“Music Server Sound Quality: Better than Disc”

[dallasjustice]

I just got the new fourth version of Robert Harley’s book entitled “The Complete Guide to High-End Audio.” He has a section in the book related to music servers and a subtitle “Music Server Sound Quality: Better than Disc.”

 

Harley unequivocally states that no CDP can compete with a properly configured music server due to the inherent disadvantage of the moving disc versus the motionless hard drive and the difference in how a computer reads data and how a CDP reads data. He says he’s compared the SOTA CDPs with properly configured music servers via the same digital cable and DAC. He claims the music server is invariably more spacious, less fatiguing, more convincing, etc.

 

Although I am not surprised by Harley's findings, is it becoming accepted orthodoxy among audiophiles that music servers are superior to CDPs?

 

 

[audioengr]

Depends on the equipment. There are turntables that dont sound as good as CD players. This does not make every CD player better than vinyl...

 

However, there is computer audio equipment for the first time in history that will outperform even the most expensive CD players and transports. This is good news because there are no moving parts and no media to degrade over time. Many of these systems are also less expensive. For the first time in history, best of show was awarded to a computer audio room at RMAF.

 

Steve N.

Empirical Audio

 

[bdiament]

Hi dallasjustice,

 

My experience, using the original master as the sole criterion for comparison, has me in full agreement with Mr. Harley.

 

I've been saying this since I mastered my first CD in 1983: Pressings from different plants (even different lines within the same plant) all sound different from each other and none sounds indistinguishable from the master used to make it.

 

While the very best pressings come quite close (still no cigar but quite close), playback from hard disk (which incidentally, does spin) does sound indistinguishable from the master.

 

In fact, proper extraction from a CD has always, in my experience, resulted in a file that sounds like the master from which the CD was made.

 

(Note, I'm talking about extraction to a raw, non-data reduced format like .aif or .wav.)

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

[labjr]

" no media to degrade over time."

 

I'm not comfortable with the thought of downloading everything. I like the artwork, liner notes etc. Would be nice if we could buy a DVD data disc or USB memory stick in a cardboard album jacket and load it into the computer. At least have some physical media to collect and put on the shelf.

 

 

 

[bdiament]

Hi labjr,

 

"I'm not comfortable with the thought of downloading everything. I like the artwork, liner notes etc. Would be nice if we could buy a DVD data disc or USB memory stick in a cardboard album jacket and load it into the computer. At least have some physical media to collect and put on the shelf."

 

I'm with you 100%. Part of the Art for me, is in the object. With my CDs, I miss LP sized cover art but still, the booklet and the physical disc are things I value along with the music they support.

 

This is one reason why Soundkeeper has not gone the download route with our files-on-disc versions. Though each has the liner notes in pdf form and the cover art as a jpg, they are packed in jewel boxes with the same artwork the pressed CD comes with, i.e. a booklet that one can hold in their hands to read away from the computer screen. And of course, a physical disc for one's real (not "virtual") collection that is also a backup for what goes on the hard drive.

 

Reference and (I believe) Chesky are using a similar model and I think it is a good one.

 

Best regards,

Barry

www.soundkeeperrecordings.com

www.barrydiamentaudio.com

 

 

 

[mwheelerk]

Once the artwork and booklet get to the point of being able to be displayed on an HDTV from a download file, and I think it will but how soon is a big question, I will be totally happy with a complete download solution. To me it is the only thing lacking.

 

I can understand others enjoyment of the physial aspect of an album jacket or CD booklet but personally I don't need it, I just need the information. And, viewing your album artwork on a large screen HDTV is awesome. I have seen details I have never noticed before in artwork I have looked at (I thought closely) for 30 years.

 

[astrotoy]

What is more, I can clearly hear the improvement that a SS drive has over a spinning drive. I have both in my current computer system and with 192/24 files the difference is easily heard - less so with 44.1 files - though there is an improvement there also. I wish it were not true, since the capacity of an SS drive is so much more limited than a spinning drive and at a much higher cost per byte. Since my 192/24 files take about 2GB per half hour, a 128 GB SS drive fills up very quickly! (BTW, I am using a Zalman PC running Windows XP with Merging Technologies software and hardware (Pyramix and Mykerinos card), all running into a Pacific Microsonics Model Two.)

