ssd high end nas?

Harry posted:

Nothing wrong with conventional HDDs in a NAS that needs to be overcome by SSD. M2 SSD for fast caching video works, I can attest to that. I wouldn't dream of down sizing my HDDs and replacing them with more expensive SSDs. Don't see the point.

This is where I am.

What Bert says is interesting. I try to keep an open mind about stuff. I have had two ReadyNAS and three QNAP NASs (still have them all) with a variety of processors, RAM and make and capacity of HDD. I have never been able to hear a difference when playing music. Ditto Minimserver versus Asset on the same NAS or different ones.  The only bottle neck I've ever heard was my old HDX-SSD, running off a QNAP. Minimserver running on the QNAP, serving the NDS with the HDX taken out of the chain brought a big improvement. So I'm sure there are all kinds of factors making all kinds of contributions and maybe I need to hear some SSDs in one of my NASs. I expect that in time I will. Based on anecdotal personal experience I don't feel compelled to.

The spec of a NAS might make a difference if you want to run Roon on it. It's processor intensive (and likely to become more so in the future) and really needs the database to be on SSD for best performance.

Harry posted:

What Bert says is interesting. I try to keep an open mind about stuff. I have had two ReadyNAS and three QNAP NASs (still have them all) with a variety of processors, RAM and make and capacity of HDD. I have never been able to hear a difference when playing music. Ditto Minimserver versus Asset on the same NAS or different ones.  The only bottle neck I've ever heard was my old HDX-SSD, running off a QNAP. Minimserver running on the QNAP, serving the NDS with the HDX taken out of the chain brought a big improvement. So I'm sure there are all kinds of factors making all kinds of contributions and maybe I need to hear some SSDs in one of my NASs. I expect that in time I will. Based on anecdotal personal experience I don't feel compelled to.

you said that your hdx was out of the chain with your nds and that gave your progress in sound quality. So, if i understand well, you prefer your nas with nds vs nds/ hdx?  why?  it seems curious to me... but the world of streaming seems to have some mysteries....

Because the HDX sat on the music. It was a bottle neck, a veil, a limiting factor. I would not have believed it but the difference was obvious. All it was doing was sitting on the network, passing files onto the NDS. How could it have such a detrimental effect? Rhetorical in my case, because once removed it was never going back. Better it found a home that would appreciate it, which it did.

Harry posted:

Because the HDX sat on the music. It was a bottle neck, a veil, a limiting factor. 

One or two forumites have similar opinions with UnitiServe - others disagree of course.  A friend of mine changed US for NAS & not expecting anything was surprised, your description is similar to his.   

Same with NAS variations where opinions vary - e.g. Bert's post & your.  My experience is I've heard little or no difference in SQ between NAS brands.   My only NAS/NAS comparison was an evening at a Linn friends house where we compared Synology & two QNAP's,  the QUAP's were 2 & 4 bay & had different CPU power & RAM, & all were running different UPnP/DNLA software.      I have never detected any SQ change on my own NAS comparing UPnP/DLNA software.    

Its an interesting thread,  but it seems we are liable to not conclude anything except maybe the difference is subjective & small

I can only speak for myself Mike and will not presume to tell other people what they should be hearing. But the case against the HDX-SSD was unequivocal for me. Mighty fine multifunctional X level network player, iRadio source, server, CDP, Swiss Army Knife. But the evolving capabilities of out system appeared to tax it (or maybe it just needed a service?) and the money I got back for the sound gain helped fund the SL upgrade. I don't suppose I'll ever be fortunate enough to do that kind of upgrade again!

no noise more reliable no difference .  music servers in operation  for over 7 years/ none failed,  keep the away from listening area.   have safe back up . and forget them. Iam sure they are better not messed with at all. 

I am slowly acquiring more SSDs in devices. From experience so far the percentage failure rate is higher for SSD than for HDD. So I wouldn't be able to agree with the claim that they are more reliable although I don't have a big sample to go on.

I use an HDX as a server to an NDS and as a streamer in a separate system connected to a QNAP NAS. I have compared playing the same files on the NDS direct from the NAS and via the HDX and there is no difference that I can hear, so the HDX is certainly not a bottleneck or limiting factor.

