Naim rip vs PC rip

Is it just me and my Firefox 4.0 browser, or are others seeing AMA's post in a

typewriter-like font, running out of the text box, and overwriting the space to the right?

I've long complained we should be utilizing that space for posts, but not like this! 

Hook

Hi, Hook. I use Chrome and the text fits the frame properly. Though the font is different -- I have no idea how did it come that way...

I agree the frame should be extended for the post space. Very strange design. But each time I pay a visit to Linn forum I realize that Naim designers at least were not double paid by Hades.

Hi AMA -

And now, as seen by my browser, this second post of yours is back to normal....weird.

Hook

Hi, Hook -- I have edited the post ( I have actually started it as new). Hope it works now.

Shall we go back to the topic ? 

What does the header do apart from the memorizing the replay gain?

I do believe the gain adjustment may sound different and even pleasant.

But this is what we can do in Foobar by will.

What else?

I recall reading that the format and use of the extended header information in a WAV file is "user defined".   So unless Naim comments, we are simply guessing at how that metadata is used. 

Hook

Hook, no the extended data header is used by most modern encoders, it is specified by the Microsoft file format, and is essential for hires audio, and it essentially tells the receiver how to interpret the included sample data (word length, number of channels, sample frequency, compression.used etc.)

You tend to find the 'chunks' of data that are non standardised occur, but don't have to be, at the end of the wave file. DBpoweramp for example puts a whole lot of tagging info at the end of the wave file in it's own chunks.

So in summary, the extended data header is specified in the wave format( so the wave file can play on any device). The older 'canonical' format of the wave file only has a simplified header  but is only really useful for unto 16bit stereo.

Simon

OK, let me set up 3 points and 2 questions.

1. The PCM data of Naim rips is the same as original CD (and the same as EAC/dbPoweramp rips).

2. Naim rips contain some extra information in file header which as I can see only Naim players can benefit from.

3. The extra data was generated by Naim ripping software and it was sourced from PCM data as there is nothing else in original CD but the PCM data.

Question 1. Are we sure that sonic improvement some people allegedly hear with naim rips is not just a small change in replay gain which is often misinterpreted as sonic improvement especially in direct A/B tests? 

Question 2.  If Naim has managed to generate extra data from PCM then possibly other mathematician could do the same. I doubt there is a rocket science in this -- but I can admit that Naim engineers came across some brilliant ideas which may take some time before the others will pick them up. If this is the case so why Naim has put these brilliant ideas into the headers of their rips for public peer view rather than secure them by training the naim players to generate this extra data on the fly from the regular PCM bitstream (as a part of DAC pre-processing)?

Hello Simon 

Is your view that the additional metadata would have no effect on the SQ of a 16-bit CD rip? 

Thanks

All the best, Guy

Hi AMA

I think your point 1 could be extended to iTunes, XLD and Max if we assume the original CD is in good condition. 

All the best, Guy

Simon,  do we have references (on the forum) to the cases that Naim rip and PC rip have been played through the SAME player (let it be Naim player like HDX) and the Naim rip sounded better?

Here is a thread where Paul Stephenson claims the Naim rip 'usually outperforms' a non-naim rip.

https://forums.naimaudio.com/di...ent/3960068604537349

However, on the same thread, Phil Harris was much more equivocal:

"However the global question of "Do you believe that a UnitiServe rip sounds better than an EAC or dBpoweramp rip?" is not really in itself possible to answer in any way that is valid"

Make sense of that how you will!

We have a statement from Paul Stephenson that naim rips 'usually outperform' dBpoweramp rips.

We have an observation by Jack that a the audio portion of a dBpoweramp rip is identical to the audio portion of a UnitiServe rip.

This means to me that a Naim player does not correctly play dBpoweramp rips.

Whether this is due to some differences in the file header, voodoo, or might even be deliberate is just speculation. But it is certainly very disappointing to anyone with a large collection of good lossless dBpoweramp or EAC rips that might have looked to Naim for a DAC and/or server to play them from.

Would be nice if Naim were to fill in the blanks...

1. All rips sound better using Naim network players because _______ .

2. Naim rips sound better using Naim network players because _______ .

3. Naim rips sound better using any network player because _______ .

I'll help.  #1 is easily answered.  Just read the Technical Detail section for the NDX, and it is clear that a lot of great engineering for high sound quality went into building this component.  (I'll save AMA the bother by adding "And I cannot wait to hear the NDS!"  )

I have been trying to hold out the possibility that #2 had an answer -- that there was some chance of unique synergy between Naim rips and Naim network players.   But if there is nothing unique to Naim in the header, and if the PCM data is the same, then I don't see it.  And if #2 cannot be answered, then #3 has no chance either.

Simon - I am still trying to understand your last post.  I think you are saying that it is possible, in theory, for a Naim network player to make good use of WAV extended header information, but that there is nothing unique to Naim in that extended header, and that the exact same information could be provided by dBpoweramp.   Is that correct?

So unless I am continuing to miss something here, then I am at a total loss trying to understand how Naim rips can sound better than any other bit-perfect rip.

Hook

PS - Is it just me, or does this not feel exactly like the old "can sources for the Naim DAC sound different" debate?   Simon - do you know Andy S by any chance? 

Wrong conclusion–the same difference can be heard on a Linn DS system–therefore it is not player-dependent, but file dependent (header + data). And I'm not the only person who can hear the difference.

The difference can be also heard on a Naim server (player / OS) as well as a Naim streamer (different player / OS).

Does this mean that the DS is broken? Hardly–but it may point to a flaw in the header information in dBpoweramp rips, if that is the sole reason for the difference (which I also doubt).

If so–then a fix should be available from the dBpoweramp folks–much like the uncompressed FLAC builds of late. Who's catching up now?

David, why don't you make a UnitiServe rip available online so that people can listen to it for themselves, and judge for themselves.

You have also clearly suggested there may be  a 'flaw in the header information in dBpoweramp rips', and alluded that there may be other sources of difference.

Making a UnitiServe rip public would allow the dBpoweramp folks to defend their product and fix it if necessary.

I have raised this issue on the dBpoweramp forum.

Likes, Nothing is stopping dbPoweramp from buying a Serve and and figuring it out themselves if they gave a shit (which I guess they dont). Naim doesn't owe dbPoweramp anything. -p

dBpoweramp do evidently give a shit Patrick. Spoon, the author of dBpoweramp, has already replied to the thread I posted raising this issue on the dBpoweramp forum. Here is his reply:

"I have read that, a fact chunk makes no difference to a wave file, it simply informs the size of the wave file in samples and should not even be used for normal PCM wave files, it is only needed for compressed wave files (such as mp3 in a wave file) where the sample count is unknown.

How it effects the quality, Naim would have to specifically code to degrade the audio stream if it does not contain a fact chunk (which would be absurd if they did), otherwise there would be no degradation"

I still believe one does not need spend money on US to know that his ripping engine is OK.

Take WAVE record, burn CD, rip it with dbPoweramp and if it goes clean you're done.

You don't need to care of what the others do because you know you do the job OK.

If they do, I am sure you could email them to David Dever and he would gladly modify the headers so they sound the same.