 

 

 

[audioengr]

"At least have some physical media to collect and put on the shelf."

 

This is risky. If you want it to play 30 years from now, it may not. It also may play badly if the pits become deformed over time.

 

If you have the data on a RAID system, the data will never be lost and never be compromised. It will always be the same. Its like those older movie films that they are video digitizing now because the film is deteriorating.

 

This is like any other change. People generally dont want change. However, this is not that painful and it is a positive because of the improved SQ. Learning is good for you anyway. Those that stop learning as they age are destined for dimentia or even Alzheimers.

 

Steve N.

Empirical Audio

 

[dmccombs]

Dallas,

 

I think computer based servers now beat CDPs of equal costs.

 

I would stack out Mach2 MacMini up against any $1500 CD player even if we use the headphone output on the MacMini and no DAC.

 

If we pair the Mach2 with a good $1500 DAC, I would put it up against any $3000 CDP player.

 

I am using our Mach2 Music Server as just one example. People are using stock MacMinis, other Macs, and PCs as music servers and I think generally, they will beat equally priced CDPs.

 

If you can find a friend with a decent Music Server and compare it to a similar priced CDP, I think you will agree. ;-)

 

Darrell

 

[Windows X]

With fully hiend music server containing:

-RME HDSP AES-32 (Best in market under $5k interface...I tried all in this price range except Mykerinos but still tempting)

-Highend grade toroidal linear power supply (same level as what used in about $10k DAC/Pre)

-Modded motherboard/RAM/etc. (changed capacitor in various places and some mods to get higher fidelity)

-Fully optimized system (Solid state drive/lowest power consumption design/no vibration/HFX case for solid resonance controlling)

-Best possible software configuration (native and stable 2ms playback latency...500hz!/dual-wire support/ethernet connection enabled for router access with iPhone/iPad)

-0.02ppm Antelope Isochrone OCX-V (I plan to modify its PSU and change capacitors in future so it needs thicker case)

 

Its linearity and sound quality puts most CD players and transports under $5k to shame, let alone shabby Linn/Naim/Olive/PS/Mac. However, when it faces the real challenge like Esoteric P-05.....not the best ones in the world but little brother from best CD Transport creator. I still can't beat the crap out of it though transport for the real sound.

 

I think Mykerinos can keep up to be in about the same league as Esoteric P-05 but I should consider its price to compete with Esoteric P-03 instead which sounds A LOT better than P-05. And if you disagree with this, please build up at least $100-200k system in your house and toy around for a year or two before complaining.

 

[ontherocks]

hello windowsX

i am interested in the mods you have done to the motherboard can you suggest some pists of improvment? I finished building a caps2 with a paul hynes linear power supply with separate rails for the mobo the ssd and the usb card (yet to be built)

 

[ItemAudio]

The point about a spinning disc/disk is useful: too often we hear negative comments about 'disc-spinners' in the same sentence praising computer audio systems with hard drives, which are comparably unhelpful in an audio system.

 

@ontherocks: we find the 12V Paul Hynes PSU is great for 12V: are you also using linear 5V and 3.3V?

 

[20hertz]

I have been listening to the Aune S1 player for the last couple days. It is better than the two transports I have here and a lot better than my PC. Playing 44.1 on the Aune sounds better than 24/192 from the laptop.

 

[ontherocks]

itemaudio:

the starting point is a SR7EHD with a 400VA mains transformer which can accomodate up to 4 separate rails. I planned initially to use 3rails for the 12V motherboard, the 5V ssd and the 12V usb card. I sent to him the currents requirements and voltage tolerances of these devices that i found. but since there is an additional available rail i am thinking to use it to power the DDR2 memory as well, i have read posts on computeraudioasylum of people doing so. I still dont know the current requirements of the RAM nor the voltage nor how to do that. I have to finalize the specifications with him and just sent my caps2 to him. Any advice much welcome!

Concerning the regulators to be used it will be his hipower PR3iEHD except for the SSD for which the use of a shunt SR3 regulator is planned. I will have customized regulator modules using Z foil SM resistors and elna silmic 2 caps. What do you think of the approach? Concerning connectors it will use industrial Jaeger connector on the psu and no connector on the music servor with hardwiring on the circuit boards except for the ssd. Internal wiring on psu music server and umbilical will be silver solid core with rfi rejecting braiding geometry.

