Post Reply 
 
Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Sound quality differences between NAS units
29-08-2016, 00:49
Post: #31
RE: Sound quality differences between NAS units
(22-08-2016 09:38)winxi Wrote:  
(21-08-2016 22:49)Peter@57m Wrote:  I have tried different NAS drives including different Melco models and have found Melco to be the best sounding and their more expensive model to sound better than the entry level.

I have swapped drives on visitors without telling them what was being tested and they have consistently found Melco better.

So I would dispute your statement there is no causal link to sound quality and different NAS drives.

Can you be sure that you didn't (unconsciously) influence your guests?
Have you already tried a blind test, were another person being in another room plays a track from a highend NAS or a regular NAS and you guess which NAS is used? To get a significant result, this experiment should be repeated at least 20 times for the same track. I'd be interested in the results of such a test.
Did somebody other do such a blind test?

Quote:Have you tried this yourself?

I've tried two different NAS and also streamed from a PC and I observed no difference in sound quality. But I had not the chance to try a highend NAS yet.

I can only chip in based on blind tests of the Melco vs QNAP servers. I spent a week pestering anyone that turned up at the house plus some educated listeners in a series of blind tests while I was considering the purchase. Both servers were outside of the listening room with no way of the listeners being able to see which device was connected. On almost every track tested the Melco was consistently chosen as the preferred version by all parties. The exceptions tended to be on one specific track (for the life of me I can't remember now). It may be clear from above but the Melco isn't simply offering an alternate NAS it also acts as the path to the network for the DS. I can't explain and neither did I expect the improvement, all I can say is that I was therefore convinced enough to buy it. I'm happy with that. I suspect, but don't know that it has more to do with the connectivity differences but I don't know that and TBH don't really care too much - I just prefer the sound. It has other benefits for some users.

The test was in my house not a lab - because that's where I listen Big Grin

hope that helps Big Grin
Find all posts by this user
Quote this message in a reply
29-08-2016, 19:35
Post: #32
RE: Sound quality differences between NAS units
I know I said I would sign off from this debate, but today I did a quick test with my wife as the listening subject. She could not care less about servers, file formats, transcoding etc. She does not even really know what a server is. She only wants the hi-fi to work. I played her the same Diana Krall track twice, once from the Synology, and once from the WD MyBook, both via Minimserver, and asked her whether she felt the track sounded the same each time. She said that they were not at all the same, and that on one version the vocal had a really harsh quality to it that was not present on the other version. The harsh version was the Synology. Yet in both cases the same 24/96 AIFF file was being streamed to a Linn Akurate DS/2 via a fully wired network. I'm convinced there is more to streaming SQ than we presently understand. In the ideal world, streamer manufacturers would build streamers to be essentially immune to choice of NAS, but at the moment, that does not seem to be the case.
Find all posts by this user
Quote this message in a reply
30-08-2016, 08:38 (This post was last modified: 30-08-2016 13:31 by winxi.)
Post: #33
RE: Sound quality differences between NAS units
Thank you for your efforts in doing these tests, but I hope you understand that such quick tests are not significant in a scientific/statistical sense.

A significant test would require a controlled setting, for example:

Preparation phase: There are two NAS types and two persons, the test person and the instructor. The test person would be ideally somebody who claims to hear a difference between NAS types. The test person chooses a track for which he/she claims to hear a clear difference and he/she is allowed to listen to this track as often as he/she likes, while knowing which NAS type is used.

Test phase: The test person and the instructor are in different rooms. In this phase it is very important that there is no kind of communication (verbal or non-verbal) from the instructor to the test person. Now the test starts:
1) The instructor flips a coin to choose a NAS, notes the result of the coin flip on a sheet of paper and starts playing the track from this NAS on a renderer in the room of the test person.
2) The test person listens to the track and tries to decide which NAS was used to play the track.
3) When the decision of the test person is ready, the test person notes the result of the listening test on a sheet of paper.
4) The test person informs the instructor that the decision is ready (claps with hands, for example).
5) The instructor stops playback and starts over at step 1)
These steps should be repeated N times, where N >= 20. The higher N is, the more significant the result will be.

Evaluation phase: To evaluate the results, the matches of the test person have to be counted. In terms of statistical testing, there would be two hypothesis: the null hypothesis H0 and the alternative H1.
H0: There is no audible difference between two NAS types (p = 0.5, where p is the probability to decide for the right NAS type in a single listening test).
H1: There is a audible difference (p > 0.5)
When, for example, N would be 20 and there were 15 matches of the test person, H0 could be rejected with a probability of error of pe ~ 2.07%. If there were more matches, pe would be even lower. pe is the probability that H0 is rejected although H0 is true: pe = P(H1 is accepted | H0 is true).
This means that 15 or more matches (for N = 20) would indicate a significant difference in sound quality between the two NAS types.

