Naim NAC-N 272 Streaming Preamplifier
Posted by: bicela on 15 February 2015
Dear All,
following the posts and rumors around for months I'm pleased to see that Naim is releasing this new model.
Some useful comments are already here: https://forums.naimaudio.com/topic/bristol?page=1
I'm really interested to buy it as upgrade of my Uniti2, with NAP250/2 that powering the Quad 2905 (these last two are already very well matched for my taste).
Maybe is not perfect as we wished, we can so post here any comments and hopefully get answer by Naim.
Let me to start:
- The analog input for phono (Stageline) is not powered? If so, is possibile to use SuperCap for powering the 272 and simultaneously the Stageline? Special cables are so needed?
- Has 272 a new DAC and/or improvements that are not present in previous streamers?
- When it will be shipped to costumers?
Warm regards, Maurizio
The major advantages of a pure MPD system against a UPnP-based streaming solution are, to my understanding, bandwidth savings and stability. I can control my MPD server from a client (for instance one running on an iPad) as I would do with an UPnP server. But an MPD system does not rely on transferring music data over Ethernet at replay time: you get the same flexibility of UPnP-based streaming solutions but you do not need wired connections to the router. This is, for many users living in old buildings, an important advantage.
?
MPD is a much lower-level implementation of a player / renderer queue, but there's very little functionally different once you press play, and you have the same issues with network bandwidth and sound quality over wired vs wireless that you have with anything else....
As for customizable tags, how would one consume them over an abstracted interface? Search by tag? Browse tree (add node for each tag)?
None of this is very well defined for network consumption but probably works fine if you are sitting in your bedroom/basement with a PC. The key here is social use - can someone else take your arbitrary tags and consume them to equeue and play music in the same house without being bound to a specific client (in this case, JRiver)?
If UPnP / DLNA servers are so substandard by your measure, why was mpd-dlna incorporated into MPD (from version 0.19 onward)? Must have been some sort of demand for that functionality....?
But both MPD and UPnP currently lack support for customizable tags. This is the common and most crucial limitation. Both MPD and UPnP are acceptable for occasional listening and mass consume of musical contents. They are fine for playing background music in a shop, for instance. But the view of a music collection they present to the user is hardly less primitive than the view presented by an iPod or an MP3 player. Even for classical music, let apart opera or more specific musical interests, they are essentially unusable.
"Unusable" is a bit strong IMO. I've been streaming to my DSs since 2009 and this would not have happened if UPnP (plus Linn's extensions for playlist support) would be unusable. Use a good server like the Minimserver. Its browse trees have never hindered me from listening to the album I intended to listen to, so job done.
Reading back on this topic I realise I may be an exception to the rule in soe respects.questions of the relevance and appropriateness of the 272 are fair enough, ultimately it's horses for courses as it were, some will choose to buy for perfectly sound (sic) reasons and others not and that's that really.
As for the filing and access of music I must, I realise be a flawed personality. A friend, a Windoze ser constantly rattles on about awful iTunes and I never really understand what irks him so much. Similarly here, the limitations of the app etc etc I find odd, no oddence meant or inferred.
Then I realise that, from my first childhood vinyl purchase, a still treasured Segovia recording, I have 'referenced' and 'indexed' all my music pretty much in my head. My over 1000 records and a similar number of CDs reside in a physical space that is somehow remembered by me (and I am notoriously forgetful!) and only when moving or re-organising my r
house do I land myself in trouble. I can never remember track names and very often forget album names and I could not recite more than a few artists or albums off the cuff. But the musical, emotional, genre links between everything is just their.
