DAS vs SAN vs NAS

0 Members and 1 Guest are viewing this topic. Read 1038 times.

James Tanner

  • Facilitator
  • Posts: 17527
  • The Demo is Everything!
    • http://www.bryston.com
DAS vs SAN vs NAS
« on: 3 Feb 2019, 03:13 pm »
Hi Folks,

Doing some experiments - any thoughts on which of the following would provide the best fidelity?

DAS is a block device from a disk which is physically [directly] attached to the host machine.
You must place a filesystem upon it before it can be used.
Technologies to do this include IDE, SCSI, SATA, etc.

SAN is a block device which is delivered over the network.
Like DAS you must still place a filesystem upon it before it can used.
Technologies to do this include FibreChannel, iSCSI, FoE, etc.

NAS is a filesystem delivered over the network.
It is ready to mount and use.
Technologies to do this include NFS, CIFS, AFS, etc.

Pundamilia

  • Jr. Member
  • Posts: 193
Re: DAS vs SAN vs NAS
« Reply #1 on: 3 Feb 2019, 03:48 pm »
I suspect that if one has a full FibreChannel network, the SAN would offer the fastest response and the cleanest signal and therefore perhaps better audio quality. However, given the associated costs, I think most people who don't have a FibreChannel in their home (99.9% of Bryston users) would be better off spending the money on audio equipment or content. Leaving the other two options.

Krutsch

  • Full Member
  • Posts: 546
Re: DAS vs SAN vs NAS
« Reply #2 on: 3 Feb 2019, 07:16 pm »
NAS, without question, as I describe in this post:

https://www.audiocircle.com/index.php?topic=161895.msg1725809#msg1725809

@Pundamilia ...

Quote
I suspect that if one has a full FibreChannel network, the SAN would offer the fastest response and the cleanest signal...

FC is pretty over-the-top for audio applications  :lol:

If you are suggesting this for optical isolation purposes, my post above discusses this (i.e. there are inexpensive, consumer routers that offer SPFs for optical gigabit ethernet, though it's unnecessary for this application - copper Ethernet is already galvanically isolated).

Pundamilia

  • Jr. Member
  • Posts: 193
Re: DAS vs SAN vs NAS
« Reply #3 on: 3 Feb 2019, 07:34 pm »
I certainly wasn't advocating running fibre through your house (unless you're running a Data Centre out of your garage!). I was just saying that I think a SAN is overkill.   :lol:

zoom25

  • Full Member
  • Posts: 770
Re: DAS vs SAN vs NAS
« Reply #4 on: 3 Feb 2019, 09:34 pm »
Lumin X1 has both copper and optical fibre ethernet inputs.

(I haven't followed the Uptone EtherRegen developments in the past few months, so don't know if John Swenson's research has been published. Aside from leakage and noise measurements with ethernet, they were looking into jitter and phase noise with ethernet which is asynchronous and non-realtime. Yet they think something might be there.)

Anyways, James, what kind of experiments/setups are you doing?

James Tanner

  • Facilitator
  • Posts: 17527
  • The Demo is Everything!
    • http://www.bryston.com
Re: DAS vs SAN vs NAS
« Reply #5 on: 3 Feb 2019, 09:49 pm »
Lumin X1 has both copper and optical fibre ethernet inputs.

(I haven't followed the Uptone EtherRegen developments in the past few months, so don't know if John Swenson's research has been published. Aside from leakage and noise measurements with ethernet, they were looking into jitter and phase noise with ethernet which is asynchronous and non-realtime. Yet they think something might be there.)

Anyways, James, what kind of experiments/setups are you doing?

Just doing blind listening test with the same music files on the 3 different setups.

James


zoom25

  • Full Member
  • Posts: 770
Re: DAS vs SAN vs NAS
« Reply #6 on: 3 Feb 2019, 10:48 pm »
Just doing blind listening test with the same music files on the 3 different setups.

