Introducing Uniti, our revolutionary new streaming platform.

iliria posted:
manicm posted:
Solid Air posted:
manicm posted:
iliria posted:
Pcd posted:
You do not need hype to sell good products they sell themselves ?

A very generalistic and misguided statement in my opinion. If good products sold themselves then Apple, Mercedes and a ton of other companies would not need to advertise at all. Apple very likely sell nearly half of their products because of hype.

 

Actually neither Ferrari, nor Rolls Royce advertise. And Linn/Naim/Cyrus advertise very little. In fact they don't need to as the websites act as advertisements.

I'm sorry, but that is a silly statement. Just type 'Ferrari advert' or 'Rolls Royce advert' into Google Images and you'll see why. Very nearly all successful businesses spend money on marketing, and those that don't have very specialist reasons not to. The term 'hype' is contentious as it implies over-blown claims, and I certainly don't advocate those, but marketing is essential. The term 'advertising' is elastic and falling out of fashion in the modern era, but websites, PR and, yes, even this forum are not free - it's all marketing. 

No it's not a silly statement, why don't you ask Linn how much they spend on traditional marketing - very little actually. They actually cannot afford to - their profit margins are really not that high. They instead spend on events like listening sessions and demos at selected locations. 

You may be a marketing graduate but you come across as a bit naive. Linn and Naim's biggest marketing spend remains their websites. Reviews also bring them product awareness. They're completely different from Apple.

Ferrari and Rolls Royce are legends and dont need to advertise as much. And still, they advertise quite a lot. By comparing Naim to Ferrari and RR  and then say that Naim and Linn are completely different from Apple you are contradicting yourself. I assume you work for both Naim and Linn based on the fact that you are making a certain statement about their marketing expenditure?

Rolls Royce is affordable by the mega rich and the mega rich are well aware of the brand. Middle and lower classes have no hope of buying a Rolls. Thus Rolls' advertising needs may not be as high as those of Naim.

Naim is aimed at the middle and lower classes especially considering that nowadays it is quite easy to get a loan or CC for £3k. Therefore the advertising needs to be more intensive, if the company is to expand customer base.

Listening demos and sessions is not as effective at reaching new customers. Rather It may prove more beneficial to convince existing customers to upgrade. Whereas an indepth review of the product which has been carried out by someone impartial (or at least someone that appears to be impartial and with no vested interests) has more "convincing power". In my opinion that is. I would put more weight on a review carried out by someone impartial rather than a review from an authorised reseller of the product.

Naim is patently not aimed at only the lower to middle classes. Their top end streamers, for example, cost around 3k +, then you need to add amplification and speakers to match and you're nudging a 15-20k system. Naim and Linn are over 40 years old and people know who they are. I would agree Naim would add more marketing for products like Muso, but their top end products will advertise through demos, sessions, exhibitions.

Crabby posted:
Hungryhalibut posted:

Who's that then?

Ha ha ha thats you :-) 

just trying to bring back the discussion to the actual units and sound. Looks like I failed !!! 

 

Keep trying! I'm far more interested in that than listening to people banging on about corporate marketing.

ChrisSU posted:
Crabby posted:
Hungryhalibut posted:

Who's that then?

Ha ha ha thats you :-) 

just trying to bring back the discussion to the actual units and sound. Looks like I failed !!! 

 

Keep trying! I'm far more interested in that than listening to people banging on about corporate marketing.

Thanks for the back up 

HH, do us the honour of sharing the feedback of your listening session today. 

As a heads up, HH took away my doubts. He was impressed by build quality and how the Atom drove easily a pair of £6k supra. His feedback was very positive. 

I was debating keeping my uq2 but his feedback   made me forget that illico presto 

hope more people with actual listening sessions to come forward 

 

 

jon honeyball posted:

It isn't finished either in firmware in the device or software on the controller. Commenting on sound quality especially given the former is unwise at present. 

There's me thinking I was the only one who left things last minute.

Penarth Blues posted:
Phil Harris posted:
yeti42 posted:

I was being slightly tongue in cheek with my coice of word there Phil.

I'm just very conscious that things that are posted here often get taken and built upon...

By extrapolation it would probably end up at "Phil at Naim said never to use WD Reds because they give errors all the time".

Cheers

Phil

I've already noted that I should never buy a WD Red drive and I'll let all my friends know too. Thanks Phil!