The HDX sounds great in its own right and is very straightforward for ripping and storing music on the NAS. The big advantage for me is that I use it in a separate system(in an annexe) for  playing rips/downloads and internet radio. 

Which NAS based server did you use Jonn?

Harry posted:

Which NAS based server did you use Jonn?

It's a QNAP TS-253A

Nice NAS Jonn. When you played music on it back direct to NDS with the HDX out of the chain, I’d be interested to know what server you used to feed the NDS. Asset, Mininserver, Twonky, Plex, something else?

We originally had the HDX-HDD but I was always nervous about storing music in a proprietary enclosure. In those days, there was no off HDX backup function, just a mirrored HDD in the case. I listened to music stored on a NAS and compared it with the same files stored internally in the HDX. No difference. It made sense to me to store music on NAS and we had the HDX converted to SSD. Lower noise, less to go wrong, longer operational life (not).

The big difference occurred when the HDX failed (another SSD gone west!) and had to go back to the factory. We had always been more than happy with the HDX as an X level player and, post the arrival of the NDS, as a server/alternative streaming source/alternative iRadio source/CDP. It was a nice easy fit, it didn’t owe us anything and we had no reason to think the HDX was doing anything other than going about its business with sonic transparency.  But when the HDX-SSD packed up, we needed a server to tide us over until it could be fixed.

I installed Minimserver on the same QNAP TS-410 that contained our music files and was instantly bowled over by the clarity, communication, apparent extra detail and more realistic sounding texture and timbre.  I’ve kept Minimserver but also added Asset to another QNAP. We dip in and out of both. Equally striking was the speed and simplicity in installing Minimserver and configuring it. The app, in those days good old nStream picked it up instantly too.

I never heard a difference with:
HDX playing back from internal HDD versus HDX being fed by a NAS.
Type of NAS.
Type or size of HDD used in NAS.
NAS processor or RAM/cache capacity.
Asset versus Minimserver on any NAS or different NASs.

I can hear differences between:
Asset/Minimserver on a NAS compared with using HDX as a server – into NDS.
Different file formats, in particular WAV compared to FLAC.
Ethernet switch make/model
Ethernet cable.

It’s obviously going to depend on a lot of things. Although our streaming stuff and amps are similar we have different speakers, brains, ears, tastes and rooms.  I wonder if we were using the same server on the NAS when comparing HDX into NDS and NAS to NDS?

Our NAS based servers seem to run more smoothly than the HDX. Reindexing is instant and all maintenance, redundancy, upgrades, scaling etc. can be done without boxing anything up and sending it away. But this would be meaningless if such convenience was at the expense of musical enjoyment. The music always comes first. That's why our beloved HDX never came back home after it was fixed by Naim. And why I expect your's will stay firmly in place in your system. If it sounds good in your system, then it is. 

Mrs Wogan's lemon drizzle cake posted:

Anyone who claims to be able to detect any difference sound wise from material stored on one NAS or the other is frankly deluded. Audio woo at it's worst.

Afraid not, there are differences where the media server is on the NAS and denying it shows a naivity on what goes under the covers. Yes on a non revealing system you might not notice, but if one can hear differences between WAV and FLAC, Tidal vs local DLNA  FLAC or noisy mains on their Naim streamer then you are going to most likely hear the different sonic footprints of media servers. Over on the beta forum I posted some network TCP traces measuring average, minimum and peak inter frame network timings between different NAS media servers... and yes each timing 'profile' gave rise to a different sonic picture on a revealing system. Now Naim are aware of this apparently, and to some extent has been mitigated on the new Uniti series with the new streamer architecture where the media is transferred rapidly into memory and is played from memory rather than playing more or less realtime from a stream across the network.

Simon

Simon,    You are saying file formats have varying levels of quality (which is easily measurable and provable) so therefore "you are going to most likely hear the different sonic footprints of media servers" which is not measurable or provable.   

Taking you argument to it's logical conclusion,  one might say "the needle I use on my record player has an effect on sound quality, so therefore red record sleeves also improve sound quality, in comparison to green ones."  It's just a bogus comparison.  

Also i'm not sure what you mean by "TCP traces" ?  Do you mean a packet capture?  If it's digital, the data either gets through to your device or it doesn't.  TCP is designed to tolerate delays and data corruption using packet windowing (buffering) and acks.

Cheers!

