Bristol Show Demo and Ripping

I did mention this on the Bristol Show thread in Hifi Corner. Basically the last part of the Naim main dem was comparing 2 files of the same music on a usb stick - one ripped to WAV by a Unitiserve and the other ripped to AIFF by a Mac. This was plugged into the front socket of the streamer.

I fail to see the point of this as ripping to AIFF and then copying to a usb stick is not something that makes a lot of sense. If you wanted to compare Unitiserve rips with computer rips you should  either rip to WAV on the Mac or stream from the Mac transcoding to WAV on the Mac as you go which is what I do (using a PC instead of a Mac). 

Others who attended the show were also unimpressed by the dem judging by posts on that thread. Nobody has yet posted giving a defence of that part of the dem.

There are two possible aggregates of a rip:

a) bit perfect

b) not bit perpect

If we assume that both a Naim ripper and a PC/MAC are capable of producing a) the results will be identical to the last bit.

It would be the slimiest snakeoill ever, if anybody claimed that there is an audibile or any other difference between the two.

I agree with you Pink Hamster and it would seem that so do a good few forum members. The demonstration seems to have upset some of the people who attended.

If the demonstration was carried out as described then you may as well be comparing an orange with a onion (or any two very dissimilar things, the choice is yours).

I agree that Pev's suggested demo would have been more convincing. I concluded that Naim think these differences exist but they chose an extreme example to make the point In a noisy hotel room. I do not think they were presenting non-existent differences as cynical attempt to sell more Unitiserves. 

if Naim think these differences exist, we should not lightly dismiss the possibility that it might be true.

i suppose this is a partial defence :-/

Keith

I was not at Bristol, but Doug did a similar demo at an event a while back and made the same claim. The second play sounded much better...., but whenever I play any unfamiliar track a second time it always sounds better to me. I've put this down to a psychoacoustic effect as I'm more familiar and start to anticipate rhythms and patterns and get more out of the experience.

Was the 'better' track at Bristol always the second one played?

I've seen other people claim (elsewhere) that the quality of the rip can be affected by (for example) powering the CD ripper from a linear PSU vs the computer's SMPS and that this quality can be kept when transferring from one device to another via memory stick or internet transfer.  To my mind this is impossible.

My purpose in asking about the demo was to understand if Naim were trying to propagate this claim... (I need to read the Bristol Show thread)

As I recall, in various threads Naim employees have stated both that the UnitiServe (and HDX, etc) produce rips "as good as it gets" while acknowledging that a properly set up computer with EAC or dbPowerAmp can get just as good results; and at other times claimed that somehow the UnitiServe / HDX get results better than any PC setup.

Any demo like this is open to suggestion ... "we're going to play two samples, one from the UnitiServe rip and the other from the Mac; see if you can work out which is best" ... so unless very carefully controlled is open to spurious claims.

Eloise

Oh yes, and I would have been astounded if they had played the WAV first, but I don't think we should be too disappointed by that. What would you do?

Also, why didn't the disappointed and disapproving speak to Doug and the crew about their concerns? The best thing about Bristol is you get speak to the key people at Naim and the other manufacturers.

Keith

Now let's see..

Hypothesis - computers are infallible. Copying a digital file 'perfectly' will give exactly the same result, no matter what you use to do it.  If correct, why would using a usb stick to remove any differences in playback be an issue?  It wouldn't. It would be eminently sensible.

Hypothesis - computers only record 1s and 0s and if they aren't exactly correct the file is rejected.

Well no, many a browsing session will bring up a "Done but with errors on the page" message. Or open an Office document and get the odd glitch. It's still working but not correctly.

Hypothesis - bits are just bits they can only be in one of two states.  Again not exactly correct, otherwise quantum computers would not be under development. And 1s and 0s is just an analogy to try and explain how a computer works. It isn't that brilliant an analogy.

Any encoding/decoding requires energy. That energy is electrical. Now if we are saying that electrical energy cannot affect a music file will someone please explain to me why the CD555 needs a separate power supply.  It reads the disc using a laser, which according to one prolific forumer, being optical, couldn't possibly  transmit intereference. After all when it reaches the DAC all the 1s and 0s will get exactly realigned in their prisitine original state by the buffer.  The DAC will zap all electrical nasties and the music signal will pass on its way completely unaffected.

Yes when I look at it logically it all makes sense. I don't have sufficient knowledge on acoustic theory and electronic engineering to explain exactly why there are differences in ripping so it can't possibly be true. Stick to the textbook, or that old standby Wikipedia. That's the ticket.

Yes if I don't think outside the box I'll not have to consider if 'string theory' will prove that the theory of relativity is wrong. And there can't possibly be a fourth dimension because no one can prove it. Curved space? black holes? other planets? say it isn't so. I might have to go back to religious conviction.

