Naim Ripping (again)

Posted by: james n on 25 April 2012

I've yet to read a satisfactory explanation of why the Naim ripping engine is superior to others so was interested to read comments on the Stereophile website from the Sound Organisation. Links aren't allowed (google - sound organisation stereophile naim nds) so i'll quote -

 

Naim’s server products sound so great for a couple of reasons:

 

First, they rip into WAV, which is an uncompressed format that requires less processing to play back than lossless formats like FLAC.  For example, there is no caching, zipping, or unzipping that occurs before the music data is sent out via Ethernet.  A good analogy to help understand this concept is that of a tire – WAV, FLAC, or any lossless file is like a completely inflated tire.  A WAV (or AIFF) tire stays inflated permanently, but with a FLAC or other lossless format, a computer extracts the zeroes (the air) from the file (the tire) so it takes up less space in storage.  Then during playback they are put back in (the tire is re-inflated).  This requires processing and unfortunately this creates noise in an audio system.

 

The second reason deals with the way they rip the CDs into WAV.  The optical drive Naim has chosen has no ability to apply error correction – which means it has no ability to change the original data.  On top of this, the operating system itself does not touch the data, which translates into the purest, most musical sound possible.  A cool side effect of this design is that not only will files ripped by the Naim server sound better, but so will files ripped and stored elsewhere and streamed through the server (again, because there is no extra processing getting in the way).

Naim decided after extensive listening tests that using WAV, and keeping the data completely segregated from the operating system, resulted in the best possible sonic performance.  Because at the end of the day, it’s what our ears tell us that matters, right?

 

I'm quite happy with the first reason - that's quite plausible and i understand the reasons for this happening, but on the second, the OS doesn't touch the data, i take it there is no manipulation or mathmatical processes implemented. How does this differ from a conventional ripping engine ?

 

Thoughts ?

 

James

 

Posted on: 26 April 2012 by pcstockton
Originally Posted by Guido Fawkes:

Hi DQ

 

How are playing FLAC through your Apple TV? 

 

All the best, Guy 

Another way is to use a control point with Airplay compatibility.  PlugPlayer comes to mind.  JRemote (for JRiver MEdia Center) will have it in the next update.

 

-Patrick

Posted on: 27 April 2012 by bhaagensen

I think there is some mixing apples and oranges here with to respect to P. Harris' statements regarding the nDAC. 

 

The whole processing generated noise issue wrt. flac vs. wav only applies as long as one is using the USB input. It doesn't apply [at least not in the same sense] for those feeding the nDAC through s/pdif.  

Posted on: 27 April 2012 by Guido Fawkes

True - the problem only seems to be with the USB and I agree there is no problem putting WAV files on a USB stick as a workaround. 

 

However, what if streaming to the NDS, if it can't plat FLAC well then that would be a problem with my set-up - I don't want to have to care about file formats - I want to be able to play any lossless file and expect them to sound great irrespective if they are AIFF, ALAC, FLAC or WAV. If the NDS doesn't manage this then perhaps it is not for me - unless it can acts a super Naim DAC through S/PDIF and I can ignore its Ethernet .... WAV just doesn't work well for me because of its tagging issues (or at least the tagging issues I have). 

 

I can cope with AIFF (Amiga Interchange File Format) - though FLAC works better for me. 

Posted on: 27 April 2012 by bhaagensen

Simon - we must be misunderstanding each other? Clearly, if the nDAC receives its data over s/pdif, there is no difference whatsoever on the upstream side [processing-wise] on wether the original file was flac or wav [or format x for that matter]. Using USB, there is of course such a difference...

 

 

Posted on: 27 April 2012 by Simon-in-Suffolk

Sorry, yes absolutely correct, I'll delete the post.

Posted on: 29 April 2012 by Dungassin
Originally Posted by bhaagensen:

I think there is some mixing apples and oranges here with to respect to P. Harris' statements regarding the nDAC. 

 

The whole processing generated noise issue wrt. flac vs. wav only applies as long as one is using the USB input. It doesn't apply [at least not in the same sense] for those feeding the nDAC through s/pdif.  


Perhaps that would explain why I differ from many here in being unable to hear any difference between uncompressed FLAC and WAV using my NDX/555PS/nDAC?  I suppose I could try it without the nDAC to see (hear?) if there is any difference ... 

Posted on: 29 April 2012 by T38.45

I ripped 3 CDs I know very well on :

HDX ( ask my dealer to  do this for me , moved files to  my USB stick)

Unitiserve ( same)

Mac/ XLD in flac and wav (at home, I use this setup as my standard gear)

All files stored on my local Qnap NAS.

 

Played on Linn  KDS in random access and in playlist mode, I and my wife couldn't hear any differences btw. Naim rips, flacs and wav 

Don' t think about ripping, start listening  more music!

 

Ralf

Posted on: 29 April 2012 by likesmusic

The original quote says: "The optical drive Naim has chosen has no ability to apply error correction – which means it has no ability to change the original data. "  If this is true then, over time, it is LESS likely to deliver correct data than a drive with error correction. It simply won't be able to correct errors that a drive with error correction could perfectly. Not very smart imo. 

 

It may be that Naim can't get FLAC to sound/measure as good as WAV. But this is arguably not a necessary law of the universe, just a question of engineering. 