Another psu will be made later for my metrum acoustice dac octave (it happens Paul Hynes has just ordered one for his system) and a esata HDD.

 

 

 

 

[Mark Powell]

I think it is 'accepted orthodoxy' yes. Whether it is true or not is another matter. I have been able to convince myself that my computer setup is superior to my 5000 dollar Naim CD player, but no one else says they hear any difference.

 

Of course they think I am crazy to be bothered with such issues.

 

Another example.

 

Equally, when I purchased my 5000 dollar CD player above, I quickly convinced myself that it was better than my 300 dollar Cambridge Audio multi-purpose DVD, DVD-A, SACD, HDMI, DivXpro, Dolby Digital, DTS, CD, player (I just read all that stuff off the front) because it was more expensive, made by a 'respected' company, was more 'purist' as it plays CDs only, and wasn't covered with silly trademarks.

 

Haven't a clue whether any one is actually better than any of the others.

 

[1audio]

Powering the RAM is difficult because its low voltage high current under Bios control. The on board switcher does this really well. Doing it with a remote supply should be done with a remote sensing supply after you figure out the correct voltage for the ram. Some motherboards will sense the voltage and throw an error if it doesn't respond to system control. The supply sequencing is also not a minor issue. many parts could get fried if the supplies are not brought up (and down) in the right sequence. And finally a 3.3V and especially a 1.8V or 1.2V supply delivering 5-10 A will dissipate a lot of power (2X the load or more) to do that.

 

I'm not convinced that this would do anything useful. The noise won't affect the data reliability of the ram. The noise reduction from the remote supply may be swamped by the noise radiated across the long cables.

 

Spending a fortune on an audio system won't necessarily make it sound better, or even good at all. There is a lot more technique and execution to getting good sound reproduction than money. There is no intrinsic reason why a disk would sound better or as good as a file played from a computer. I have pointed out elsewhere that most high end sound cards were designed for sound production, not reproduction and may not be the best choice for a reproduction system.

 

[ontherocks]

To Demian:

The idea of powering separately the RAM came from this thread:

http://db.audioasylum.com/mhtml/m.html?forum=pcaudio&n=98778&highlight=memory+power+supply&search_url=%2Fcgi%2Fsearch.mpl%3Fsearchtext%3Daudiogon%26b%3DAND%26topic%3D%26topics_only%3DN%26author%3D%26date1%3D%26date2%3D%26slowmessage%3D%26sort%3Dscore%26sortOrder%3DDESC%26forum%3Dgeneral

 

thanks for your informative post, a few questions:

 

a) "Powering the RAM is difficult because its low voltage high current under Bios control":

a-1)SSD and SOtM USB card are also under BIOS control?

a-2) for the RAM what implications with regards to BIOS adjustment if you plan for a separate remote PS or rail?

 

b)"The supply sequencing is also not a minor issue. many parts could get fried if the supplies are not brought up (and down) in the right sequence"

Even if i dont pursue the goal of having a separate rail for the RAM are there any supply sequencing issues for separate rails for the MoBo, the SSD and the USB card used in the caps2? And how to cope with that operationnaly?

 

c) "The noise reduction from the remote supply may be swamped by the noise radiated across the long cables."

Seems to be a good point but what if you shield properly the supply wires? A local shunt regulator is not possible because of the current requirements of the RAM.

 

The SSD seems has low enough currents requirements enabling the use of a shunt regulator , what do you think of the idea? Generally speaking shunt is preferred over series regulator and in that case this shunt reg from paul hynes has noise rejection up to 1GHz suitable i guess for digital compared to his series hi power modules which are limited a a few hundred KHz from memory

 

[1audio]

Interesting link.

 

from easy to hard-

Shileding the cables won't help because its magnetic field. The 4 conductors would need to be very tightly twisted and that might make the remote sensing unstable. Its a can of worms you would not want to get into.

 

External to the motherboard is less likely to be an issue. If it has a separate power connector then it should not matter which sequence the power comes up in. The internal supplies are controlled by the bios (all that overclocking stuff comes in here) and in some cases need to change dynamically in operation.

 

The SSD most likely has an on board regulator since the internal flash probably runs at a much lower supply voltage, and it probably is a switching regulator.