James

So pull music files from ethernet via NAS, and a USB attached or internal hard drive/SSD for DAS, and something for SAN? What's the equipment setup like? I'm wondering if your setup or purpose is something different than what I'm imagining. Is this with MPD only or DLNA and Roon as well?

zoom25

  • Full Member
  • Posts: 770
Re: DAS vs SAN vs NAS
« Reply #7 on: 3 Feb 2019, 10:50 pm »
NAS, without question, as I describe in this post:

https://www.audiocircle.com/index.php?topic=161895.msg1725809#msg1725809

@Pundamilia ...

FC is pretty over-the-top for audio applications  :lol:

If you are suggesting this for optical isolation purposes, my post above discusses this (i.e. there are inexpensive, consumer routers that offer SPFs for optical gigabit ethernet, though it's unnecessary for this application - copper Ethernet is already galvanically isolated).

So what sounds best for MPD in your setup, your NAS or those Corsair flash drives?

James Tanner

  • Facilitator
  • Posts: 17527
  • The Demo is Everything!
    • http://www.bryston.com
Re: DAS vs SAN vs NAS
« Reply #8 on: 3 Feb 2019, 10:53 pm »
So pull music files from ethernet via NAS, and a USB attached or internal hard drive/SSD for DAS, and something for SAN? What's the equipment setup like? I'm wondering if your setup or purpose is something different than what I'm imagining. Is this with MPD only or DLNA and Roon as well?

Using the BDP 3 and the BDA3. Have Roon Server as well

James

zoom25

  • Full Member
  • Posts: 770
Re: DAS vs SAN vs NAS
« Reply #9 on: 4 Feb 2019, 09:04 pm »
Using the BDP 3 and the BDA3. Have Roon Server as well

James

Okay, so just the standard BDA-BDP hookup and see how to get the best sound from a music file via different playback methods / attached devices?

I'd be curious to hear about how many people are the Bryston listening team when developing these products and how many can reliably tell different playback methods apart? Also, among those that can tell them apart, do they all prefer the same method or is it a toss-up? Lastly, the one thing I want to know the most is does any difference in noise/jitter show up on the BDP outputs when using different playback methods or attached devices?

On the BDP, I've tried Roon Ready, DLNA Renderer with Audirvana Plus, MPD, Shairplay. I've done this with router acting as NAS, wifi USB adapter, flash drives, external portable and desktop spinning drives, linear power supplies for the drives and network gear, with or without Jitterbug, WAV and FLAC. They sure as hell don't all sound the same. Surely, it should show up on the measurements?

Anyways, the one playback method that stands out very clearly among the group is using the portable USB WD hard drive plugged into the BDP with MPD use. The few times I take the BDP into the large hall, the difference just stands out so clearly. I haven't found any other transport device or BDP playback method that quite fills the room like the portable hard drive. The image and layering just leaps from the speakers and touches you. The centre stage is more locked and as a result you get smoother vocals rather than having this smeared feeling. There is also this amount of air and ambience that gets lost with other method. I can only conclude that either this portable hard drive is adding something artificial that other methods are not, or the other methods are not as good at retaining the ambiance. All other playback method sound more 2D and flatter in comparison. If I had to convince someone that digital devices can sound different, I'd use the portable HDD as the example against whatever else.

sfraser

Re: DAS vs SAN vs NAS
« Reply #10 on: 6 Feb 2019, 04:24 pm »
The bit rate is extremely low for uncompressed audio delivery , assuming appropriate application buffering is available i would not expect an audible  difference between any of the 3  storage and delivery mechanisms.  The use of TCP ensures error free delivery over a network medium. However WiFi might prove to problematic in some instances.  I would suggest if you can  hear differences between the 3 content/delivery mechanisms, the difference might be at the individual  OS/application level and how it deals with USB VS Network  delivery and resources  etc.