Attempts at adding tags in iTunes (and indeed photographs in Aperture) haven't failed because of any inadequacy of the application but rather because whatever words I use to reference my collections appear to fail to register in my head, when I look objectively at my physical music collection I realise that, at best, there is a vague grouping by genre and running a finger along the spines of the discs usually comes up with what I am thinking about. So my tagging is an odd physical one and once I have located the CD is a simple matter of locating it on my server and, if I fancy, adding it, or specific tracks to a playlist ( not that n-Stream makes that intuitive now). For those more rigidly organised than me I see the issues I just can't share the frustration as my feeble tagging methodology only frustrates me more.
where this does fall down is with downloads and I suppose this will become more pertinent as that method becomes more widespread - I ought to learn to take appropriately!
so, my own foibles, which in a way I hope are shared, aired, I presume that most of the issues raised here are essentially software ones? Some sort of clever box, maybe black, receiving some square waves steamed from another box somewhere in accordance to a set of instructions issued by some sort of software running on something tablet like? In which case, given sufficient memory and initiative it can be squirted into some at least of Naim's delightful black boxes?
Reading back on this topic I realise I may be an exception to the rule in some respects. Questions of the relevance and appropriateness of the 272 are fair enough, ultimately it's horses for courses as it were, some will choose to buy for perfectly sound (sic) reasons and others not and that's that really.
As for the filing and access of music I must, I realise be a flawed personality. A friend, a Windoze user, constantly rattles on about awful iTunes and I never really understand what irks him so much. Similarly here, the limitations of the app etc etc I find odd, no offence meant or inferred.
Then I realise that, from my first childhood vinyl purchase, a still treasured Segovia recording, I have 'referenced' and 'indexed' all my music pretty much in my head. My over 1000 records and a similar number of CDs reside in a physical space that is somehow remembered by me (and I am notoriously forgetful!) and only when moving or re-organising my house do I land myself in trouble. I can never remember track names and very often forget album names and I could not recite more than a few artists or albums off the cuff. But the musical, emotional, genre links between everything is just there.
Attempts at adding tags in iTunes (and indeed photographs in Aperture) haven't failed because of any inadequacy of the application but rather because whatever words I use to reference my collections appear to fail to register in my head, when I look objectively at my physical music collection I realise that, at best, there is a vague grouping by genre and running a finger along the spines of the discs usually comes up with what I am thinking about. So my tagging is an odd physical one and once I have located the CD is a simple matter of locating it on my server and, if I fancy, adding it, or specific tracks to a playlist (not that n-Stream makes that particularly intuitive now). For those more rigidly organised than me I see the issues I just can't share the frustration as my feeble tagging methodology only frustrates me more.
Where this does fall down is with downloads and I suppose this will become more pertinent as that method becomes more widespread - I ought to learn to tag appropriately!
so, my own foible aireds, which in a way I hope are shared, I presume that most of the issues raised here are essentially software ones? Some sort of clever box, maybe black, receiving some square waves steamed from another box somewhere in accordance to a set of instructions issued by some sort of software running on something tablet like? In which case, given sufficient memory and initiative it can be squirted into some at least of Naim's delightful black boxes and improved upon?
The major advantages of a pure MPD system against a UPnP-based streaming solution are, to my understanding, bandwidth savings and stability. I can control my MPD server from a client (for instance one running on an iPad) as I would do with an UPnP server. But an MPD system does not rely on transferring music data over Ethernet at replay time: you get the same flexibility of UPnP-based streaming solutions but you do not need wired connections to the router. This is, for many users living in old buildings, an important advantage.
?
MPD is a much lower-level implementation of a player / renderer queue, but there's very little functionally different once you press play, and you have the same issues with network bandwidth and sound quality over wired vs wireless that you have with anything else....
David,
in a UPnP-based Naim solution one typically has 3 processes running at replay + control time:
a) the UPnP server,
b) the UPnP client and
c) the control application.
The UPnP server runs on a uServe or on a NAS; the UPnP client runs on a Naim streamer and the control application runs on a iPad. When you press play, a significant amount of data is transfered from the UPnP server to the UPnP client over the LAN. This is hardly unavoidable because, by design, UPnP clients send requests to the server and receive data over LAN.
In a MPD-based solution, in contrast, one typically has only 2 processes running at replay + control time:
a) the MPD server and
b) the MPD client.
The MPD server runs on a microserver directly connected to a DAC-V1 or (via USB->SPDIF interface) to a nDAC; the MPD client runs on an iPad or whatever device of your choice. When you press play, no data (except for the "play" command itself, of course) need to be transferred through the LAN if the MPD server is configured to access local data.
