NDS with 555ps vs Klimax ds

Interesting post Cris.

I've made a similar comparison few weeks ago at my local dealer

KDS > 252 > 2x250.2 > Ovator

vs

NDS +555PS > 252 > 2x250.2 > Ovator

Selected music for the test

Joni Mitchen:Shine 

David Sylvian: Secrets Of The Beehive

K.Jarret: Paris Concert 

Keith Jarrett Trio: Standards in Norway

G.Gould: Goldberg Variations

Brad Mehldau Trio: The Art Of The Trio, Vol. 3: Songs 

Ben Harper & The Blind Boys Of Alabama: There Will Be a Light

Alfred Brendel: Mozart: Piano Sonatas #10, 11 & 16

Cassandra Wilson: Traveling Miles

Charlie Haden & Pat Metheny: Beyond The Missouri Sky

Ensemble Mare Nostrum: Bach: Orgelbuchlein

Isabelle Faust: Bach: Sonatas & Partitas for solo violin

Morelenbaum & Ryuichi Sakamoto: As Praias Desertas (Casa)

Mstislav Rostropovich: Bach_Cello Suites

William Carter: Fernando Sor Early Works For Guitar 

I have to say that both players were excellent.

Really excellent.

But to my ears,

in that room,

that day,

the NDS was better: 

NDS+555PS was more exciting, it had more timing and more PRAT.

I think that as always, at this level,  is a matter of personal taste:

For example, I've made long listening session comparing

CDX2+555ps

vs

MacBook pro + AudioNirvanaPlus + Antelope Audio ZODIAC+ 

TO MY EARS, in a fast A/B comparison there were no difference. Nothing:

Antelope Zodiac+ is an amazing dac !

But in a long listening session the CDX2+555ps gave me more pleasure,

had more fluidity,

was more exciting...perhaps, "just " more NAIM.

that's it.

The NDS is on his way.. looking forward to listen to it in MY room

Orfeo 

( ex Buk aka Bo ) 

"NDS+555PS was more exciting, it had more timing and more PRAT."

Yep, I get this, which is why I now have both... 

But if Peter and Phil (I guess more Phil than Peter ) instructed uPnP server to transcode FLAC to WAV then both KDS and NDS were in the same conditions.

IME, FLAC served up as WAV does not make it sound like a WAV. if anything, it made all my WAVs sound like FLACs. Bleah!

Julian, I don't hear the difference between the bitstream originated from WAV and the bitstream which has been transcoded to WAV from FLAC on the PC/Asset uPnP. As soon as PC is powerful enough to do transcoding on the fly -- the result will be the same.

Just ot avoid misudnerstanding between us: if you compare a WAV file to the compressed with FLAC and then transcoded back to WAV -- do you think

1) those two will be the same bit-to-bit?

2) will they sound the same?

My server should be well up to the task. Highly over specified and running on tickover when serving music, even when serving FLACs as WAV's on the fly. Regarding the bits, of course they should be the same. All I know is I can hear a difference and therefore I have converted any FLACs to WAV for playback. Instant delight! If you can't tell the difference, you are lucky since you will not need to transcode.

I'm also in the "lucky, can't hear the difference camp" also I struggle to rationalise how it would be possible to hear a difference in this situation. For a difference to exist, there must be a change in the way the data packets are presented to the streamer - I suppose this is possible, as the transcode step requires some buffering and processing so maybe the resulting packets are a different size (smaller) so there is more reassembly required in the streamer? Simon once postulated that TCP window size could have an audible impact - maybe this is a similar phenomena. 

+1.  I too struggle to understand how WAV could be different than FLAC transcoded on-the-fly to WAV.   I have tried but cannot hear any difference between the two, but I will accept the possibility that my inability to understand how they could be different may actually be coloring what I hear.   Listening for subtle differences is tricky in the best of cases -- sometimes I hear 'em, and sometimes I don't.  Also, a while back I asked Mrs. Hook (whose hearing is much better than mine) to listen for differences between these two streams, and she too could not detect any.

