Naim Ripping (again)

Plausible - when I look at the horizon I can almost believe that I'm looking at the edge of the world - ..... almost 

We have been here before and we will be here again if that doesn't sound too much like a bad episode of "Twin Peaks" or "Battlestar Galactica" . There is an equal amount of so called "evidence" that refutes these claims floating around and although I'm sure Naim's ripping software is wonderfully magical but I don't believe personally I can hear the difference between a file ripped to wav, aiff or flac and wav has very poor support for tagging. 

I use xld - rip to flac and avoid overpriced servers.

Tog

WAV files don't support tags in the same way FLAC and AIFF do, but iTunes can compensate for this. I prefer FLAC or ALAC because they take up less space and I can hear no difference through my system between them and WAV or AIFF. 

As for ripping XLD and iTunes both make bit perfect rips that in terms of PCM are identical those from a Naim US and from dBPA - at least for the tracks I've compared. The metadata is different and dBPA is clever with HDCDs. 

Legend has it Naim does better rips and WAV sounds better, my tests show no difference and my ears hear no difference. I can hear a difference between between different hardware connected to my Naim DAC and more so through the S/PDIF on my UQ. 

To me, from the kit, I have Vortexbox sounds best in to the UQ by a small margin. Into the Naim DAC I find little to choose between the Mac Mini and W4S Sonos Connect - possibly just in favour of the latter (its use of up-sampling may be the reason). 

My thoughts on the article is it is nonsense. I can capture the output from the ripped files from my UQ's digital out and they are bit for bit identical. Differences in sound, if they exist, are not down to ripping - others may disagree, of course. 

It is just if Naim ripped better then I'd expect to see some difference in the files I have that were ripped on a US and ripped with iTunes, but perhaps I'm mad after all. 

I'd never say people can't hear differences if they say they can - I just find it hard to accept some of the explanations I read. 

What do your ears tell you James? It's what matters, right?

I have not heard Naim rips served by Naim servers so I can't say whether they sound better or not. I have tried WAV vs. FLAC vs. AIFF through my ND5XS though and they all sounded the same to me. This is not to say that others don't hear a difference but I didn't. My gut feeling tells me that, if there were a difference between Naim rips served by Naim servers and standard rips from a standard NAS using another UPnP server, than these are likely to be heard on very high end systems only. I'd say for the vast majority of Naim owners the difference is too little to worry about let alone justify the substantial amount of extra money the Naim ripping solutions cost, when you can just as well produce your own rips on your Mac or PC. Now if you wanted a Naim ripping & serving solutions because you like their ease of use, fair enough but there are other options out there that are no less intuitive at a fraction of the money Naim charge.

when having a chat with a friend about my hdx, we got onto the topic of ripping v's cd.  as you can imagine these chats can go on for a while.  i was quite adamant that a naim rip sounded awsome, which it does, my friend not so convinced.

the test came when he brought round with him a muse cd and the muse usb that this album came with (the usb contained the superior copy of the album with the muse embossed on the side)

when comparing the usb played on hdx via ndac against the hdx rip, would you credit it, the hdx rip sounded better, we both agreed.

he now has a hdx!

You have no problems until you try and use them wavs with anything other than dbpoweramp.

This is the problem, all the different apps including itunes have sorted out the inherent problem of wavs, you cannot tag them!

James, I agree the second point defies real meaning to me. Also last year from an engineering perspective I analysed Naim rips with rips from EAC and dBpoweramp with various  optical drives and the pcm data in the WAV data chunk was identical bit for bit. 

In  fact the only benefit I can see from different ripping engine/  optical drives is thier reliable ripping speed,  strategy for reading damaged discs and tolerance for reading discs outside of the Redbook standard. Also accuraterip is of great benefit for fast effective rips, only if the checksum doesn't match or is not available to validate against  a lower speed rip is then performed to mitigate any optical drive / CD Terence issues.

To Gary... It's only the id3 tags in WAV that are supported by a few program's like dbpoweramp, however the LIST INFO tags is also part of the WAV standard, and is also used for tagging in both commercial and consumer software. If the WAV software only supports the old 1980s Canonical WAV standard (like itunes) as opposed to the Extensible WAV standard it won't support LIST INFo and unlikely to support reliably (if at all) hidef. Dbpoweramp writes both id3 tags and the standardised WAV tagging method, LIST INFO for interoperability. However because of the RIFF structure of WAV the  parser can skip over WAV segment/chunks it has chosen not to support. This is one of the great advantages of WAV, but it is double edged..... As we see with poor or badly written software. But that is a software issue not WAV format issue.

The main benefit of id3 tags over Listinfo is that contains album art and allows easy custom tags! Its disadvantage unlike Listinfo tags, id3 tags is not officially in the WAV standard yet.

Simon

Simon, lots of geekey words there, but the conclusion for most people is that wavs are a pain to transport between software packages, and from that perspective its a problem of wav.

Ah! 