Of course, one can implement MPD based solutions that require significant bandwidth (for instance, by setting up MPD to fetch data from a NAS) but one does not have to do so.
This is the crucial difference between MPD-based solutions and Naim's streaming design! Naim's approach needs significant bandwidth and, for high-resolution data, wired connections. Higher bandwidth requirements, and the presence of 3 active processes at control + replay time might explain some of the reliability problems that some users have experienced from time to time.
Originally Posted by DavidDever:
The major advantages of a pure MPD system against a UPnP-based streaming solution are, to my understanding, bandwidth savings and stability. I can control my MPD server from a client (for instance one running on an iPad) as I would do with an UPnP server. But an MPD system does not rely on transferring music data over Ethernet at replay time: you get the same flexibility of UPnP-based streaming solutions but you do not need wired connections to the router. This is, for many users living in old buildings, an important advantage.
?
As for customizable tags, how would one consume them over an abstracted interface? Search by tag? Browse tree (add node for each tag)?
Every decent MPD client supports searching and browsing the MPD database according to a fixed set of about 10 tags.
In a client search engine, users typically enter values for one or more of these (fixed) tags and receive a list of items fulfiling the searching criteria.
In a client browser, users typically selects one the (fixed) tags, say "Composer", and receive a list of values. They can then select another tag (or choose a value from the list) to refine their view of their music collection.This is also what most UPnP control applications and Naim's app support.
A server supporting customizable tags would read the custom tags at startup from a configuration file. If users have not provided any custom tag, it would use the default tags. Clients would request and receive the list of the currently supported tags after connect.
On the user side, nothing would change in the searching / browsing experience except for the fact that users would be able to search and browse their music collection according to their own criteria.
This is the natural way we usually organize search and browse our data collections, for instance, our digital photographs.
It is just for collections of musical data that software developers appear to have meant, they would know how to organize our music better than we know.
The major advantages of a pure MPD system against a UPnP-based streaming solution are, to my understanding, bandwidth savings and stability. I can control my MPD server from a client (for instance one running on an iPad) as I would do with an UPnP server. But an MPD system does not rely on transferring music data over Ethernet at replay time: you get the same flexibility of UPnP-based streaming solutions but you do not need wired connections to the router. This is, for many users living in old buildings, an important advantage.
?
None of this is very well defined for network consumption but probably works fine if you are sitting in your bedroom/basement with a PC. The key here is social use - can someone else take your arbitrary tags and consume them to equeue and play music in the same house without being bound to a specific client (in this case, JRiver)?
David, to be honest I do not understand how feature customization can relate to social use in any diminutive or problematic form: you select the number of clients which are allowed to connect to a given server. The server report the current set of supported tags to every client in much the same way as it reports the current number of albums available or whatever other piece of information stored in the database. Different client will search and browse according to (arbitrary subsets of) the current set of customizable tags in exactly the same way as they now search according to the fixed set of tags the software developers of their UPnP server and client have come up with. Best, nbpf
If UPnP / DLNA servers are so substandard by your measure, why was mpd-dlna incorporated into MPD (from version 0.19 onward)? Must have been some sort of demand for that functionality....?
But both MPD and UPnP currently lack support for customizable tags. This is the common and most crucial limitation. Both MPD and UPnP are acceptable for occasional listening and mass consume of musical contents. They are fine for playing background music in a shop, for instance. But the view of a music collection they present to the user is hardly less primitive than the view presented by an iPod or an MP3 player. Even for classical music, let apart opera or more specific musical interests, they are essentially unusable.
"Unusable" is a bit strong IMO. I've been streaming to my DSs since 2009 and this would not have happened if UPnP (plus Linn's extensions for playlist support) would be unusable. Use a good server like the Minimserver. Its browse trees have never hindered me from listening to the album I intended to listen to, so job done.
jfritzen, I agree "unusable" is a bit strong. And you are perfectly right, no UPnP server will prevent us from playing an album. But even Minimserver will prevent you from being able to search in your music library according to your own criteria.