Mrs Wogan's lemon drizzle cake posted:

Simon,    You are saying file formats have varying levels of quality (which is easily measurable and provable) so therefore "you are going to most likely hear the different sonic footprints of media servers" which is not measurable or provable.   

 

Simon said he was measuring different NAS media servers, not different file formats. I am assuming the files were the same, and the servers different, so any difference in the TCP trace can be deemed to be due to the different servers.

 I have a pair of 1TB SSDs in my Mac Mini. I can't comment on their reliablity as they have only been going 20 months or so, and I haven't attempted to compare any sound difference from HDDs, but I would not expect there to be because my renderer loads each entire track into RAM before playing. What I do like is their complete silence, and speed of access.

Reliability is likely to be different from normal computer use of storage devices because on a music store it is normal for files to simply be saved once, and virtually never deleted. Melco apparently modifies the algorithm designed to spread the 'wear' of SDDs presumably reflecting this difference, and claim a posiitive difference in sound in the N1Z, though as already suggested that is not necessarily due to the drives.

If we must contend that a NAS (used in its correct setting setting in a garage way away from your hifi) makes an audible difference to your sound, then we must also accept that every electrical thing in your house does.

Spending mega bucks on a NAS to improve sound is nonsense, as is expensive ethernet cables. If you are such an 'audiophile', turn off your fridge and freezer, the 'audible' difference will most likely be more  profound, and even with the food wastage less expensive.

We need to get a grip, people from the IT department are laughing at us, this used to be the preserve of the Psychology department.

garyi posted:

If we must contend that a NAS (used in its correct setting setting in a garage way away from your hifi) makes an audible difference to your sound, then we must also accept that every electrical thing in your house does.

Spending mega bucks on a NAS to improve sound is nonsense, as is expensive ethernet cables. If you are such an 'audiophile', turn off your fridge and freezer, the 'audible' difference will most likely be more  profound, and even with the food wastage less expensive.

We need to get a grip, people from the IT department are laughing at us, this used to be the preserve of the Psychology department.

That can be said about a number of things.

However there is a simple alternative to a  NAS as a music server which removes the network aspects completely (from the main music system if more than one), which is to use a combined store and renderer. I have been anticipating the Core to be excellent in this area, to make Naim a real challenger to the likes of Melco, Innuos Zenitha d MacMIni/Audirvana, though to date it seems to have had teething troubles that make its position rather unclear. However, the OP question of SSD vs HDD would remain the same.

garyi posted:

 

We need to get a grip, people from the IT department are laughing at us, this used to be the preserve of the Psychology department.

To your average IT guy it probably does look a bit daft. After all bits are bits aren't they ?  When you look at it as a complete system (with a digital to analogue stage at the back end) the theories of why there can be differences in the end result make a lot more sense.

The IT dept can happily go back to their Pizza...

garyi posted:

If we must contend that a NAS (used in its correct setting setting in a garage way away from your hifi) makes an audible difference to your sound, then we must also accept that every electrical thing in your house does.

Spending mega bucks on a NAS to improve sound is nonsense, as is expensive ethernet cables. If you are such an 'audiophile', turn off your fridge and freezer, the 'audible' difference will most likely be more  profound, and even with the food wastage less expensive.

We need to get a grip, people from the IT department are laughing at us, this used to be the preserve of the Psychology department.

Yup, I absolutely agree with this. Similarly humorous thread in one of the other forums regarding expensive kettle leads, the owners of which use appeals to magical thinking and pseudoscience to explain their expensive gear's mechanisms

Of course, we are all in control of what we spend (though subject sometimes to tighter control of purse strings by someone who might not share our interests), so we do not need to spend on "upgrades" unless we want, and we should all be in control of our ability to judge whether or not supposed upgrades actually make a positive difference to the sound of the music we listen to, to make that the deciding factor, and if no positive difference to our own ears, or if uncertain whether or not, there is no reason to spend (unless of course a goal is to have a system that either is composed solely of the alleged very best components or is the most expensive we can afford to buy, even if it makes no aural diffeence that we can hear.

i say 'should be in control'' because of course to some extent we can all be susceptible to persuasion by external suggestion, though the extent to which that is the case in practice can vary very considerably. Also relevant is that, in some cases at least, how likely an effect is likely to be evident can depend very much on the rest of the system, which includes the listening environment, and the type of music, as well as the quality of recording/mastering. And as mentioned, ultimately our hearing is a fundamental factor, which will depend on age and health or exposure effects, and on genetic differences, and on training. Aging people having with untrained hearing in some ways at lease may be the luckiest here, as they might be likely to find least differences, and so likely to be more readily content with their systems, saving money, time and frustration...