I have been using a NAS for years now, with a rack mount 16 TB freeNAS ZFS  server hosting my audio/video content. I use a Raspberry Pi running Squeezebox Server  software with  CIFS mount to the aforementioned NAS server. I have a variety of squeezebox clients streaming content from the Pi server . Works and sounds  great, and the Pi  seems to be able  to deal with the 78K track library. I would move the Squeezebox Server off the Pi and over to virtual machine instance on the FreeNAS server and consolidate things, but I have yet to be successful with that endeavour.


zoom25

  • Full Member
  • Posts: 770
Re: DAS vs SAN vs NAS
« Reply #11 on: 13 May 2019, 08:50 pm »
So what were your conclusions, James?

Blueshound

  • Jr. Member
  • Posts: 9
Re: DAS vs SAN vs NAS
« Reply #12 on: 25 May 2019, 02:26 pm »
I've just received and installed a new BDP-3 into my system, which includes an SP3, BDA-3 and pair of Active Model T speakers driven by the BAX-1, (2) 7B3s & (2) 4B3s.  For sources the system includes a Bryston turntable and an Oppo UDP-203 multi-disc player.

Until now, I've been using the Oppo's network media player utility to stream digital audio files to the BDA-3, which has generally worked OK, but has become increasingly flaky and unreliable (for reasons I haven't fully sussed out, and are no longer important).

While I'm not surprised that the BDP-3 makes for a sound quality improvement over the Oppo as media player, I am surprised at the EXTENT of improvement in sound quality overall. Had I known the BDP-3 would make this significant a contribution, I would have ordered one a long time ago!

That's the background.  As to James' topic, I installed a 2TB hard drive (spinner, not SSD) into the BDP-3, and am just beginning to compare the same music files sourced from either this internal drive, or from a Western Digital NAS drive connected via a Netgear router. The router and NAS are in the same location as the system, no wireless needed.

My preliminary reaction is that I'm tending to prefer the NAS / router in pure sound quality terms over the internal drive.  This is based on only a couple of hours of listening, with a limited number of tracks.  I'll have more to say on this once I've auditioned enough material to know if what I've heard so far remains true over a much larger number of samples, and for a broader survey of genres and recording styles.

Having said that, I'll go out on the limb, however fragile it may be, and suggest that the files from the internal drive sound like very good hi-fi. The same files from the NAS drive sound like music.  I can put more nuance to how the two sound in comparison, but I'll save that until I know more.  It will be interesting to see if this impression holds up with extended listening.

Brian




zoom25

  • Full Member
  • Posts: 770
Re: DAS vs SAN vs NAS
« Reply #13 on: 25 May 2019, 08:26 pm »
I'll be interested in hearing your impressions after you've had due time. Also, if you are currently using FLAC or ALAC, go ahead and try uncompressed (FLAC uncompressed, AIFF, or WAV which is what I opt for) version of it to see if it makes a difference on the BDP-3 / your setup. Cheers and enjoy!

Blueshound

  • Jr. Member
  • Posts: 9
Re: DAS vs SAN vs NAS
« Reply #14 on: 25 May 2019, 10:57 pm »
@ Zoom - I've traditionally used uncompressed Flac, but a few months ago I started comparing those files to WAV. I have to say,  much of the time WAV sounds slightly better and more articulate, particularly in the bottom end.

I also very much like DSD, but haven't spent enough time comparing WAV with DSD to have much opinion there.

Brian
« Last Edit: 26 May 2019, 02:09 pm by Blueshound »

zoom25

  • Full Member
  • Posts: 770
Re: DAS vs SAN vs NAS
« Reply #15 on: 25 May 2019, 11:36 pm »
People already think it's crazy for there to be any difference between lossless vs. uncompressed with regards to CPU processing, much less between uncompressed formats that are the same size and perhaps only differ in endianness. The uncompressed formats are very close, but my 'gut' ends up with WAV. I wouldn't want to bet on it in a blind test, but there just seems to be something there that I can't shake off. AIFF especially stuck out for some reason.

I keep 2 backup copies of my libraries in FLAC compressed (highest setting 8) as my master archival for the smallest size and keeping all metadata. I used to have a secondary WAV library just for use with the BDP and MPD. I haven't done any DSD or MQA. Everything is in its native PCM format.