You are free not to care, of course. But this does not change the facts. And not being able to organize our data according to our own criteria is not just dull. It is absolutely pointless. It does not buy us any advantage, just limitations. And I am sure none of us would have taken digital photography seriously if we had not been free to organize our pictures according to our own needs. Except maybe if we had been already Apple-brainwashed! Best, nbpf
@nbpf - after using Windoze for twenty years I switched to Mac. Imperfect and corporate for sure, but fewer problems in 10 years, less hardware issues too than in one year of microsoftband the army of invaders who constantly assailed it!
I am no Apple evangelist but I have not, for a moment missed Windows. My one remaining PC fund 8! Is that an operating system at all?!
You are free not to care, of course. But this does not change the facts. And not being able to organize our data according to our own criteria is not just dull. It is absolutely pointless. It does not buy us any advantage, just limitations. And I am sure none of us would have taken digital photography seriously if we had not been free to organize our pictures according to our own needs. Except maybe if we had been already Apple-brainwashed! Best, nbpf
Actually, EXIF is a good analogy - simply embedding arbitrary tags into the metadata field of an object does not get you scalability when it comes time to search over the set of available objects - on the other hand, if these tags correspond to an agreed-upon set of fields, it's much easier for the server to sift through and to return meaningful values to a suitable client that can consume them (or choose not to consume them). Remember here, we're not talking a server cluster, but a handheld device, so the tags have to be easily accessed and allow sorting of the ContentDirectory at the client, if they are not already handled at the server.
Standards are really important when it comes to interoperability, and without the ability to invite friends over to browse your library and add items to the running playlist, what's really the point?
You are free not to care, of course. But this does not change the facts. And not being able to organize our data according to our own criteria is not just dull. It is absolutely pointless. It does not buy us any advantage, just limitations. And I am sure none of us would have taken digital photography seriously if we had not been free to organize our pictures according to our own needs. Except maybe if we had been already Apple-brainwashed! Best, nbpf
Actually, EXIF is a good analogy - simply embedding arbitrary tags into the metadata field of an object does not get you scalability when it comes time to search over the set of available objects - on the other hand, if these tags correspond to an agreed-upon set of fields, it's much easier for the server to sift through and to return meaningful values to a suitable client that can consume them (or choose not to consume them). Remember here, we're not talking a server cluster, but a handheld device, so the tags have to be easily accessed and allow sorting of the ContentDirectory at the client, if they are not already handled at the server.
I completely agree with your observation but there is no contradiction between customizable tags and scalability. Scalability problems would arise if the UPnP (or MPD) server would have to parse the music collection to deduce the "current" set of tags. But the server does not do so in the current implementations and would not have to do so in implementations that support customizable tags. It is in fact exactly the other way round: users would be free to use only those tags which are important to them thereby reducing complexity.
Your post suggests that, beside bandwidth requirements and number of processes involved in controlling music replay, there might be another difference between approaches based on MPD and approaches based on UPnP: if I am not mistaken, MPD clients do not rely on locally stored content directories. According to your post, UPnP control applications do so.
You are free not to care, of course. But this does not change the facts. And not being able to organize our data according to our own criteria is not just dull. It is absolutely pointless. It does not buy us any advantage, just limitations. And I am sure none of us would have taken digital photography seriously if we had not been free to organize our pictures according to our own needs. Except maybe if we had been already Apple-brainwashed! Best, nbpf
Standards are really important when it comes to interoperability, and without the ability to invite friends over to browse your library and add items to the running playlist, what's really the point?
Nothing against standards but I fail to see how customizable tags could prevent anyone from being able to invite friends or for these to browse one's library. They would come over, fire up their app on their iPads or on whatever devices they bring with and add whatever they want to the running playlist just as they do now. What's the problem?
Put me, too, in the camp of being too dim to realize that I cannot do what I've been doing for the 3 years I've owned a UnitiServe (and a Naim Dac, followed by NDS and Qute). Me and Daunt, I guess 
(To add insult to injury, I too have gone 'all Apple' at home. Well my wife and kids prefer Apple too; I don't make them use it. Apple brainwashed -- all of us. I'm surprised we can even eek out a living.)