I have ferrite rings on the mains leads of all the appliances you mention (and quite a few more that you don't mention).  I also have ferrites on the Ethernet cables.  Collectively they do make quite a difference...

RFI is real folks.

Different mains leads can alter the effective impedance of the mains supply to the component at 50/60Hz and at RF.

Also, the effect of switching noise in the processor of the streamer is real.

So, there's no reason to assume that that the pattern of the reading and processing of Ethernet frames by the streamer cannot have any effect.

Harry posted:

Nice NAS Jonn. When you played music on it back direct to NDS with the HDX out of the chain, I’d be interested to know what server you used to feed the NDS. Asset, Mininserver, Twonky, Plex, something else?

We originally had the HDX-HDD but I was always nervous about storing music in a proprietary enclosure. In those days, there was no off HDX backup function, just a mirrored HDD in the case. I listened to music stored on a NAS and compared it with the same files stored internally in the HDX. No difference. It made sense to me to store music on NAS and we had the HDX converted to SSD. Lower noise, less to go wrong, longer operational life (not).

The big difference occurred when the HDX failed (another SSD gone west!) and had to go back to the factory. We had always been more than happy with the HDX as an X level player and, post the arrival of the NDS, as a server/alternative streaming source/alternative iRadio source/CDP. It was a nice easy fit, it didn’t owe us anything and we had no reason to think the HDX was doing anything other than going about its business with sonic transparency.  But when the HDX-SSD packed up, we needed a server to tide us over until it could be fixed.

I installed Minimserver on the same QNAP TS-410 that contained our music files and was instantly bowled over by the clarity, communication, apparent extra detail and more realistic sounding texture and timbre.  I’ve kept Minimserver but also added Asset to another QNAP. We dip in and out of both. Equally striking was the speed and simplicity in installing Minimserver and configuring it. The app, in those days good old nStream picked it up instantly too.

I never heard a difference with:
HDX playing back from internal HDD versus HDX being fed by a NAS.
Type of NAS.
Type or size of HDD used in NAS.
NAS processor or RAM/cache capacity.
Asset versus Minimserver on any NAS or different NASs.

I can hear differences between:
Asset/Minimserver on a NAS compared with using HDX as a server – into NDS.
Different file formats, in particular WAV compared to FLAC.
Ethernet switch make/model
Ethernet cable.

It’s obviously going to depend on a lot of things. Although our streaming stuff and amps are similar we have different speakers, brains, ears, tastes and rooms.  I wonder if we were using the same server on the NAS when comparing HDX into NDS and NAS to NDS?

Our NAS based servers seem to run more smoothly than the HDX. Reindexing is instant and all maintenance, redundancy, upgrades, scaling etc. can be done without boxing anything up and sending it away. But this would be meaningless if such convenience was at the expense of musical enjoyment. The music always comes first. That's why our beloved HDX never came back home after it was fixed by Naim. And why I expect your's will stay firmly in place in your system. If it sounds good in your system, then it is. 

Interesting findings. I have Minimserver installed on the NAS. Using a CISCO switch and Chord C Ethernet cable. The HDX was recently serviced by Naim and is running the latest software which may have made a difference.

I had another go today to see if I could detect any differences comparing files fed to the NDS using Minimserver direct from the NAS and from the NAS via the HDX. I used downloaded WAV files to make the comparison. 

The outcome was the same as before, virtually no difference between the two that I could detect. If anything the files via the HDX had slightly more body but nothing significant and I doubt that I could distinguish between the two "sources" in a blind comparison.

Sounds like my HDX needed a service. However, having taken it out of the loop, had I not been able to hear a difference compared to Minimserver running on our NAS, our decision would have been the same.  £5.6k versus a voluntary donation ? For the same result?

But each to our own as always.

When Jason came over to rebuild the system he took the HDX-SSD out of the rack. Best put elsewhere was the advice. So we did.