DarqueKnight

Re: DAS vs SAN vs NAS
« Reply #16 on: 1 Jun 2019, 04:15 am »
Hi Folks,

Doing some experiments - any thoughts on which of the following would provide the best fidelity?

DAS is a block device from a disk which is physically [directly] attached to the host machine.

SAN is a block device which is delivered over the network.

NAS is a filesystem delivered over the network.


I have used DAS (directly attached storage) and am currently using a NAS (network attached storage) with my BDP-3.

DAS scenarios:

DAS-1. 5 TB Seagate HDD installed inside BDP-3.
DAS-2. 5 TB Seagate HDD installed inside external hard drive enclosure and attached to BDP-3 with stock USB-3 cable and enclosure's stock switch mode power supply.
DAS-3. 5 TB Seagate HDD installed inside external hard drive enclosure and attached to BDP-3 with iFi Audio Mercury3.0 USB 3 cable and enclosure's stock switch mode power supply.
DAS-4. 5 TB Seagate HDD installed inside external hard drive enclosure and attached to BDP-3 with iFi Audio Mercury3.0 USB 3 cable and iFi Audio iPower switch mode power supply.
DAS-5. 5 TB Seagate HDD installed inside external hard drive enclosure and attached to BDP-3 with iFi Audio Mercury3.0 USB 3 cable and TeraDak DC-30 linear power supply.

DAS-1 and DAS-2 sounded identical. DAS-3 was sonically superior to DAS-1/DAS-2. DAS-4 was sonically superior to DAS-3. DAS-5 was sonically superior to DAS-4.

NAS scenarios:

NAS-1. Synology DS718+ NAS with two Western Digital 6 TB Red drives in RAID 1 (mirrored) configuration, with stock switch mode power supply.
NAS-2. Synology DS718+ NAS with two Western Digital 6 TB Red drives in RAID 1 (mirrored) configuration, with TeraDak DC-150VA linear power supply.

NAS-1 was sonically superior to DAS-5. NAS-2 was sonically superior to NAS-1.

Ethernet cables are ordinary computer grade Bluejeans Cat6a.

I have no plans for a SAN implementation and evaluation in the foreseeable future.

zoom25

  • Full Member
  • Posts: 770
Re: DAS vs SAN vs NAS
« Reply #17 on: 1 Jun 2019, 04:11 pm »
I have used DAS (directly attached storage) and am currently using a NAS (network attached storage) with my BDP-3.

DAS scenarios:

DAS-1. 5 TB Seagate HDD installed inside BDP-3.
DAS-2. 5 TB Seagate HDD installed inside external hard drive enclosure and attached to BDP-3 with stock USB-3 cable and enclosure's stock switch mode power supply.
DAS-3. 5 TB Seagate HDD installed inside external hard drive enclosure and attached to BDP-3 with iFi Audio Mercury3.0 USB 3 cable and enclosure's stock switch mode power supply.
DAS-4. 5 TB Seagate HDD installed inside external hard drive enclosure and attached to BDP-3 with iFi Audio Mercury3.0 USB 3 cable and iFi Audio iPower switch mode power supply.
DAS-5. 5 TB Seagate HDD installed inside external hard drive enclosure and attached to BDP-3 with iFi Audio Mercury3.0 USB 3 cable and TeraDak DC-30 linear power supply.

DAS-1 and DAS-2 sounded identical. DAS-3 was sonically superior to DAS-1/DAS-2. DAS-4 was sonically superior to DAS-3. DAS-5 was sonically superior to DAS-4.

NAS scenarios:

NAS-1. Synology DS718+ NAS with two Western Digital 6 TB Red drives in RAID 1 (mirrored) configuration, with stock switch mode power supply.
NAS-2. Synology DS718+ NAS with two Western Digital 6 TB Red drives in RAID 1 (mirrored) configuration, with TeraDak DC-150VA linear power supply.

NAS-1 was sonically superior to DAS-5. NAS-2 was sonically superior to NAS-1.

Ethernet cables are ordinary computer grade Bluejeans Cat6a.

