they feel that their labor is at no cost.  I disagree rather vehemently. Do some rough numbers since you built recently.  How much time in Research, development and acquisition and finally production?    

 

Time tuning and adjusting is nearly zero.  With Euphony (and Daphile, which MarceINL is enjoying), there is no tuning.  There are some settings, in setting up the software, a bit about the cache and how you want that to be deployed etc., but these sorts of things are more of a learning curve with the software than trying various frequency or turbo settings and trying to figure out which one sounds best.  One doesn't and can't do that with Euphony (probably not with Daphile either).  And Euphony still sounds better than AudioLinux + Roon (in my set-up, and this seems to be the view), despite the fact that you can't tune or adjust it.  And Euphony is all you need to install.  Some want to install HQPlayer also, which is fine, of course.

 

Time in acquisition?  There's maybe ten things you need to buy.  Once I decided on Sean Jacobs power supply, that took minutes.  It is likely the same if one decides on a different power supply.  An H5 case is minutes.  If you limit your decisions on network and USB cards to either JCAT or Pink Faun (I don't know who the others are), that is quick.  Overall, acquisition time is, I don't know - 90 minutes?

 

Time in research and development?  That's what this thread is for.  People on this thread provide their experiences with this CPU or that one, this motherboard or that one.  Nenon provided those specifics in the first few pages of this thread.  Others have chimed in.

 

Time in production?  I don't know how long it takes to drill two holes in the back of the case for connectors to the CPU and the motherboard, but not long.  The connectors just screw into those holes.

 

Inside the case, the motherboard is just a screw in item.  The CPU is a screw-in item also, but it has to be epoxied with thermal paste to the cooling elements, so it is not limited to fasteners.  The network card and the USB card drop right into slots on the motherboard, if the motherboard permits.  The power supply is purchased.  The DC cables between the PS and the server are purchased.  The time spent wiring the from (a) the connectors to (b) the connection points inside the server takes time and knowledge that I don't have (yet).    

 

A DIY build of a high-power server allows for audiophile-grade network and USB cards.  That isn't possible with a NUC.  Euphony also sounds better on a high-power server than it does on a NUC.  So you lose some big potential advantages with a NUC.  You can't fix that later.  With a DIY, you can choose not to have the network and USB cards, but you can address those when finances permit.

 

Production seems straightforward, save for the wiring inside the server to power the motherboard and CPU.  That may be straightforward also, but I don't know. 

 

The USB card and the network card don't need to be wired.  They each have DC jacks - you just insert the DC barrel into them from outside the case.  They don't need to be wired internally inside the server - they just drop into slots in the motherboard slots.  

[MarcelNL]

I agree that a NUC cannot hold it's water against a proper sized MB and case for the very reason everything you add sits on top, hangs out or at least needs an umbilical unless you get a larger case to put the NUC MB and peripherals in!

 

I just completed the raw POC build, installing the PF I2S card in the AMD 5 1600, which has 64Gb fast RAM on a GigByte B450 ds3H V1 MB, using the regular 'silent' PSU. Only the I2S card is fed by a 5V linear PSU, the sound is better than the NUC, just not (yet) as organic. The I2S card connects to the DAC with some cheapass premade ethernetcable, as I was too impatient to build a longer proper one.

 

I simply whacked in the Daphile SSDafter getting tired of fighting boot systems (UEFI vs legacy) etc, and it worked, the I2S card was auto detected and we have a lift off.

 

That is, all electrons need to still break in, OCXO is stone cold needs to break in many hours etc. etc. yet I do not hear 'digitis' or nasty stuff. 

 

From what I hear right now there should be plenty of potential for improvement, right now with only 15 min of playing temperature monitoring shows I run the risk of having to heat the CPU as it runs at 28'C so going passively cooled is def possible.

 

Sound quality? Lows are lower highs more transparent, there is more dynamic range, across the bandwidth, staging etc is hard to say as I am playing with only one concept speaker (room constraints in a temporary house make it impossible to add another speaker, 2.10 meter by 2.10 meter, 1 meter deep....WAF you know).

 

So now up to see if putting all in RAM brings something, or that I buy an Optane to play with (though AMD does not support it like Intel-duh- it might sound better due to a different working mechanism, and I want to heat that)

 

To the investment: I bought the computer used, some 300 euro all in all. PF I2S card 1000euro (VAT is a great invention). The real cost will come with building several linear PSU's and a passive case. I don't count the hours, this is a hobby I get enjoyment out of it and I only bill my clients. 

 

[bobfa]

14 minutes ago, gererick said:

but I don't know.

The above quote is a bit out of context but it is apt to where I am going: That is part of what kills you.  What do you do not know.   How many hours did you have fun reading the forum?  Is that a cost item to you?  The “Massively” thread that @austinpopis guiding over 700 pages that wonders around it multiple directions.   WOW what I have learned from interacting on that one.   How many days to read that one folks?

 

Or we can talk about my Linux shootout thread.  That was a lot of fun and we all learned a lot.  Testing all of the OS’s for my review of the Allo USBridge Signature was so interesting.  

 

 It is what drives the DIY world.  Great stuff.  Part of this is journey.  When I was into Model RR I finished three models, I never ran them for long.  I am not down on this at all.

 

 I want to challenge the thought that we can build cheaper than Commercial in many cases.   We have to know when to stop, or not.  I was looking at this from the business side. It is fun, but not viable for me.  I write all of this to explain my story and present a cautionary tale.  Have fun, know your limits.  Fail faster!  Do not run to a moving target.

 

 

We have a lot of pre-conceived thoughts in our heads.  Many have been soundly trounced here on AS.  Make something new and exciting.

 

 

I may be going back to assembling nice neat Lego Blocks, you should pick your own paths.  Figure out something new, tell us how it sounds or works.  Tell us about what did not work so we do not trip over the same crack in the sidewalk.   I will do the same!

 

So GO BUILD SOMETHING, HAVE FUN!  Really, I mean it!    

[MarcelNL]

28 minutes ago, bobfa said:

I want to challenge the thought that we can build cheaper than Commercial in many cases.   We have to know when to stop, or not.  I was looking at this from the business side. It is fun, but not viable for me.  I write all of this to explain my story and present a cautionary tale.  Have fun, know your limits.  Fail faster!  Do not run to a moving target.

What if that moving target is part of the journey towards sublime sound?

 

I like that path, as much as the sound it created... the pitfall is that no sub-project is ever finished, so there your warning is justified IMO. 

 

My target keeps moving as I learn new options for making things sound better, more natural. My audio buddy started making his own resistors and capacitor, you can do all sorts of crazy (far out too) stuff and find out what works, and that opens up new avenues.