I'm aware that thoroughly carrying out such a test requires a lot of time and effort. I think, however, the result would be very interesting and if such a test showed a significant result, it may be able to convince even me that the NAS has an influence on the sound quality Wink
Find all posts by this user
Quote this message in a reply
30-08-2016, 14:05
Post: #34
RE: Sound quality differences between NAS units
Martin H and ac16161, following my earlier post I urge you to invest less than $100 (£80) in adding optical isolation to the last ethernet leg of your systems (and please, please do not use USB connectivity to your DACs). I'm sure that this will level the playing field and you find it hard to tell the difference between NAS and server.

You can find a full parts list in this blog by Andrew Everard, which I used to set up my optical isolation:
https://andreweverard.com/2015/06/08/hig...ded-fibre/

David

System: ALAC iTunes library on Synology DS412+ (running MinimServer) > Airport Extreme bridge > Optical isolation > dCS Network Bridge (controlled by Galaxy Tab S2 tablet running BubbleUPnP&Mosaic) > PS Audio DirectStream DAC > Primare A60 > Harbeth SHL5plus 40th Anniversary model
Find all posts by this user
Quote this message in a reply
02-09-2016, 07:42
Post: #35
RE: Sound quality differences between NAS units
hi David,
you're right !
A small optical "bridge" as you mentioned brings a nice improvement of SQ.
Anyhow, a complte optical network goes much further in SQ, by far over other current technical solutions.

Here are some topics about it :
in English on computeraudiophile.com forum :
http://www.computeraudiophile.com/f22-ne...ons-24641/
or in French the thread I created about it (I had to face quite large skepticism anyhow:coolSmile : http://www.homecinema-fr.com/forum/post1...p178699294
Rgds
Find all posts by this user
Quote this message in a reply
02-09-2016, 12:37
Post: #36
RE: Sound quality differences between NAS units
(30-08-2016 14:05)DavidL Wrote:  Martin H and ac16161, following my earlier post I urge you to invest less than $100 (£80) in adding optical isolation to the last ethernet leg of your systems (and please, please do not use USB connectivity to your DACs). I'm sure that this will level the playing field and you find it hard to tell the difference between NAS and server.

You can find a full parts list in this blog by Andrew Everard, which I used to set up my optical isolation:
https://andreweverard.com/2015/06/08/hig...ded-fibre/

David

Interesting and probably good timing. By coincidence I had a demo of the N1Z yesterday - I have the N1A currently. Again I was suitably sceptical about a meaningful difference between the two but there was enough for me to try this again. The environment here doesn't lend itself well to A vs B tests and I either need to temporarily change all of that in a further test or more likely go the shop for another session. The situation was slightly confused as we had just made some other changes (the point of the visit) and I need those to settle I think (in my head as much as the system).

I have for a while now surmised that this is likely a major element of the Melco's success but TBH don't know. I currently have a number of DS co-located with the servers that feed the various rooms in the house and they are connected via ethernet cable to the network/switches/router. I was wonder if it might make sense now to create their own mini network and bridge that to the main network via this type of setup. It may be a little over the top for anything other than the main DS though because the others are more casual listening in truth.

One thing that bothers me a little and I think it is briefly covered in that article is that the converters need power and by definition risk introducing a new problem as I try to fix another. TBH the article was interesting but I was a bit confused with the timeline and what looked like an incomplete shopping list?

It sounds as if you have been impressed by how this has helped you though?
Find all posts by this user
Quote this message in a reply
02-09-2016, 15:15
Post: #37
RE: Sound quality differences between NAS units
(02-09-2016 12:37)Martin H Wrote:  .....................

One thing that bothers me a little and I think it is briefly covered in that article is that the converters need power and by definition risk introducing a new problem as I try to fix another. TBH the article was interesting but I was a bit confused with the timeline and what looked like an incomplete shopping list?

It sounds as if you have been impressed by how this has helped you though?

You are correct about being bothered by power supplies for the ethernet wire-to-fibre converters. As sold the converters come with switch-mode power supplies and, at least, the one for the downstream converter (i.e. that nearest the renderer/DAC) should be replaced by a linear power supply. I replaced both. These are readily available but rather than quote some examples it's probably best for you to search locally.

The shopping list is complete (2 converters & a pair of 1 meter [say] fibre cables). As noted later in the blog you can simplify the setup if you only have 1 device you wish to isolate (your main DS), by putting the back-to-back pair of ethernet wire-to-fibre converters in the last leg of the ethernet connection to your renderer/DAC i.e. after your router and any switches in your network.
This Audiostream blog gives a useful pictorial view of the setup for optical isolation:
http://www.audiostream.com/content/barn-...OFbdYIS.97
6 & 8 are TP-LINK converters connected by a pair of fibre cables 7. Upstream of 6 the ethernet connects to modem/router, switch etc. Downstream of 8 the ethernet connects to your player (renderer/DAC).

