SuperNAS and parts availability

hcl · Post by **hcl** » 2013-04-27 13:36

Split from "The Linnofil Super NAS!"

/Az

I think ThomasOK demonstrates my reservations against this (super-NAS) approach. It might, at any time, be possible to put together a very good NAS, soon to be impossible to build due to that most of the pieces is discontinued. This can go on forever. AClose to a waste of time as I see it. If we could put this time into something showing a more lasting result I would consider it to be alot more interesting!

Azazello · Post by **Azazello** » 2013-04-27 14:46

What do you mean by "lasting"? All equipment is continuously developed, why would this NAS need to be an exception?

hcl · Post by **hcl** » 2013-04-27 15:16

Azazello wrote:What do you mean by "lasting"? All equipment is continuously developed, why would this NAS need to be an exception?

What I mean is that alot of the experienses gained is and will be comtinuously lost due to the continueing obsolesens issues with computer stuff.

The long term value is that the efforts made by Linnofil and others will influence the deelopers to make a stronger effort to prevent the network side of the DS players from affecting performance of the players.

Nicolav · Post by **Nicolav** » 2013-04-28 22:28

hcl wrote:I think ThomasOK demonstrates my reservations against this (super-NAS) approach. It might, at any time, be possible to put together a very good NAS, soon to be impossible to build due to that most of the pieces is discontinued. This can go on forever. AClose to a waste of time as I see it. If we could put this time into something showing a more lasting result I would consider it to be alot more interesting!

Only the Asus E45M1-I Deluxe is discontinued, all other components are still available.
On the other hand all products sooner or later becomes not available on the market, including many excellent hifi products. But this does not prevent me to enjoy the benefits.
To remain in the NAS area, the excellent Readynas NV+ is discontinued too.
Of course the development and testing of components is expensive, but buying a qnap nas or any nas is even more expensive because you have the possibility to modify it if not in small part.
As for the development of a pc based nas, I don't think it's a problem: there will always be a better components that will replace the discontinued.
I see this an advantage not a limitation! IMHO of course.

ThomasOK · Post by **ThomasOK** » 2013-04-29 14:55

Nicolav wrote:
hcl wrote:I think ThomasOK demonstrates my reservations against this (super-NAS) approach. It might, at any time, be possible to put together a very good NAS, soon to be impossible to build due to that most of the pieces is discontinued. This can go on forever. AClose to a waste of time as I see it. If we could put this time into something showing a more lasting result I would consider it to be alot more interesting!
Only the Asus E45M1-I Deluxe is discontinued, all other components are still available.
On the other hand all products sooner or later becomes not available on the market, including many excellent hifi products. But this does not prevent me to enjoy the benefits.
To remain in the NAS area, the excellent Readynas NV+ is discontinued too.
Of course the development and testing of components is expensive, but buying a qnap nas or any nas is even more expensive because you have the possibility to modify it if not in small part.
As for the development of a pc based nas, I don't think it's a problem: there will always be a better components that will replace the discontinued.
I see this an advantage not a limitation! IMHO of course.

For the most part I agree with what you are saying which is part of why I decided to go ahead with this project (the other being the increasing rarity of some parts). However, I also see hcl's point. For one thing the Intel series 320 drives are also discontinued with the 600GB and 300GB becoming increasingly difficult to find. And they must have a good reputation outside the LS-NAS community as prices on the remaining ones are going up a lot!

Also while it is true that "progress" in the PC industry has always been rapid, leading to more powerful parts for the same or less money, this doesn't mean they are better parts for our musical purposes. The Intel 320 Series SSDs are a good example of this as they are at least a couple of generations old but the reports are that the newer replacements are musically inferior. It is definitely possible that - while searching for ever faster speeds, larger capacity, higher circuit density and more efficient manufacturing - the SSD companies (or other parts manufacturers) could continue to make less musical devices. Certainly I hope not. But I doubt there is anybody at Intel who cares about the musical quality of their SSDs, if they even believed such a thing existed.

This does actually leave us with the possibility that it might not be possible to make as musical an NAS in a couple of years as we can now and this would be a real shame.

hcl · Post by **hcl** » 2013-04-29 17:47

Music Lover wrote:
ThomasOK wrote: It is definitely possible that - while searching for ever faster speeds, larger capacity, higher circuit density and more efficient manufacturing - the SSD companies (or other parts manufacturers) could continue to make less musical devices.
I think the opposite!
Not maybe every generation, but over time the digital technology evolution will result in more musical devices

I can recognise arguments in both directions, but am not sure which will be pre-dominant. How do you figure it will develop?

ThomasOK · Post by **ThomasOK** » 2013-04-29 20:49

Music Lover wrote:
ThomasOK wrote: It is definitely possible that - while searching for ever faster speeds, larger capacity, higher circuit density and more efficient manufacturing - the SSD companies (or other parts manufacturers) could continue to make less musical devices.
I think the opposite!
Not maybe every generation, but over time the digital technology evolution will result in more musical devices

I certainly hope you are right, and I believe overall that improvements in digital technology will result in more musical devices as has already happened in newer versions of DS units.