300px-Paris_Tuileries_Garden_Facepalm_statue

Phil

Attachments

Photos (1)
iliria posted:
jon honeyball posted:

It isn't finished either in firmware in the device or software on the controller. Commenting on sound quality especially given the former is unwise at present. 

There's me thinking I was the only one who left things last minute.

Everything is to be adjusted up to the last minute, of course. Naim is historically notorious for adjusting and adjusting and adjusting -- its the only way to get quality.

Hungryhalibut posted:

Indeed. It's total bollocks. I believed it though, when I bought my UnitiServe. Then you discover that both the Serve and dbpoweramp look at the perfectrip database, just the same. 

Hmm, that's weird. 2 years ago I tested this to check my sanity and the difference was substantial. I made several tests to verify. 

I ripped some music using US and got it onto a USB stick. Ripped same music using optimized dbpoweramp on my home PC and for devil-may-care same again with itunes. Placed all on same stick. Played the tracks via my BD player through 282/hi/250.2. The US ones were consistently better. Played the tracks for SWMBO and asked if a) there was a difference and b) if it was valuable. She much preferred the US tracks (without knowing what she was listening to) and she thought they were valuable though the US was a bit of a tall one. The dbpoweramp vs itunes was a no-brainer.

Back in the shop, I tried the same rips using our NAS drive's own server (Twonky) vs our US as a server. Significant difference and did a blind test on my colleague with same results. Again the US sounded better serving the same track as twonky via the same client (might have been NDX, might have been NDS, I forget) into our large-ish Naim system at the time. So I don't understand your experience. 

Now, whether or not the US (or Core) is value for money is another question entirely which depends on how far up the ladder you are and your future aspirations vis-a-vis your system. 

But your experience has me surprised, and a bit baffled.

Frank.

Frank Abela posted:
Hungryhalibut posted:

Indeed. It's total bollocks. I believed it though, when I bought my UnitiServe. Then you discover that both the Serve and dbpoweramp look at the perfectrip database, just the same. 

So I don't understand your experience. 

Now, whether or not the US (or Core) is value for money is another question entirely which depends on how far up the ladder you are and your future aspirations vis-a-vis your system. 

But your experience has me surprised, and a bit baffled.

Frank.

I have never understood this argument about differences in lossless files. A lossless file is just that, nothing added and nothing taken away. Am I right? If so then how can one collection of 1s and 0s sound different from the same collection of 1s and 0s?

iliria posted:
Frank Abela posted:
Hungryhalibut posted:

Indeed. It's total bollocks. I believed it though, when I bought my UnitiServe. Then you discover that both the Serve and dbpoweramp look at the perfectrip database, just the same. 

So I don't understand your experience. 

Now, whether or not the US (or Core) is value for money is another question entirely which depends on how far up the ladder you are and your future aspirations vis-a-vis your system. 

But your experience has me surprised, and a bit baffled.

Frank.

I have never understood this argument about differences in lossless files. A lossless file is just that, nothing added and nothing taken away. Am I right? If so then how can one collection of 1s and 0s sound different from the same collection of 1s and 0s?

There's a lot of similarity between homeopathic medicine and the nonsense talked about a bit not being a bit imho. It's fun arguing about both of them though!

best

David

(but I'm not an evolutionary biologist... )

iliria posted:
Frank Abela posted:
Hungryhalibut posted:

Indeed. It's total bollocks. I believed it though, when I bought my UnitiServe. Then you discover that both the Serve and dbpoweramp look at the perfectrip database, just the same. 

So I don't understand your experience. 

Now, whether or not the US (or Core) is value for money is another question entirely which depends on how far up the ladder you are and your future aspirations vis-a-vis your system. 

But your experience has me surprised, and a bit baffled.

Frank.

I have never understood this argument about differences in lossless files. A lossless file is just that, nothing added and nothing taken away. Am I right? If so then how can one collection of 1s and 0s sound different from the same collection of 1s and 0s?

But what about a lossless file with metadata and one without...? There was interesting piece in one of the mags fairly recently about this as well as transcoding in and out of lossless formats. IIRC, I think the gist of it was that a pure WAV rip without metadata and no previous lossless transcoding sounded best.  Interesting stuff. I'm not going to dismiss this out of hand just yet. I get the feeling we still have quite a bit to learn and understand on this...

Everything is to be adjusted up to the last minute, of course. Naim is historically notorious for adjusting and adjusting and adjusting -- its the only way to get quality.
 
