Surely it's only digits?

Hurray, hurrah, another Melco thread. Just what we all need. I'm sure it's marvellous!!

Makes me pine away for the Hugo days.  

Ah yes...Hugo...I remember him well. Whatever became of him?

I am still enjoying my UnitiServe and a NAS

OK, I'll bite. Could you describe the 'serious gain'? Does it come at any cost to engagement with the music? (This latter aspect is why I've held on to my UnitiServe.)  

Thanks

Jan

Jan, 'serious gain' means more detail, lower noise, more natural instrument sounds, all that goes with an improvement at this level. Engagement with the music is immediately increased; you know that your head will be spontaneously filled with tunes for months to come.

David

David - you wrote that the biggest gain seems to come from connecting Melco directly to NDS.

That begs a question: was your network implemented correctly before (as this is the only way to connect US to a Naim streamer)?

I don't understand your question. Adam. Please explain.

David

Essentialy the optimal way to connect a UnitiServe is via a LAN cable to a network switch. NDS is then connected to the same switch. ISP router is connected to the switch, providing open internet access.

How is your Melco connected to the network, to which your NDS is connected?

David O'Higgins posted:

Jan, 'serious gain' means more detail, lower noise, more natural instrument sounds, all that goes with an improvement at this level. Engagement with the music is immediately increased; you know that your head will be spontaneously filled with tunes for months to come.

David

That sounds like a good space to be in. I'm interested to read how your listening evolves over time.

Jan

I just don't know why people say they are suprised.. I have written about this severeal times on here and I start to see references elsewhere.. The UPnP media SQ in the streamer can be determined to some extent by the interframe timing variation from the server over the network. Perhaps if it if I call it server frame jitter it might mean more to peeps.

The net effect is that different UPnP server platforms and software can sound different in apparent 'atmosphere' and 'realism'.. certainly notice this with Naim streamers with resolving DACs like NDS or NDX/Hugo.

Hungryhalibut posted:

Hurray, hurrah, another Melco thread. Just what we all need. I'm sure it's marvellous!!

Don't mention the Melco!

I mentioned it once but I think I got away with it.

David O'Higgins posted:

Well maybe, but yesterday I imported my backed up music library from a USB 3.0 external drive into a Melco N1A. The library consists of about 2,100 CDs ripped via my Userve, together with 400+ 24 bit downloads. The import took the best part of an uneventful 24 hours. Leaving aside a lot of problems which I hope to cure by using Minimserver on the Melco instead of Twonky, I am astounded at the improvement in SQ. I had been led to expect something much more modest, but this is serious gain territory, for what (in our mad world!) is a very modest outlay of less than €2,000. The biggest gain seems to  come from  the direct Ethernet connection of the Melco to the NDS. 

So now, 36,000 + new tracks ! Where to start? How to go to bed ?....... 

David

Yes.

Adam Zielinski posted:

Essentialy the optimal way to connect a UnitiServe is via a LAN cable to a network switch. NDS is then connected to the same switch. ISP router is connected to the switch, providing open internet access.

How is your Melco connected to the network, to which your NDS is connected?

Adam, that is precisely how my US and NDS are connected to my network. The twist with the Melco is that it offers a direct connection between it and the NDS (i.e. NDS is no longer connected directly to the switch) and it sounds better.

David O'Higgins posted:

Adam Zielinski posted:

Essentialy the optimal way to connect a UnitiServe is via a LAN cable to a network switch. NDS is then connected to the same switch. ISP router is connected to the switch, providing open internet access.

How is your Melco connected to the network, to which your NDS is connected?

Adam, that is precisely how my US and NDS are connected to my network. The twist with the Melco is that it offers a direct connection between it and the NDS (i.e. NDS is no longer connected directly to the switch) and it sounds better.

Sounds like you're running the whole setup with no internet connection? If so that must be a bit of a drawback at times. 

If there is a placebo or herd effect, I'm pretty sure it will not be able to tell whether it is Melco or Naim lettering on the box.

Does it beat the sound you get when reading WAVE files directly from a USB stick shoved into the front panel?
(N.B. remove any network cable when doing the test).

lowdean posted:

If there is a placebo or herd effect, I'm pretty sure it will not be able to tell whether it is Melco or Naim lettering on the box.

1 to 1 double blind testing  -  the best way to distinguish a herd effect from a heard effect!

ChrisSU posted:

David O'Higgins posted:

Adam Zielinski posted:

Essentialy the optimal way to connect a UnitiServe is via a LAN cable to a network switch. NDS is then connected to the same switch. ISP router is connected to the switch, providing open internet access.

How is your Melco connected to the network, to which your NDS is connected?

Adam, that is precisely how my US and NDS are connected to my network. The twist with the Melco is that it offers a direct connection between it and the NDS (i.e. NDS is no longer connected directly to the switch) and it sounds better.

Sounds like you're running the whole setup with no internet connection? If so that must be a bit of a drawback at times. 

Everything is connected, just that the NDS is connected via a dedicated port in the Melco, rather than dirctly to the shared switch.

Huge posted:

Does it beat the sound you get when reading WAVE files directly from a USB stick shoved into the front panel?
(N.B. remove any network cable when doing the test).