That said, the one place I have the most concern is in the storage devices themselves. In the never-ending pursuit of storing ever more data in ever smaller spaces they are already using some pretty interesting techniques to read minuscule magnetic impulses. One wonders what effect the low-level error correction algorithms being used have on the musicality of these devices? Almost makes me want to resurect one of the old Seagate Barracuda 5.25" 91MB hard drives and see if it sounds any good. Maybe there would be a market for vintage hard drives as there is for old tube gear? ;-) "Anybody want to see my server with a stack of 20 5.25" drives? It gives me almost 2GB of storage of the highest quality! Got my 3 favorite CDs on there."

k_numigl · Post by **k_numigl** » 2013-04-29 20:54

hcl wrote: What I mean is that alot of the experienses gained is and will be comtinuously lost due to the continueing obsolesens issues with computer stuff.

I completely agree, except that no experience is made. The bad thing is that it is all just about trial and error, and no knowledge is accumulating at all. No one has any idea WHY this or that is better or worse. Nevertheless, it might be worthwhile to take the chance today, and benefit from the selection made by the original posters. It is certainly above average.

jiddu_k · Post by **jiddu_k** » 2013-04-30 07:37

k_numigl wrote: The bad thing is that it is all just about trial and error, and no knowledge is accumulating at all. No one has any idea WHY this or that is better or worse.

+1

Having stopped my LS-NAS project - never being able to surpass my ReayNAS NV+ musical performance - I already started selling the components on ebay.
Sealed and unsealed Intel 320series SSD in (120gb and 300gb) and the Seasonic X-460 platinum psu are still available if someone is interested.
I will post them in the Sale section.

Music Lover · Post by **Music Lover** » 2013-04-30 11:06

k_numigl wrote:
hcl wrote: What I mean is that alot of the experienses gained is and will be comtinuously lost due to the continueing obsolesens issues with computer stuff.
I completely agree, except that no experience is made. The bad thing is that it is all just about trial and error, and no knowledge is accumulating at all. No one has any idea WHY this or that is better or worse.

Totally disagree !

A huge amount of knowledge is gained, HUGE!
I compare this work with what Ivor did 40 years back when he started tweaking a turntable discovering a massive benefit in performance.
What Linnofil has done is a TRUE breakthrough in digital reproduction.

What you forget is that R/D is all about trial&error.
You seems to discuss Engineering, something else.
ok the LS NAS also has some engineering into it but it's same as designing an amp - it's mainly trial/error based on a few engineering concepts.

To be more specific:
1. we understand that NAS components (HW and SW) can affect the performance a LOT
2. we know the important HW/SW and the not so important HW/SW (in regards to performance)
3. we know whats working
4. whats not working
5. And as a result (IMPORTANT) we have a benchmark so all future components (SSD, RAM etc) can be compared against components in the current LS NAS

That is a HUGE amount of knowledge!!
And it will be the stepping stone for future performance enhancements.

With the knowledge gained (#1 to #4), Linnofil has been able to PREDICT performance of new components.
Also others have been able to predict and buy new components that enhanced the performance, example the PSU just mentioning one...

hcl · Post by **hcl** » 2013-05-01 07:37

Irrespective of the performance gained from using the LSNAS I do not think that there is much actual knowledge gained, except for the knowledge that virtually any part on the digital side of the DS affects the final performance (#1 and #5 above). That however and I agree is in itself an important achievement!

When all the current parts of the LSNAS is discontinued, how much of the experiences gaind can then be re-applied? There seem to be no theories (at all let alone validated) able to explain why the currently suggested parts makes for its good performance. How much re-usable knowledge actually gaind could thus be argued. Only time will tell. From what I understand there are only very fragmented assumptions about the reasons explaining what is affecting the NAS performance.

What you state regarding R&D is not entirely correct. Allthough many break throughs have emanated from unexpected results it is not at all e same as if R&D is built on trial and error experiences. Most R&D is, from what I know, conducted through building new experiments or designs on known and proven knowledge. Thus expanding the knowledge around a known base of proven knowledge.

DelNaja · Post by **DelNaja** » 2013-05-01 14:15

hcl wrote:Irrespective of the performance gained from using the LSNAS I do not think that there is much actual knowledge gained, except for the knowledge that virtually any part on the digital side of the DS affects the final performance (#1 and #5 above). That however and I agree is in itself an important achievement!

When all the current parts of the LSNAS is discontinued, how much of the experiences gaind can then be re-applied? There seem to be no theories (at all let alone validated) able to explain why the currently suggested parts makes for its good performance. How much re-usable knowledge actually gaind could thus be argued. Only time will tell. From what I understand there are only very fragmented assumptions about the reasons explaining what is affecting the NAS performance.

What you state regarding R&D is not entirely correct. Allthough many break throughs have emanated from unexpected results it is not at all e same as if R&D is built on trial and error experiences. Most R&D is, from what I know, conducted through building new experiments or designs on known and proven knowledge. Thus expanding the knowledge around a known base of proven knowledge.