 

By contrast, Linn have made the following statement:

 

1. If we measure the power rail that feeds the main processor in the DS we can clearly see identifiable disturbance patterns due to audio decoding and network activity. These patterns do look different for WAV and FLAC - WAV shows more clearly defined peaks due to regular network activity and processing, while FLAC shows more broadband disturbance due to increased (but more random) processor activity.

2. If we measure the power rails that feed the audio clock and the DAC we see no evidence of any processor related disturbances. There is no measurable difference (down to a noise floor measured in micro-volts) between FLAC and WAV in any of the audio power rails.

3. Highly accurate measurements of clock jitter and audio distortion/noise also show no difference between WAV and FLAC.


Posted on: 29 April 2012 by Guido Fawkes

Interesting Ralf and Likes -

 

I cannot hear any difference through my Naim DAC/555PS (USB stick), but was beginning to think my DAC might be faulty 

 

The Linn results are as I would hope for the NDS

 

.... I, too, am totally lost as to why anybody would not want to use the error correction built in to the CD drive, it is not sufficient on its own, but it is one the weapons along with multiple disc reads and the Accurate Rip database to get the data right. In practice, I've found no difference between the PCM in different rips irrespective of whether they use all the available techniques. 

Posted on: 30 April 2012 by Phil Harris
Originally Posted by likesmusic:

The original quote says: "The optical drive Naim has chosen has no ability to apply error correction – which means it has no ability to change the original data. "  If this is true then, over time, it is LESS likely to deliver correct data than a drive with error correction. It simply won't be able to correct errors that a drive with error correction could perfectly. Not very smart imo. 


 

OK - the original quote itself isn't correct.

 

The optical drive that we use in both the UnitiServe (the slot loader) and the HDX / NS0x (the tray loader) will both apply error correction to any data frames read that contain errors by default. However both mechanism have the ability to be *ACCURATELY* queried to see if a frame of data contains any error corrected data and if so then we can disable error correction and re-read that frame making our own adjustments to focus, laser power and drive speed as necessary to ensure an error free read as far as possible.

 

However the proposal that a drive with error correction is going to deliver better data than one which has error correction disabled is a little bit of a red herring in itself - the drive will deliver data and may simply not flag it as having been error corrected or interpolated and you would never know.

 

Originally Posted by likesmusic:

It may be that Naim can't get FLAC to sound/measure as good as WAV. But this is arguably not a necessary law of the universe, just a question of engineering. 

 

By contrast, Linn have made the following statement:

 

1. If we measure the power rail that feeds the main processor in the DS we can clearly see identifiable disturbance patterns due to audio decoding and network activity. These patterns do look different for WAV and FLAC - WAV shows more clearly defined peaks due to regular network activity and processing, while FLAC shows more broadband disturbance due to increased (but more random) processor activity.

2. If we measure the power rails that feed the audio clock and the DAC we see no evidence of any processor related disturbances. There is no measurable difference (down to a noise floor measured in micro-volts) between FLAC and WAV in any of the audio power rails.

3. Highly accurate measurements of clock jitter and audio distortion/noise also show no difference between WAV and FLAC.

 

 

My original post was quite specific in that it referred to the DAC and didn't reference any of the streamers - if the processor in the DAC is unpacking the FLAC file from its USB input then it does sound slightly different to playing the same piece of music as a WAV file from USB. If you have an external device doing the "unpacking" (such as a UnitiServe) and are playing via S/PDIF into the DAC then I would not expect to hear *ANY* difference between the two file formats.

 

In that context then the statement quoted is correct and I am sure this applies to the Linn streaming clients as it does to ours - the UPnP client side of the streaming products (ND5 / NDX / NDS) is quite independent of the DAC.

 

Phil

 

Posted on: 30 April 2012 by james n

Thanks Phil - that was just what i was looking for.

 

Likesmusic - Yes i've seen that from Linn but they don't mention anything about radiated noise within the DS - one of the reasons why the Klimax uses the enclosure design to screen the PSU / Streamer and Processing / Output stages from each other - and hence why a Renew DS doesnt quite sound as good as the original Klimax, even though it's using the same board and PSU.

 

James

 

Posted on: 30 April 2012 by Dungassin
Thanks Phil, you've confirmed what I thought. As I'm taking SPDIF from my NDX to the nDAC I might just as well rip to compressed FLAC on my NAS and save space. After all, I can always set Asset to output WAV should I feel the need.
Posted on: 30 April 2012 by Guido Fawkes

Thanks Phil 

 

> the UPnP client side of the streaming products (ND5 / NDX / NDS) is quite independent of the DAC.

 

That's great Phil, you would not expect there to be differences in the Sound Quality whether the streamed input to the NDS was FLAC or WAV (assuming all other things are equal) The independent client would unpack it and send it to the internal DAC and so the DAC would process exactly the same PCM and be unaffected by whatever occurs during the unpacking process (within the NDS). 

 

That is exactly what I was hoping to hear - the forthcoming NDS will be able to play my FLAC files (in my case from the UPnP server in my Vortexbox) with exactly the same Sound Quality as it would if I were using WAV .... so no need for me to convert anything to WAV or transcode on the fly to get the full benefit from the NDS. 

 

Many thanks for clearing up the confusion - I can continue to plan for my NDS without any need to use WAV files to get top notch replay. 

 

All the best, Guy 

Posted on: 30 April 2012 by Simon-in-Suffolk

hi, that could end up being one of they differences between NDS and NDX. Ie

1) NDX-> ndac/555PS (WAV optimised)

or

2) NDS / 555PS ( FLAC/WAV optimised)


Simon