I think we all agree that FLAC compression is lossless, and that when it comes to files, the FLAC-to-WAV-to-FLAC conversion process works just fine.  We end up with exactly the same PCM data that we started with.  So I think it makes sense then to try and understand more clearly the actual process of transcoding on-the-fly to see if it has any potential for creating differences.  Mr. Spoon, are you out there?  If yes, could you comment on Asset's transcoding logic?  Thanks!

To begin with, I do not believe that the entire file is first being transcoded, and only then being sent.  My understanding is that the transcoding occurs in chunks, so maybe that process somehow introduces some differences?   It is hard for me to understand how the operating system and/or the network device driver would treat these two streams differently, as the transcoding (as far as I know) is taking place up at the application layer.  Perhaps Simon or another engineer could chime in with some thoughts here?  Hope so...

Also, I wonder if there is a way (perhaps using a network sniffer?) to capture and compare the two streams?

Hook

Notwithstanding the technology, the differences are not "subtle". Its the difference between wanting to switch it off because I am bored vs musical engagement.

Hi Julian -

To be clear, am not for a moment questioning what you are hearing, and you are clearly not alone in this regard.  Likewise, all I can do is report what I hear.  Maybe it is something in how our setups differ?  Who knows.  Just think it would be interesting to find out if there is any difference in the PCM data being seen as input by the streamer, or if we should be looking to other setup-related differences to explain what we are hearing (or not).

Hard for me to imagine that the extra processing taking place 50' away on my Asset server is somehow causing EMI/RFI noise to travel over the ethernet, through my switch, and into my streamer.  But anything is possible I suppose.  Just for safety's sake, I do follow Simon's recommendation and am using a couple of ferrite chokes on my ethernet cable, right next to my streamer's ethernet port.  Am curious to learn if you are you doing the same?  Thanks!

ATB.

Hook

Julian, what's your streamer?

I didn't know there was difference blissfully ignorant somewhat when I converted to streaming.  Ripped all to wav.  Had difficulty tagging.  So PAINFULLY ripped again to 2 different lossless formats.  Only to find sound quality flat.  Toggled back to wav files.  Significantly if not substantially... 

For special listening, I use wav.  For normal listening, I use aif.  Compromise given tag vs sound quality...  

Try Barb Jungr - free Naim hi rez download.  Rip is wav then aif.  Gulf between them...

I wouldn't have known diff if I hadn't experienced for myself somewhat inadvertently like that.

Streamer = humble SQB touch into nDAC.  Yet some seem to hear no difference using SQB.

NDX

Nope, not here. I didn't think you were doubting what I can here Hook. No probs.

From Hook:

"Also, I wonder if there is a way (perhaps using a network sniffer?) to capture and compare the two streams?"

One day when I have some time and the inclination I may well 'shark this, but right now I'm just enjoying the rather fine sounds from my NDS - Asset transcoding FLAC to WAV....

I also have a ferrite on my ethernet cable, plus one on each of 3 Powerlines. Perhaps that's the answer?

Well particularly with Hook, Simon & Guy I have the feeling of deja-vu because we have been round this discussion before. I guess you will recall my stance has always been that I prefer WAV over FLAC. I also have now had occasion to compare ALAC and AIFF and again I prefer AIFF to ALAC. 

Interestingly I did convert some WAVs into ALAC's recently to play through i-tunes/Bit perfect on my MAC via a USB -I2S interface DAC and felt the AIFF sounded as good as subsequent listening with WAV via a PC using WASAPI to feed the same USB - I2S DAC.

regards

Geoff

Geoff, you are right we have discussed this many times before, and yes just use what you prefer, with upnp servers like asset, streaming WAV or FLAC is a simple mouse click.. No sweating  about what format your 15,000 ripped tracks are in..

Btw AIFF and WAV should sound the same when played back in the real world, the file formats came from the same source many moons ago, and simply WAV has the reverse bit sequence order to AIFF for each sample, other than that they are largely the same.

Simon

I also note a useful increase in quality with AIFF over ALAC, and also had some WAVs in ITunes that I also converted to AIFF for consistency. No change was notable between WAV and AIFF, and the conversion took about one second per track [average of four five minutes of music each] compared to some quite considerable time converting from ALAC files to AIFF.On the fly conversion of ALAC must require quite a lot of CPU effort!

I do have a thread running on this in the Streaming Audio section.