Mmm, my experience differs, a Muso and UnitiQute that would randomly turn themselves on at different times of the day!! 

 

Kermit posted:
Everything is to be adjusted up to the last minute, of course. Naim is historically notorious for adjusting and adjusting and adjusting -- its the only way to get quality.
 
Mmm, my experience differs, a Muso and UnitiQute that would randomly turn themselves on at different times of the day!! 

 

That's a security feature to confuse any potential burglars and make them think there's people in

Kermit posted:
Everything is to be adjusted up to the last minute, of course. Naim is historically notorious for adjusting and adjusting and adjusting -- its the only way to get quality.
 
Mmm, my experience differs, a Muso and UnitiQute that would randomly turn themselves on at different times of the day!! 

 

That's a very bold statement. 

There are indeed differences in lossless files and the article Richard alludes to was in HIFICRITIC. The view taken there was that metadata made the biggest difference to the sound. Even adding the metadata at a different point in the process of creating the file would make a difference, and quite a substantial one at that. The research continues. I don't know why the differences can occur since the data is presumably the same (though I still have my doubts about this, even with so-called bit perfect rips). Perhaps it's something to do with how and where in the file the metadata is stored. I haven't had the time, patience or inclination to check. I just tried what I tried and got vastly different results to HH. I didn't say he was wrong, just that MY experience is different from his. There could be any number of reasons why that is, both due to the hardware being discussed as well as the environmental hardware (routers, network) so quite frankly you'd be an idiot to claim that bits are just bits when there's all that mucking about with them in IP land before they even get to your renderer! 

The fact is I heard a difference, my wife heard a difference and my colleague heard a difference, and you cannot dispute that. You can't say it's psychosomatic because these were different tests at different times. Its what we heard, period. If you want to convince yourself we didn't hear it, well that's your problem, not mine, nor my wife's or colleague's. Incidentally, we were rather bemused by the results (we expected little to no difference) so we discussed what was different and we basically came to the conclusion that we were hearing the same differences. 

Frank.

Frank Abela posted:

 so quite frankly you'd be an idiot to claim that bits are just bits when there's all that mucking about with them in IP land before they even get to your renderer!

Frank.

Chill Frank One way or another we are all victims of some crappy claim made by someone somewhere.

In this case I suppose it boils down to whether an individual would pay £1600 to rip their cd collection into one type of lossless or pay nothing for another type of lossless (or just keep using their cd player for that matter). Personally, I'll stick with geting an Atom.

I have no idea why either on a technological basis, but I heard the same exact file played through Minim and then through a Unitiserve at my dealer and there certainly was an improvement. I am a huge skeptic of a lot of this Hi-Fi hocus pocus, but I concur with you that the difference was there. Is it possible the Uniti is decoding the flac or doing some other pre-processing before sending it to the streamer? I am very curious is the Core will wind up being audibly better than a standard NAS with Minim.. I would love any explanation why this is the case, because it defies logic, yet I heard it myself.

> On Nov 28, 2016, at 4:06 PM, Naim Audio Forums <alerts@hoop.la> wrote:
>

It's said that unitiserve outputed from net serve streamed into a net device is better than when locally direct digital into a dac. Though going through all the loops and holes and this bit and that bob to get  the net loop working at a reference level seems that this presumption is just theoretical.

DUPREE posted:
I have no idea why either on a technological basis, but I heard the same exact file played through Minim and then through a Unitiserve at my dealer and there certainly was an improvement. I am a huge skeptic of a lot of this Hi-Fi hocus pocus, but I concur with you that the difference was there. Is it possible the Uniti is decoding the flac or doing some other pre-processing before sending it to the streamer? I am very curious is the Core will wind up being audibly better than a standard NAS with Minim.. I would love any explanation why this is the case, because it defies logic, yet I heard it myself.

> On Nov 28, 2016, at 4:06 PM, Naim Audio Forums <alerts@hoop.la> wrote:
>

The Core has the potential to sound better because it dispenses with the need to stream across a network - much like Mac Mini with Audirvana, Melco, innuos Zenith, feeding a rendered stream direct into a DAC

Frank Abela posted:
Hungryhalibut posted:

Indeed. It's total bollocks. I believed it though, when I bought my UnitiServe. Then you discover that both the Serve and dbpoweramp look at the perfectrip database, just the same. 