Now please excuse me I must lie down before the pink elephants get any closer.

Andy

Apologies to anyone of a sensitive nature who feels that my rant is unjustified. Only kidding, but apparently if you sign off with some platitude you can be as nasty as you like.

Hi Eloise,

Demos at exhibitions are no substitute for proper demonstrations at dealers or at home. 

Also, it was really obvious, too obvious if anything, but they no doubt chose it because they thought we would all get it. The only reason they're getting abuse is that so many people think it should be the same. Naim are telling us that they have found audible differences between rips and they are trying to  make theirs as good as possible.

i don't know as I've never bothered to do the test myself.

Keith

My purpose for asking the initial question was to try to understand what Naim were attempting to demonstrate as (elsewhere) the Naim demonstration was being held up to support a controversial view about ripping...

If the comparison was WAV vs AIFF then there is at least 2 different programming pathways to take into account which may explain the difference.  Or a collective "hypnosis" of those listening...

Hi Elouise,

I agree. I'm sure people challenged them on this point after the demo. Unfortunately, none of those people post on the forum and (last time I looked on the other more crowded thread) Nam have not addressed the issue here.

Personally, I'm happy to believe it as I own a uServe :-)

Also, while I haven't done the PC verses US rip test myself, I have heard differences doing similar tests which are widely regarded as impossible on this forum so I'm keeping an open mind. Plus, I'm sure Naim wouldn't raise the issue if they hadn't done more controlled tests in their R&D department.

Keith

We seem to have the annual resurgence of this. About 18 months ago several of us categorically demonstrated that Unitiserve, dbpoweramp, ITunes* and EAC produced identical rips encoded as WAVs, to the point of better than 1 error in 300,000,000 in the sample data. * iTunes occasionally produced an offset of a few milliseconds at the start and end end of rip, but other than identical.

However and  this is where the confusion seeps and the sales people lean on the FUD as it is NOTHING to do with rips them selves, the format the PCM is encoded as does sound different in Naim equipment due to the sonic signature caused by the EMI and Powerline modulation caused by the Naim file format decoder. In the ideal world this wouldnt happen, but in the real world it does.

Now AIF and WAV are almost identical other than the sample values are reversed, so the decoding algorithm would be different. It should be slight and not as involved as FLAC, but I dare say different none the less and so the sonic signature is different. 

We do know that in the network players such as the NDS careful tuning of the firmware is performed to minimise sonic changes, therefore it i feel it is fair to assume that adifferent file format will upset the tuning.

Therefore with Naim I always say feed in WAV files. These days for streamer servers transcoding to WAV is trivial.

But people please remember PCM file formats and how they sound in Naim equipment is different from ripping.

If any newcomers are interested, please search back to the posts where the test and analysis were performed on this forum. Digital powerline modulation and EMI is well understood and established part of design best practice. There is nothing new here, and indeed because of its unexpected (to me) impact in PLA's ( the precursor to micro controllers) was an unplanned major aspect of my undergraduate computer engineering dissertation back in the late 80's. 

Simon

Hi Jude, only Naim can categorically confirm, but in my expierience yes. The nDAC can only handle FLAC (with the later firmware) and WAV I believe from memory stick, and WAV sounds preferable. Other than that it's SPDIF via Toslink/coax and Apple PCM streams via USB.

Sorry, only Mrs S-in-S uses a mac now, perhaps others can comment?

Mixing up ripping and playback and then claiming to be thinking 'outside the box' is not such a brilliant approach.

Nobody said that playback fron a USB stick couldn't differ from UPnP streaming. And there is also a technical explanation for that. It is the way the DAC will deal with the jitter.

But a file is a file. If you think 'outside the box' on this you get a corrupted file. This is a very straight forward fact and not an issue of opinion or subjective listening.

No, this is not true. You must have misunderstood something somewhere.

Jude, PCM is simply sample data, it needs to encapsulated and encoded somehow to be sent anywhere such as with a CD rip, and if you are going from one encoded file format to another that is when one transcodes.

If you streaming the data via SPDIF, such as converting from AIF to SPDIF for example , strictly speaking you are transcoding, as SPDIF has its own way of encapsulating and encoding PCM sample data.. But the term is not usually applied to SPDIF, or asynchronous USB.

Simon

Jude, yep that will support asynchronous USB as well. To be honest you don't need to know half this stuff, after all it's all consumer software and largely plug and play. I guess the only time you need to tread into the 'mine field' is if you are insisting on doing something a particular way or you want to maximise the performance of something... No different than with a lot of things really.

Simon

What does the 'dem' prove? That the naim equipment on dem can't play AIFF files properly? Was anyone given the opportunity to investigate whether the AIFF file was correctly ripped - ie bit identical to the WAV file when upacked? How about putting these two files, and the source, up for public scrutiny?