My first post on this forum, hi to you all!

Although no answers to why a certain component works, or not, are gained, don't forget the Tune Method side of it. Because of Linnofil's work we now have a base to start from, a base of components to evaluate with newer ones as they are released. Number five in ML's list.

Is this not, at least in practical terms, a huge step forward? Even though we can't explain why, if the improvement can be heard, it is there. And it seems many have heard it.

hcl · Post by **hcl** » 2013-05-01 17:54

DelNaja wrote:
hcl wrote:Irrespective of the performance gained from using the LSNAS I do not think that there is much actual knowledge gained, except for the knowledge that virtually any part on the digital side of the DS affects the final performance (#1 and #5 above). That however and I agree is in itself an important achievement!

When all the current parts of the LSNAS is discontinued, how much of the experiences gaind can then be re-applied? There seem to be no theories (at all let alone validated) able to explain why the currently suggested parts makes for its good performance. How much re-usable knowledge actually gaind could thus be argued. Only time will tell. From what I understand there are only very fragmented assumptions about the reasons explaining what is affecting the NAS performance.

What you state regarding R&D is not entirely correct. Allthough many break throughs have emanated from unexpected results it is not at all e same as if R&D is built on trial and error experiences. Most R&D is, from what I know, conducted through building new experiments or designs on known and proven knowledge. Thus expanding the knowledge around a known base of proven knowledge.
My first post on this forum, hi to you all!

Although no answers to why a certain component works, or not, are gained, don't forget the Tune Method side of it. Because of Linnofil's work we now have a base to start from, a base of components to evaluate with newer ones as they are released. Number five in ML's list.

Is this not, at least in practical terms, a huge step forward? Even though we can't explain why, if the improvement can be heard, it is there. And it seems many have heard it.

Yes it is! thought I was clear on that point in the first part of my post!?

DelNaja · Post by **DelNaja** » 2013-05-01 18:12

hcl wrote:Yes it is! thought I was clear on that point in the first part of my post!?

Yes you were. I meant no offence.

I just thought that your outlook on this project seemed, to me, to be rather negative, despite all the practical aspects of it.

ThomasOK · Post by **ThomasOK** » 2013-05-01 19:13

I really do see both sides of this discussion. I certainly think it is great that this work has been done and that so many components that work well have been found by trial and error. No matter how much engineering knowledge you have it is still really trial and error that makes for the most musical sounding devices - whether digital or analog. So the work is certainly worthwhile and gives us a workable, high-quality starting point.

I also feel the frustration in finding the best solutions discovered so far becoming unavailable. It is obvious that the computer industry changes parts and models at a much faster rate than what we are used to with Hi-Fi gear and this can lead to the rapid obsolescence of products that may still have merit for our purposes. I think that this frustration was heightened by the rapidity of some major components becoming scarce just after the information about their quality was published. However, it should be noted that these same kind of frustrations also take place in the world of Hi-Fi and even analog equipment. Examples include those due to regulations such as ROHS meaning you can't build components with lead solder despite many feeling it is superior musically (I really don't want to get into the politics of how much sense this makes ecologically) and the requirement for front power switches adding complexity to units like the KK. There is also the replacement of items with musically inferior versions due to availability problems with components, manufacturing restraints, etc. such as a less musical power cable, RCA or tonearm cable replacing a previous better part.

So analog or digital the occasional setback is something we just have to live with and, fortunately, the overall musical quality of the best systems continues to move forward - at least in my opinion. But I do hold onto older, better performing parts where practical. My bigger frustration here is that, even though I have availed myself of the best known components for building a LS-NAS, I will not be able to offer to build these for customers at least until readily available replacements for the discontinued parts are found. So, Linnofil, I wouldn't be holding my breath waiting for those royalty checks to start rolling in! ;-) Still, I am really looking forward to hearing how this thing sounds once I've got it built (now if I can only get that Chinese seller to ship my case!!!).

Music Lover · Post by **Music Lover** » 2013-05-02 10:24

ThomasOK wrote: I think that this frustration was heightened by the rapidity of some major components becoming scarce just after the information about their quality was published.

To be fair, Linnofil published the info mid January and all of you got plenty of reminders to buy the stuff asap.
We that built a LS NAS all understand that computer parts going to be discontinued quite fast, and that info was passed to the forum.

ThomasOK wrote:
So analog or digital the occasional setback is something we just have to live with and, fortunately, the overall musical quality of the best systems continues to move forward - at least in my opinion.

Agree, that is my experience as well.
As you know Linn often run out of parts (CD12, lead soldering etc ) and despite this continuing improving the performance.
So I'm full of comfort for what the future going to bring!

ThomasOK wrote: My bigger frustration here is that, even though I have availed myself of the best known components for building a LS-NAS, I will not be able to offer to build these for customers at least until readily available replacements for the discontinued parts are found.

A simple solution...start testing :-)