Hmm, that's weird. 2 years ago I tested this to check my sanity and the difference was substantial. I made several tests to verify. 

I ripped some music using US and got it onto a USB stick. Ripped same music using optimized dbpoweramp on my home PC and for devil-may-care same again with itunes. Placed all on same stick. Played the tracks via my BD player through 282/hi/250.2. The US ones were consistently better. Played the tracks for SWMBO and asked if a) there was a difference and b) if it was valuable. She much preferred the US tracks (without knowing what she was listening to) and she thought they were valuable though the US was a bit of a tall one. The dbpoweramp vs itunes was a no-brainer.

Back in the shop, I tried the same rips using our NAS drive's own server (Twonky) vs our US as a server. Significant difference and did a blind test on my colleague with same results. Again the US sounded better serving the same track as twonky via the same client (might have been NDX, might have been NDS, I forget) into our large-ish Naim system at the time. So I don't understand your experience. 

...

 

Interesting observations Frank, thanks for sharing! 

Your findings (and, indeed, most of the comparisons between the outcomes of "bit-perfect" ripping on different platforms, replay of .wav and .flac files, replay of .flac files with different metadata, etc. that from time to time pop up in this forum) raise a number of interesting questions.

An obvious one is that of whether bit-perfect ripping on two different platforms yields, up to the metadata, identical files or not. In your case, this question could be answered by removing all metadata from the files. For .flac files, this can be done by running "metaflac" with the "--remove-all" option.

If ripping on the US and on your home PC resulted into identical files up to metadata, your observations would suggest that metadata can have a strong impact on sound quality.

In the other case case, your observations would suggest that "bit-perfect" ripping does not guarantee equivalent outcomes and that the US and your home PC produce genuinely different results.

That the same files sound differently when served by different UPnP servers running on different machines is not completely surprising. It would be interesting to know whether these differences are mainly due to differences in the UPnP servers, in the computing platforms, in ... the IP addresses of the devices or to aspects that we do not understand.

Perhaps we should start, whenever we come across unexpected or surprising results, to share the data required to reproduce such results. There are many platform for sharing data and protocols of experiments or tests, GitHub is what first comes to my mind. Perhaps Naim could even set up a repository for sharing intercomparison data and protocols and help making comparisons and empirical findings more understandable and reproducible.  

 

DUPREE posted:

Anyone know how they are going to handle this on the Naim App side.. For instance, is there going to be a seamless way to control the core if you attach it to say a 272 or other iOS controlled preamp?

Are you meaning when they are connected to a common network via Ethernet or are you meaning when connected directly to each other via S/PDIF?

Obviously they will both need to be connected to your network to be able to be controlled by the app anyway...

Cheers

Phil

FWIW, I believe that bit perfect rips really are that - i.e. that the bits stored in the files are the same. However, I suspect that either the metadata is stored differently, or that the reading of the file in real time by the server or the rendition of the file in real time on the renderer differs depending on the location of the metadata in the file or differs in the means that is used to read and expand said file in real time. The article in HifiCritic pretty much seemed to say this and their research is ongoing. Very interesting article. Incidentally, if memory serves (I don't have the article nearby) they found that WAV files were affected less than FLAC by the metadata, and they found that FLAC with metadata removed was pretty much identical to WAV in quality whereas FLAC was 'worse' than WAV when metadata was present.

Frank.

That makes no sense either. The metadata is at the beginning of the file then it is followed by frames of the music stream. The stream itself does not have the metadata interspersed; it should be the same stream regardless of what the metadata is. The file sequentially just contains
The four byte string "fLaC"
The STREAMINFO metadata block
Zero or more other metadata blocks
One or more audio frames

I don’t see how it having or not having metadata could be the issue.
DUPREE posted:
That makes no sense either. The metadata is at the beginning of the file then it is followed by frames of the music stream. The stream itself does not have the metadata interspersed; it should be the same stream regardless of what the metadata is. The file sequentially just contains
The four byte string "fLaC"
The STREAMINFO metadata block
Zero or more other metadata blocks
One or more audio frames

I don’t see how it having or not having metadata could be the issue.

Well, for one thing the renderer or perheps the control point might query a file's metadata at replay time. This does not need to happen, of course, but can happen. Perhaps it never happens, perhaps it happens with a fixed frequency, perhaps randomly.