Count me in too. Not really getting this discussion, and I don't have any trouble finding what I want to play - the searching is fine for me. It also doesn't play records or feed the cat.
Count me in too. Not really getting this discussion, and I don't have any trouble finding what I want to play - the searching is fine for me. It also doesn't play records or feed the cat.
No worry, the problem with records and cats will be fixed by firmware upgrade! 
Best, nbpt
Contrary to what some of my family think I enjoy these discussions - and stoking the embers a bit too in order to move it along. I live in the world of broadcast where much of what we talk about is considered not just silly but mostly over dramatic, silly or merely self deluding snake oil. Only a few think beyond their little 19 inch rack world and I am just a nutter to them (for other reasons too I suspect!).
as a techie some of the terminology passes me by but the manner of the debate is interesting and the points raised, of course, are well intended and pertinent.
Our love of music has brought us here and, in the end we make do with what's on offer and try, as. Eat we can, to add a little to the debat.
After an evening of downloading from highresaudio.com I am thinking I should now be luxuriating with music and letting knowledgeable protagonists carry on and move the experience into more accessible territory.
Put me, too, in the camp of being too dim to realize that I cannot do what I've been doing for the 3 years I've owned a UnitiServe (and a Naim Dac, followed by NDS and Qute). Me and Daunt, I guess (To add insult to injury, I too have gone 'all Apple' at home. Well my wife and kids prefer Apple too; I don't make them use it. Apple brainwashed -- all of us. I'm surprised we can even eek out a living.)
"all Apple" is a bit too much but we too have iMac (and iPad and iPhone!) at home and manage to survive. If you are happy with your UPnP server, it's fine. You do not have to care about the posts.
But if, no matter for which reason, you will happen to come to the idea of tagging some of your music files with a "First Violin" tag (or with whatever other tag of your interest which does not belong to Apple's universe of tags) you will immediately realize that your server happily ignores your efforts.
This is both unnecessary and dull and is one of the reasons why some of us prefer iTunes, JRiver or Quodlibet based solutions to Naim UPnP based streaming solutions.
But both MPD and UPnP currently lack support for customizable tags.This is the common and most crucial limitation. Both MPD and UPnP are acceptable for occasional listening and mass consume of musical contents. They are fine for playing background music in a shop, for instance. But the view of a music collection they present to the user is hardly less primitive than the view presented by an iPod or an MP3 player. Even for classical music, let apart opera or more specific musical interests, they are essentially unusable.
Not necessarily true nbpf. The JRiver UPnP server does allow you to add your own tags, including ones based on expressions (even regular expressions), and will present views of your music based on those tags to a control point. Also it allows you to apply filters at every level. So if you wanted to see a tree in your control point of let's say all the Haydn String Quartets you've bought in the last two years but not listened to yet, split by artist, then you can have it, easily.
Very interesting likesmusic, thanks for the precisation! Do you have any experience in using the JRiver UPnP in conjunction with Naim's control app? Does the app honour JRiver's support for user-defined tags? If not, which control applications on which devices would you suggest? Unfortunately I have little to no experience with Windows systems. But if the JRiver UPnP server can act as the engine of a flexible client-server solution, I would certainly like to learn and test it. Many thanks, nbpf
hi nbpf - alas (or maybe not!) I don't have any experience of naim's control app. I've been put off naims UPnP offerings for several reason - the ghastliness of the standard UPnP model, naim's inability to play FLAC indiscriminably from WAV, and Paul Stephensons failure to substantiate his (to me ludicrous) claim than naim rips sound better than dBpoweramps bit-identical rips.
I've fiddled a bit with Linn's Kinsky - it certainly seems to see the custom views and tags I set up in JRiver's DNLA server, though as I have a USB DAC I tend to use JRemote to control JRiver from my iPad. It would be 2 minutes work for me to add a "First Violin" tag to the J River library fields and (assuming it was populated by magic) make appear in a DNLA view. Or use it to search on directly through J River. I doubt J River will do everything you want, but if you have a spare hour or two you can download it free and play about. I certainly agree with you about the general useless of the UPnP model. You tag your media, painstakingly. But you can't use the tags to search on? Gi's a job as Yosser used to say.