Don't know, but that wouldn't be a viable way of listening to a library this size.

David O'Higgins posted:

Huge posted:

Does it beat the sound you get when reading WAVE files directly from a USB stick shoved into the front panel?
(N.B. remove any network cable when doing the test).

Don't know, but that wouldn't be a viable way of listening to a library this size.

It's useful reference point to use as a test to see how good your network connection is from an audio perspective.

I wasn't suggesting that you try to fit 36,000 tracks on a memory stick!
(Even if some fleabay scammers are selling what they claim are 1TB+ capacity sticks sold at knock down prices - invariably they're small capacity devices frigged to appear much larger).

Interesting discussions and very timely for me. My logitech streamer has finally died and so i got a replacement, not a Naim (too expensive at the moment) but a Moon Mind. My primary digital system is a unitiserve SSD (music on a Qnap NAS) into an NDac with an XPS2 on it. So before I put the Moon Mind in the room to replace the logitech I thought I would take a listen to the Moon against the unitiserve. Both are going through the Ndac.

Not really any SQ difference. Using the Moon I could serve up the music either from the unitiserve or the logitech media server (which pulls an identical copy of the music from an external drive on my old mac pro which I'm using as a server) as control points and I couldn't perceive any difference in SQ between them. I know it's weird to have an identical copy but I like to have an extra backup!

However the difference in cost of the Moon Mind and a Naim streamer is quite large and I would expect the Naim streamer to sound better.

Now I thought lets try using a UPnP on the Qnap. So I tried the latest version of Twonky and also Minimserver. Also I installed Minimserver on the Mac Pro to see if that would affect sound quality. Didn't really get very far as each of these systems just didn't work very well. Files not picked up, stuff in the wrong order, generally quite a mess. That's when I realized how good the unitiserve is for indexing music collections and also I suppose the Logitech music server as well. Thank god I didn't get a streamer first, I think I would have given up on it.

Tim

Simon-in-Suffolk posted:

I just don't know why people say they are suprised.. I have written about this severeal times on here and I start to see references elsewhere.. The UPnP media SQ in the streamer can be determined to some extent by the interframe timing variation from the server over the network. Perhaps if it if I call it server frame jitter it might mean more to peeps.

The net effect is that different UPnP server platforms and software can sound different in apparent 'atmosphere' and 'realism'.. certainly notice this with Naim streamers with resolving DACs like NDS or NDX/Hugo.

"Surprised" vs. "skeptical" perhaps more accurate?  The history of hi fi is replete with claims that certain technology, from the green ink pen up, provides sonic benefits.  So how do we, the pool of hi fi consumers, figure out which is which?  When does "psychoacoustic" become just "acoustic" with respect to hardware changes and resultant acoustic changes?  Only with blind testing???

We're obviously all entitled to our own reality, and decades now of hi fi talk and message board chatter has been consumed with "if you think it's better and you want to spend your money on it, go ahead" type advice.  As well as, "if you don't think this is better you're a dunce" type advice.

We all as individuals put various technologies in different "buckets" - perhaps 3 of them are "utter silliness," "I don't really know if it's better or not," and "it's definitely better. So when does any given piece of technology move from the "skeptical" to the "accepted" category?  When does it move from "utter silliness" to "I dont really know" to "definitely?"  No generalizations, I am sure. 

Don't take it personally, Simon.  Yes you have written about "interframe timing variation."  Does that make it so?  Does that mean that all humans with decent hearing can and should be able to discern a difference in the sound of servers based on this parameter?  I don't know!  I'm sure you believe what you write and I mean that most respectfully . . . but surely people look for more than "because Simon says."  And even assuming, arguendo, that "interframe timing variation" can affect sound quality, is that WHY a Melco server connected via its dedicated ethernet port to an NDS results in some purchasers reporting that it sounds "better?"  Is the Melco's timing in this way different in a way that leads to objective improvement?

Me -- I'm not willing to do the experiment for myself at home, because I couldn't hear a difference when I moved from the UnitiServe to QNAP running Asset.  But I absolutely know that I could buy a Melco server, connect it as described in this thread, and convince myself and others that I heard an improvement

Bart - interesting post. I haven't measured a Melco yet, but if i was designing it I know how I would want to do it for optimum SQ. The only point that slightly clouds this is that the effects in terms of SQ do seem to vary from firmware version to version with Naim equipment. But having said almost right back at the start of the NDX when I bought mine i noticed a difference difference between media servers - and I didn't know why. Some research into an issue at work gave some insight of what I might need to look for in these issues at home - so I  set up a test harness around my NDX, wrote some macros and set an hypothesis  and viola the measurements correlated to my subjective SQ assessment. I was well chuffed.

I am a firm believer there is a sound engineering reason behind every SQ variation we hear and it doesn't always have to cost a lot of money- and its nice to demonstrate this - even more so when its not yet main stream web info. The term 'RFI' is too often used as catch all for differences in digital audio - and at the level we are using here that is simply too simplistic.

Simon

Bart posted:

Makes me pine away for the Hugo days.  

And the Lavry days.