That explains why I can't find certain artists albums under the artist folder on my twonky server qnap nas.

I add some albums and they don't show in the artist folder, but if I look under 'by folder' option  and then artist they are there. It is always the wav files that do this. 

It strikes me that streaming audio is a bit like having a PC as opposed to having a Mac. I was originally going to go for a Uniti as it had the CD player in it and streaming was a nice add on, but after being impressed with the sound quality of the streaming I went for a qute with NAS, but didn't realise just how problematic that would be. Seems fairly sorted now thanks to support from forum gurus.

Hiya Tog,

It's a very easy comparison to do if you have access to a Naim DAC and I did this test a year or so ago.

Take two identical memory sticks - put an album on one as WAVs, put the same album on the other as FLACs. Label them up as "A" and "B" but don't say which is WAV and which is FLAC - play them to an assembled "audience".

I tried this with two groups of people here and without exception they could quite easily (and very reliably) tell one stick from the other even though they didn't know which stick I had plugged in at the time.

Cheers

Phil

I find that comment intriguing. Does that mean the Naim drive never re-rips on the fly by comparison against a reference database such as 'AccurateRip'. That is what is implied by this statement as I read it. I am not sure I would want to rip that way personally.

Beyond that I have recently hosted a listening session using my Linn KDS/1 as the replay box. The files involved were sometimes FLAC, some ALAC and some WAV and in several cases at least two alternate format files were available to compare coming from different rip hardware. Couldn't hear any differences really.

For these reasons apart from the theory discussion I am very dubious of this claim. It is a nice thought to hold in your head when listening to a Naim DAC product from a personal replay point of view though.

regards

Geoff 

I can tell WAV and FLAC apart easily, WAV being the better of the two.

I used to have a lot of problems with WAV tagging (artwork in particular), but thanks to help from fellow Naim forum members here, nStream now displays WAV and FLAC files equally well.

I rip classical and jazz music to WAV, and pop/rock music to FLAC because generally I find quality of pop/rock recordings not very high so I should not waste valuable hard disk space on them.

James N: as for the 2nd reason I believe it (too) is a bit far fetched [as an argument for placing nRipper over other quality options]. Nonetheless, its a possible parameter. Basically, since reading of a RedBook CD, as opposed to a data-cd, is done at all cost [errors allowed], it may be that some drives implement their own type of error recovery* by filling in the possibly erroneous parts with whatever with no way of overriding nor knowing when it happens. However I don't think newer drives work this way.    

* I think most definitions take error correction to mean 100% recovery of the original data - c.f. Reed Solomon coding etc.

Was this using uncompressed FLAC?  I ask because I STILL can't tell the difference between uncompressed FLAC and WAV.  Just done this comparison now, using the Linn 192kHz Berlioz Fantastique.  Perhaps my hearing is finally going - although my ENT Consultant friend tells me I have excellent hearing. 
 

I can't hear any difference using my Naim DAC with a USB stick using WAV or FLAC, but I agree that could be a limitation of my ears or my all Naim system or both.  

If it is the case that the Naim DAC plays WAV better than FLAC then I hope this will be put right with the NDS so it plays identical PCM data, well identically. I'd like my FLAC files to sound just as good as WAV files. 

Strangely, if you capture what comes out the digital out of a UQ then FLAC and WAV send out an identical bitstream - does this mean the rendering software in the UQ is better or worse than in the Naim DAC.

This is very confusing ......... 

I'm not keen on WAV as a format because it has real world compatibility problems - ones I do not experience with FLAC or ALAC or AIFF. 

As you can turn a FLAC in to a WAV on the fly with some software then surely a top line Naim system should be able to extract the PCM with no problem from both formats ... 

Phil, will the NDS play AIFF with the same SQ as WAV as I could keep my files in AIFF?  

Or alternatively should I retain the Mac Mini and simply send a PCM stream from it to NDS's optical input (seems a bit strange, but I can easily do that if it improves the SQ)?

I think computer audio is becoming a lot more complicated, just when I thought it was settling down.

Perhaps I should investigate the KDS after all as Geoff says it works really well with FLAC.   

As I've never tested WAV vs. FLAC I have no idea if you can tell the difference.... Just wondered why though when people report an audible difference between WAV and FLAC It's always WAV that is reported to sound better???

When I used a minimum build laptop running CMP2 I couldn't tell any difference between FLAC & WAV through the nDAC via two different systems:

Laptop>nDAC>552>300>Audiovector speakers (from memory)

Laptop>nDAC>EAR864>EAR534>Living Voice speakers

When I got an NS01 the difference was obvious, with WAV being the clear winner - and I am NOT saying that the NS01 is better than the laptop; although it is a damned sight more convenient.

Sorry Phil, must be my clothe ears. I can hear no difference between my Naim rips and those of Ruby Ripper; nor can anyone else I have tested this on.

The issue for me is simply one of usability. The NS01 is a joy to use, and since buying it my CD purchases have gone up at least tenfold.