@likesmusic - can you expand on what you mean by, 'Naim's inability to play FLAC indiscriminately from WAV'? I too have suspicions about claims about better rips even after reading the Naim white paper on the subject. I get the advantage of using a selected transport and taking care to read many times to ensure data integrity. I use XLD occasionally and that app,I cation has only once thrown up an error in the log, so presumably my humble Mac Mini is doing as good a job as my UnitiServe?
@likesmusic - can you expand on what you mean by, 'Naim's inability to play FLAC indiscriminately from WAV'? I too have suspicions about claims about better rips even after reading the Naim white paper on the subject. I get the advantage of using a selected transport and taking care to read many times to ensure data integrity. I use XLD occasionally and that app,I cation has only once thrown up an error in the log, so presumably my humble Mac Mini is doing as good a job as my UnitiServe?
Just search this forum for WAV vs FLAC and you'll find many examples of the claim that uncompressed (WAV) sounds better than compressed (FLAC) on naim equipment. Here's just one, from the Naim NDX white paper:
"Uncompressed audio data will always give better results than compressed. Even lossless compression may not reproduce audio with equivalent quality to the uncompressed original as the processing required to uncompress the data increases the computational load. This raises the power supply noise floor, which detracts from the sound quality"
Maybe the reason that naim are taking so long to get a Tidal interface is that Stephenson is busy re-ripping all Tidal's music to WAV! On a Unitiserve no doubt!
Still don't get it as you can set the Unityserve to stream in original format surely and to rip in WAV or FLAC as you desire?
Still don't get it as you can set the Unityserve to stream in original format surely and to rip in WAV or FLAC as you desire?
My understanding is that you can setup a UPnP server accessing flac files to
1) directly send the files (a streamer-dac has then to unpack the files and convert them)
2) unpack the files and send the unpacked files (a streamer-dac has then to convert them)
and that many users adopt 2) (transcoding) to minimize computational requirements on the streamer-dac.
That's fair enough, in their words you choose which end of the bit of wire to unpack the data?
Either way you either have the choice of ripping straight to WAV or downloading WAV usually, or no choice at all which isn't really Naim's fault. And Naim downloads are available WAV or FLAC unlike Linn who appear to favour FLAC.
Reading through various forums the jury is still out as to the quality differences between the two though and maybe that is dependant on processor load, data throughput?
i keep everything WAV when I can and opt for WAV decoding at the UnitiServe but then I am only currently using a Uniti with 192 DACs so I am well behind the curve.
I've fiddled a bit with Linn's Kinsky - it certainly seems to see the custom views and tags I set up in JRiver's DNLA server, though as I have a USB DAC I tend to use JRemote to control JRiver from my iPad. It would be 2 minutes work for me to add a "First Violin" tag to the J River library fields and (assuming it was populated by magic) make appear in a DNLA view. Or use it to search on directly through J River. I doubt J River will do everything you want, but if you have a spare hour or two you can download it free and play about. I certainly agree with you about the general useless of the UPnP model. You tag your media, painstakingly. But you can't use the tags to search on? Gi's a job as Yosser used to say.
Thanks likesmusic, this is a crucial piece of evidence. If Kinsky can deal with custom tags, there is no reason why Naim's app should not be able do so. Of course, customizable tags would not eliminate the two major potential weaknesses of streaming approaches as compared to MPD based solutions: bandwidth requirements and number of interacting processes. But streaming would become a viable solution at least from the point of view of software usability. Thanks a lot, nbpf
If the SQ is best by giving the streamer/dac files in WAV format, then it makes sense to do that. The two ways this is achieved most commonly is either storing all files in WAV format or by having the NAS 'transcode' from FLAC to WAV. The streamer/dac will not 'know' which of the above options has been used - it simply gets a WAV file either way, so the SQ of these two options is 100% identical - no-one could tell, not even with a Statement amp.
The choice becomes simple preference, usability and so on. I prefer to store in FLAC and have the NAS transcode because I find the tagging better in FLAC, but there's no right or wrong to it.