 

The larger question is why use the linear regulator? Will the lower noise on the supply for the PC make the data better? (probably not.) Will it radiate less noise than the switching supply its replacing? (probably, however it could radiate 60 Hz and 120 Hz hum). Will it reduce the noise going back up the AC? (yes). Will it "break" a ground loop? (probably). Will it use more power? (Lots more)

 

There are valid reasons for a liner supply, however it should be used intelligently or you are trading noise for noise.

 

30 years ago computers all had linear supplies with remote sensing. The supplies were large, heavy and expensive. The cabling and connectors were also a pain and expensive. That's all gone now. I hope we don't need to bring that back.

 

[ontherocks]

Demian,

 

The more I know...

 

To fully understand:

 

a) "External to the motherboard is less likely to be an issue. If it has a separate power connector then it should not matter which sequence the power comes up in. The internal supplies are controlled by the bios":

 

In my case i plan to use for the caps2 an external power supply from paul hynes, one case, one mains transformer and 4 separate fully regulated rails (see above post)for the SSD, the MoBo (which has a P4 Molex power connector), the SOtM USB card and may be the RAM. The umbilical will go through the caps2 chassis directly to the components to be supplied and molex connectors on the MoBo and USBcard will be removed (hard wiring). In that case should i understand power ignition sequence is not an issue and no change in the BIOS is needed? Please help!

 

b) "The larger question is why use the linear regulator?"

 

For all the reasons you mention + in addition I use separate rails for the music server main components to eradicate ground return current inter-modulation caused by powering more than one circuit from the same power supply. Excerpt from paul Hynes circle:

 

Up to four fully regulated supply rails can be fitted depending on the voltage/current requirements of the equipment to be powered. The supplies have their own separate secondary windings on the high performance custom mains transformer for complete galvanic isolation from each other. This eradicates any ground return current inter-modulation caused by powering more than one circuit from the same power supply, giving a more stable musical presentation both dynamically and dimensionally and it also reduces the masking of low level information caused by the mixing of these ground return currents with a common supply.

 

c)"There are valid reasons for a liner supply, however it should be used intelligently"

 

Regardless of cost/benefit efficiency ( I am already commited...) is there anything in what i plan to do which i should avoid or lead to inferior SQ?

Please note that I have an isolation transformer inside the paul hynes PSU as well.

 

d)"you are trading noise for noise": you speak of the 60 120Hz hum?, where does this comes from and does it affect all linear psu or only the off the shelves ones? Any way to avoid, mitigate it?

 

 

 

 

[bstcyr]

20hertz, I'm curious what software you're using on the PC that does not sound as good as the Aune S1 and what are you using to store files that feeds the Aune. The Aune looks like a great device, I always thought that something like this made sense- a customized computer with a stripped down OS like linux rather than windows, which puts it in reasonable price range.

-where is the Aune available?

 

[ItemAudio]

I believe they have a list of resellers on their website...!

 

[Mark Powell]

I thought about building a CAPS 2. In the end I thought why not a 'toaster' player, in my case the lowest cost Naim? All the work, linear power supplies, a great big toroidal transformer working at well below its capacity, 28 linear regulators dotted around where needed, shielding, all the rest of it, is done. And it was up and running within ten minutes of opening the cardboard box. I still have to use a regular computer for ripping/downloading and putting the files on my NAS, but that is all. And most of Naims' development work is guided by listening.

 

(You can buy an optimised ripping NAS from them too, but it is rather expensive.)

 

And if you do all this optimising on a CAPS 2 the cost will not be a whole lot lower, if you don't count the optional Naim NAS.

 

Few of us buy a low cost preamp, power amp, speakers, etc. and then work on 'improving' them. Why do it with the player? And some of them are a rather lower cost than the Naim, though maybe not as physically 'optimised'.

 

Just a thought...

 

[The Computer Audiophile]

Item - Let me guess, you're on that list :~)

 

 

 

[1audio]

But eBay is. . . .

 

Makes it tough to be a distributor when you are competing with a trading company in China.

 

Cute product.

 

[Audio_ELF]

Chris, with respect I think your response to Item Audio was unfair (or at least had an unfair subtext).

 

He was answering a direct question. He didn't promote his own company.

 

If he can't make such a comment (though perhaps a link to the aforementioned list may have been a more useful reply) then you may as well ban him from being here...

 

Eloise