M

I've been on a bit of a CD buying spree as well since getting the NDX/555PS/nDAC.  I did my uncompressed FLAC vs WAV comparisons in a very simple way : Asset on my laptop with FLAC set to output as WAV, and the same filesFLAC direct from Twonky on my QNAP NAS.  Still can't tell any difference, and I've tried a few more  recordings since my last post.  I had previously tried the comparison using uncompressed FLAC and WAV on USB sticks directly into the nDAC - again no difference that I could discern.  Perhaps I'm just cloth-eared - although I doubt it.

Guy, the bitstream, or more accurately the PCM sample data from a decodedwav file and FLAC (assuming no software errors) will be identical. The difference we hear is the noise on the sensitive analogue and possibly digital click circuits caused by the decoding of FLAC,which requires more processing cycles to decode than WAV, hence more noise.

You could argue this noise should  be completely decoupled from said circuits, and perhaps it should, from what I hear from Linn users it might be the case, although the KDS was never my cup of tea with WAV or FLAC, so I can't really say.

We see there is greater decoupling and screening of the streaming board in the NDS, so these issues and RFI issues might be significantly reduced with the NDS. Time will tell.

However transcode FLAC to WAV using your upnp server.. You get the space saving benefits of FLsupport ore widely supported id3 tags, and the Naim is happy. Metadata is sent by the upnp server.

Finally Garry, geeky words? Possibly, however I am a geek and immensely proud of it, and I maintain if software can't read a file properly it's the software's issue not the file's.....At work for compliance recording analytics and home audio I have no problem reading WAV metadata (listinfo) other than with Twonky and iTunes.

Simon

Simon,

Personally, I am greatly indebted to your "geekiness". I have learned much by reading here.

On what is ripped where: my results are clear:

On either Apple TV or NDX,

a) WAV sounds better than FLAC

b) WAV ripped via HDX = WAV ripped via DBPoweramp

c) Tagging WAV is straightforward using DBPoweramp

d) Things ripped by itunes did not give me good quality

Of course NDX makes Apple TV sound puny

Hi DQ

How are playing FLAC through your Apple TV? 

All the best, Guy 

FLAC converted to ALAC!

Hi Simon

I can't measure the processing difference on the NDX, but on OS X - iTunes playing ALAC or WAV or AIFF all put a light load on resources on the Mac - playing the Sensational Alex Harvey Band's Swamp Song - WAV coming in at 6.9%, AIFF at 6.8% and ALAC at 7.2% average CPU load and I couldn't detect any difference with my ears. I tried it twice and the results were consistent. The memory usage didn't change, WAV was more disc intensive caulsing the total load on the CPU to be higher than ALAC and about the same as AIFF. I know this means nothing in the context of the NDX. Though I'm happy to agree the processing required is about 0.3% more for ALAC than WAV. 

I really can't hear a difference though; I though AIFF sounded slightly better first time, but second time I thought no .. it's the same. 

I don't know a way of transcoding to WAV on the fly from any of my systems; I know Asset can do this, but I can't run that until Spoon does the promised Linux port. I'd need a UPnP that did this and ran on Linux or OS X. 

The NDS was superb at the demo, but if makes a pig's ear of playing my files then .... well what's the point. 

If both Meridian and Linn can achieve great results with FLAC then surely Naim can - I hope you right and the use of fibre optics in the NDS enables it to play FLAC as well as it can play WAV. 

Perhaps I should forget the science and just listen and go with which ever player works best with the Vortexbox. 

Still one thing everybody agrees on nowadays is that there is no such thing as a magic ripper  

Perhaps my inability to hear the difference is down to record choice, but I'd sooner listen to Shirley Collins, Sandy Denny, the Beatles, the Kinks, ELO, ELP and HMHB than audiophile music. If only Barbara Thompson was available in high resolution. 

Glad to hear there is another Sandy fan on the forum (I know there are quite a few), as she is as good as it gets for contemporary singer songwriters and the world lost a great deal when she left this world. 

All the best, Guy 

Thanks DQ - ALAC and WAV sound the same to me on my ATV, but I'm not playing it through anything special. 

You have been unlucky with iTunes (do you use a PC?) - iTunes rips on a Mac that I have compared with some from dBPowerAmp have been 100% identical in terms of PCM so give just the same sound quality - there is no way of telling which PCM came from what ripper. I didn't do the dBPA rips, as I can't run it, but I did XLD, CD-Paranoia and iTunes rips and the PCM was the same every time. Nonetheless they were all as good as one another both subjectively (according to my ears) and by comparing the bits in files - Naim UQ also creates identical PCM. 

XLD is probably the best ripper I have if you want it to check with Accurate Rip and so on. 

dBPA is good - I like it, it just WAV files that I don't like, not because of the sound, but the way they handle metadata. 

Correct...

...and in the NDS the decoding is able to be decoupled from the playback more than it is in the DAC.

Phil