Incidentally I picked up on Andrew Everard's blog after reading a comment he made in his review of a couple of Melco server / players in HiFi News "very good sound quality but this could be equalled or exceeded (with far less expense) by a NAS with an optimised network" or words to that effect
I did some brief (not really scientific) listening comparisons with and without the optical isolation in place, feeding my PS Audio Direct Stream DAC which has an internal renderer (Bridge II). I felt there was an improvement in SQ but it was not large, presumably because I also have a wireless bridge in my network between the NAS, router, switch, computers etc and the hifi gear.

System: ALAC iTunes library on Synology DS412+ (running MinimServer) > Airport Extreme bridge > Optical isolation > dCS Network Bridge (controlled by Galaxy Tab S2 tablet running BubbleUPnP&Mosaic) > PS Audio DirectStream DAC > Primare A60 > Harbeth SHL5plus 40th Anniversary model
Find all posts by this user
Quote this message in a reply
02-09-2016, 15:32
Post: #38
RE: Sound quality differences between NAS units
(02-09-2016 07:42)tgb Wrote:  hi David,
you're right !
A small optical "bridge" as you mentioned brings a nice improvement of SQ.
Anyhow, a complte optical network goes much further in SQ, by far over other current technical solutions.

Here are some topics about it :
in English on computeraudiophile.com forum :
http://www.computeraudiophile.com/f22-ne...ons-24641/
or in French the thread I created about it (I had to face quite large skepticism anyhow:coolSmile : http://www.homecinema-fr.com/forum/post1...p178699294
Rgds

Hi tgb
I'm familiar with the computeraudiophile thread and made a couple of contributions. I decided not to go down the route of switching all my wired ethernet to fibre as most of my network components are upstream of a wireless bridge (formed by 2 generation 6 Airport Extremes) so wire-borne noise is eliminated. Of concern to me however has been airborne noise (RF) from the downstream wireless bridge 'end' which also acts as an access point for control (from an Android tablet running BubbleUPnP). I got a noticeable improvement in SQ when I moved the Airport extreme 3 metres away from the hifi gear.
regards
David

System: ALAC iTunes library on Synology DS412+ (running MinimServer) > Airport Extreme bridge > Optical isolation > dCS Network Bridge (controlled by Galaxy Tab S2 tablet running BubbleUPnP&Mosaic) > PS Audio DirectStream DAC > Primare A60 > Harbeth SHL5plus 40th Anniversary model
Find all posts by this user
Quote this message in a reply
03-09-2016, 21:53
Post: #39
RE: Sound quality differences between NAS units
(30-08-2016 14:05)DavidL Wrote:  Martin H and ac16161, following my earlier post I urge you to invest less than $100 (£80) in adding optical isolation to the last ethernet leg of your systems (and please, please do not use USB connectivity to your DACs). I'm sure that this will level the playing field and you find it hard to tell the difference between NAS and server.

You can find a full parts list in this blog by Andrew Everard, which I used to set up my optical isolation:
https://andreweverard.com/2015/06/08/hig...ded-fibre/

David

Just bought a pair of MC200CMs and some fibre cable and at first listen it does indeed bring a very nice improvement.

Thanks David for highlighting Andrew's article.

Next purchase is a linear power supply for the downstream MC200CM. Don't know if this will make any improvement but I don't like switch mode wall warts near the KDSM, Solos etc.

I am using the fibre bridge between my GS108 switch and my Melco N1ZH

** No blind tests were used in the testing **

Peter
________________________________________________________
Linn Klimax DSM O&U & Solos, Wilson Benesch ACT C60s & Torus
Linn Selekt Edition DSM Organik, Wilson Benesch P1.0
Linn Klimax Renew DSM, Chord Mezzo 75, ProAc 1SCs
QNAP 559 ProII, Melco N1ZH, MinimServer, Linn, Lumin & Kazoo
Find all posts by this user
Quote this message in a reply
04-09-2016, 15:55
Post: #40
RE: Sound quality differences between NAS units
I've been fascinated by this discussion and have decided that for the relatively low cost this has got to be worth a go. There certainly seems to be some merit in the theory that electrically decoupling the nas from the renderer could have benefit and it cant be "snake oil" when so many respected opinions are categorically stating there is improvement to be had.

I've ordered two MC100cms and interconnect to decouple my Synology DS1511+ from my Cyrus XP2 QX. I ordered the MC100cms as my renderer doesn't support gigabit, and there obviously wont be an issue even streaming 192/24 flacs as wav files.

I'm hoping that someone can actually spec the linear mode power supplies for the TP Link adapters and a supplier as I have no idea what's required, volts amps or even the plug on the end!

Streaming from a Synology 1511+ to a Cyrus Stream XP2 QX, 2x Roberts 93i, A Revo Superconnect, a Mede8er 600 3D and a few Squeezeboxes thrown in!
Find all posts by this user
Quote this message in a reply
Post Reply 


Forum Jump:


User(s) browsing this thread: 1 Guest(s)