I have no plans for a SAN implementation and evaluation in the foreseeable future.

Is the BDP and NAS connected to a switch or router, both to the same one or different one? Are you powering the router or switch with a LPS as well?

DarqueKnight

Re: DAS vs SAN vs NAS
« Reply #18 on: 1 Jun 2019, 05:01 pm »
Is the BDP and NAS connected to a switch or router, both to the same one or different one? Are you powering the router or switch with a LPS as well?

The BDP-3 and NAS are connected to the same Amped Wireless Athena EX WiFi range extender. The range extender is currently powered with its stock switch mode
power supply.When I get around to it, I will try the range extender with both an iFi Audio iPower switch mode power supply and a TeraDak DC-30 linear power supply.
There might also be some sonic benefit by connecting the BDP-3 and NAS to a switch and then connecting the switch to the range extender.

I also tried five different Ethernet cables, and I perceived no audible or tactile sensation difference among them: AudioQuest Cinnamon Cat7 ($100 - 5 feet),
Bluejeans Cable Cat6a ($18.50 - five feet), Amazon Cable Matters Cat6a ($3.40 - 5 feet), Synology stock Cat5e cable that came with the NAS ($0 - five feet),
Belkin Cat5e ($1.87 - five feet). I chose the Bluejeans Cat6a because of its build quality and because it was certified to pass specification.

DarqueKnight

Re: DAS vs SAN vs NAS
« Reply #19 on: 4 Jun 2019, 04:33 am »
***Update***

I had to improvise a little. The Amped Wireless Athena EX WiFi range extender uses a 12V/3A switch mode power supply and I do not have a compatible
linear power supply in house. I configured an old Amped Wireless SR10000 range extender to function as a five port unmanaged Ethernet switch by removing
the antennas and turning off the wireless functions.

I did not perceive a difference in sound quality or tactile sensation between the BDP-3 and DS718+ connected to the Athena EX (with its stock SMPS) and
the the BDP-3 and DS718+ connected to the SR10000 (with its stock SMPS).

The SR10000 runs on a 12V/1A SMPS, therefore I could compare its stock SMPS to the 12V/1.8A iPower SMPS and the TeraDak 12V/2A DC-30 LPS.
The most noticeable improvement was in the bass, with improved clarity, articulation, and tactile sensation. There was no difference in the size
and placement of images within the sound stage. There were also improvements in overall clarity and detail and a lowering of the noise floor as
evidenced by the music sounding apparently louder.

The iPower SMPS provided a little bit of improvement. The TeraDak LPS provided a moderate amount of improvement. The improvement with the TeraDak
LPS was similar to upgrading to a more resolving DAC.


Figure 1. Proof of concept: Better sound was obtained from the SR10000 with better power supplies.
Top, left to right: Amped Wireless Athena EX WiFi range extender, Amped Wireless SR10000 range extender,
TeraDak DC-30 linear power supply.


The SR10000 was plugged into one of the Athena EX's Ethernet ports with a Cable Matters Cat6 cable. The BDP-3 digital player
and DS718+ NAS were plugged into the two of the SR10000's Ethernet ports with Bluejeans certified Cat6a cables.

My digital front end consists of a Bryston BDP-3 digital player, Synology DS718+ NAS, dCS Debussy DAC, and dCS Puccini
word clock.


Figure 2. Left to right: SR10000's stock SMPS, iPower SMPS, TeraDak DC-30 LPS.

I have been keeping an eye on the development of Uptone Audio's EtherRegen low noise Ethernet switch. Unlike the current crop of stock Ethernet switches
that have been extensively modified for low noise computer audio applications, the EtherRegen switch was designed from the ground up for computer audio applications.
In the meantime, I have ordered a Cisco SG110D-08NA 8-port Ethernet switch ($36) to connect the BDP-3 digital player and DS718+ NAS. The SG110D comes with a
12V/1A SMPS. No other devices will be plugged into the SD110D  switch other than the BDP-3 and DS718+.
« Last Edit: 5 Jun 2019, 06:03 am by DarqueKnight »