ATB from George

Yes - just use the one you prefer .... on my system they all sound the same to me. Whether they would to somebody else I don't know, whether they would on another system I don't know. 

I have seen a test where a WAV file was compare with a FLAC file ... by capturing the output to Audacity and inverting one of the outputs and adding them to get a straight line at 0. Please don't ask me to find it again ... I think I posted links once, but can't remember what they were or find them. 

The argument at the time was that it didn't prove anything .. it was merely an analysis of one file. 

Neither my Vortexbox nor my Mac can transcode - though ports of Asset and J-River to OS X are interesting. 

Has anybody noticed when you record a digital file on to cassette and then play it back it sounds more analogue?  Perhaps when should all be analogising our digital music collections. 

Gentlemen -

Please note that I was commenting specifically on 1) WAV versus 2) FLAC transcoded on-the-fly to WAV.

I can understand when someone says they can hear a difference between FLAC and WAV, or ALAC and AIFF.  The theory is that a streamer is doing more or less work, and that more work equals more noise.  I have no problem conceptualizing how these differences may occur, regardless of whether I can hear them myself.

What I am having trouble understanding is how WAV can sound different than WAV. I do not understand why a WAV file should sound different than a FLAC file that is being transcoded to WAV on-the-fly.  The only thing I can think of is that the transcoding process must be broken.  Unfortunately, unlike file conversion, the accuracy of on-the-fly transcoding is not as easily tested.

If someone has already explained how this specific difference is possible, then I would appreciate a pointer to that explanation.  Also, if that explanation has already occurred, then I apologize for being repetitive and will exit the thread.

Hook

Hi Guy, an interesting test would be to use a good quality recording and mic system and record different systems playing back WAV and FLAC or ALAC, AIFF etc. Average over a few recordings and then invert and compare... Might be interesting. I know aligning the recordings will be challanging.. But I'm sure it could be done.. Food for thought... Any undergraduate wants a thesis idea?

BTW is it a full moon tonight? My music is sounding exceptionally enjoyable this evening.. And not a drop of alcohol in sight

Simon

Hi George -

The context of this discussion is UPnP, and so it is different from your setup using a local hard disk for playback.

The transcoding of FLAC to WAV takes place on the remote UPnP server, which can be quite a distance away from the network player.  In theory, the network player should be seeing the exact same PCM bitstream whether it comes from a WAV file or from a transcoded FLAC file.  But some folks are hearing a difference, so that what is what I am trying to understand.

ATB.

Hook

Hello Simon 

Please google "fordgtlover Here is the difference between WAV and FLAC" (posted on George's thread too). You may find it interesting. 

All the best, Guy 

I went to CES a few of years ago and there was a room there full of gloriously coloured Reel to Reel tape decks, lovingly refurbished:

The guy responsible was playing music off tape with one of the decks that sounded superb. It was quite recent music so I asked hime where he had got his hands on a taped version and he replied he made it himself by recording to tape on a Reel to Reel off CD.

I was dumbstruck how fluid and analog sounding the tape music was. If I had the money and a mode of transport from Vegas to home I would have been tempted to buy one of those Reel to Reels there and then.

Geoff

> What I am having trouble understanding is how WAV can sound different than WAV.

Hello Hook

If I have two CDs of the same recording is it possible they could sound different. [I don't know]. Some people claim if they burn a CD-R then it can sound better than the original CD especially if you use a Yamaha burner. 

I don't understand if the same digital stream is coming out of a computer in to a buffer in a DAC and being re-clocked and played from the buffer how it can possibly sound different ... the DAC has the same stuff in its buffer however it got there.

There are lots of things I don't understand ....

Do you here a difference between WAV and WAV? Perhaps the transcoded WAV is not the same as the vanilla WAV and the receiving streamer has more work to do to strip out the PCM?

Does a file with embedded artwork sound as good as one without?

In the end if you can sing along then I think that may be all that matters ... altogether now, one, two, three, four, can I have a little more, five, six ..... 

All the best, Guy 

Amazing ... ages since I heard a reel-to-reel, but used to love the sound. 

I did make some cassettes on my Dolby S cassette deck and played them back at a party and got told how good they sounded and they were all taken from CD. I think the tape medium had a lot to offer. 

All the best, Guy