Of course, there could be other reasons why metadata of .flac file can affect the sound quality. I personally have never experienced these effects. But, in absence of detailed knowledge on how data streams (and queries) are handled, it is very difficult to exclude a priori the possibility that metadata can have an impact on sound quality. In particular, when there are empirical evidences that seem to support this possibility.

It would be nice if files that have been found to sound differently with and without metadata would be shared. We could then attempt at independently confirm or confute these findings.

Metaflac (https://xiph.org/flac/document..._tools_metaflac.html) is available for virtually all platforms and allows one to remove all metadata from a .flac file with the "--dont-use-padding --remove-all" options.

DUPREE posted:

I more meant integration - would you have to flip back and forth between devices to control - i.e. go to core to select track, go back to device menu go to 272 for volume etc or would it understand they were two parts of the same system

If you're using the Core as a UPnP source then it will work exactly as it does now ... you'll have the app controlling the NAC-N272 (for example) and in the 272 control select UPnP and browse the Core's UPnP server.

There's no reason that would change as far as I can see?

Phil

David Hendon posted:

Phil

I think he means if you are using the Core as an HD Player into a 272.

best

David

Hi David,

That's not really the way that you would normally use it but in that case then you would have to switch between controlling the 272 and Core in the Naim app if you wanted to control them.

Phil

nbpf posted:
DUPREE posted:
That makes no sense either. The metadata is at the beginning of the file then it is followed by frames of the music stream. The stream itself does not have the metadata interspersed; it should be the same stream regardless of what the metadata is. The file sequentially just contains
The four byte string "fLaC"
The STREAMINFO metadata block
Zero or more other metadata blocks
One or more audio frames

I don’t see how it having or not having metadata could be the issue.

Well, for one thing the renderer or perheps the control point might query a file's metadata at replay time. This does not need to happen, of course, but can happen. Perhaps it never happens, perhaps it happens with a fixed frequency, perhaps randomly.

Of course, there could be other reasons why metadata of .flac file can affect the sound quality. I personally have never experienced these effects. But, in absence of detailed knowledge on how data streams (and queries) are handled, it is very difficult to exclude a priori the possibility that metadata can have an impact on sound quality. In particular, when there are empirical evidences that seem to support this possibility.

It would be nice if files that have been found to sound differently with and without metadata would be shared. We could then attempt at independently confirm or confute these findings.

Metaflac (https://xiph.org/flac/document..._tools_metaflac.html) is available for virtually all platforms and allows one to remove all metadata from a .flac file with the "--dont-use-padding --remove-all" options.

I think FLAC is a great way to store music, but comparing two FLAC rips made by different software may be hard. It is compressed after all, so there are other variables than the software. Comparing WAV makes more sense to me.

A CD ripped to WAV *should* sound the same no matter how it was ripped . . . but of course it doesn't. There are re-reads and C2 and lots of other ways of correcting for errors, but they do happen, due mainly to hardware and firmware, but also the ripping software (eg EAC and DBPowerAmp have different error levels).

That's why we have AccurateRip. What that does is compare lots of people's rips of the same album. If there are 50 rips and 40 of them are identical to the bit, then it's reasonable to assume that those are accurate. (As an academic, I'm forced to say that this is an assumption - all 40 may be inaccurate. But the more rips there are and the more consistency there is, the higher the chance they are accurate. Also, being accurate doesn't necessarily mean the best-sounding, but that is subjective, so I'm going to stick with 'accurate' as the goal.)

If two rips to WAV are both deemed accurate by AccurateRip, they are the same. It matters not one jot what software was used to rip them. If they sound different, then there's another reason for it: eg one or other ripping software wasn't using AccurateRip, or one or other rip failed AccurateRip (at least to an extent). It can also happen that AccurateRip doesn't have enough examples of an album to be certain, or the examples it has are not consistent. In that case it will tell you. But that's not relevant to this exercise: there's no dispute that errors in rips can happen, only that rips from different software sound different even after AccurateRip has approved them.

Best of all, we can test this. Take EAC, DBPoweramp, and a Unitiserve, and rip the same CD to WAV on each - make sure it's an album AccurateRip has plenty of examples of (I believe Dido's No Angel is considered the go-to for this). If they pass, put each resulting file on a USB and run a blind test (double blind is better). My firm belief is: you won't be able to tell which is which. If that's the case, you may safely assume it makes no difference to sound quality which software you use.

Likes (6)
Cbr600fred40Doug GrahamKlydeIan G.Artferg20
×
×
×
×