ND5XS vs nDAC

Posted by: Lightkeeper on 02 January 2012

Hi,

 

just interested did anyone compared ND5XS and Naim DAC regarding sound quality?

 

thanks,

oz.

Posted on: 09 January 2012 by RudiNee

Hi Guys,

 

Nice to receive this feedback about decoding. Since I have no doubt in believing what you are saying I really wished I knew this from the beginning. Imo iTunes is also the best software to use and I kind of regret using FLAC in my little apple world. I'm not going to convert all items again

 

What I also mentioned was that I recently switch from an old Twonkey Media version to the latest one.My SU is connected via the router to my NA. This really changed SQ a lott. All got clearer, more detail and more soundstage. Can anyone explain this? Because I really can't. Nothing else changed.

 

To the OP and also to Iver: I would definitly go for the comfort of the ND5XS. You can still upgrade the machine with an extra PS and do not need extra cabling and an extra streamer. Maybe in a 1-1 comparision the DAC is better but I'm sure you are going to like the 1 box solution much much more. Also the tuner and iradio feature is already worth this choice.

 

Cheers,
Rudi

Posted on: 09 January 2012 by Aleg
Originally Posted by RudiNee:

Hi Guys,

 

Nice to receive this feedback about decoding. Since I have no doubt in believing what you are saying I really wished I knew this from the beginning. Imo iTunes is also the best software to use and I kind of regret using FLAC in my little apple world. I'm not going to convert all items again

 

... 

Cheers,
Rudi

 

Hi Rudi


Why not?

 

If you have access to a  PC and use dBPoweramp you can run a batch convert that will convert all your files from one lossless format to another lossless format in one go.

 

-

aleg

Posted on: 09 January 2012 by RudiNee

Ok ... I think XLD has this feature also. Just some questions than ...

 

- What is the best format to convert to, when using the superuniti, ALAC?

- Does this not compromise the quality of the previous rips?

- What sourcefiles does the naim like best (aside from wav which is really to large)?

 

I than can use iTunes as the engine and link it to my NAS for storage and back-up.

Posted on: 09 January 2012 by Aleg
Originally Posted by RudiNee:

Ok ... I think XLD has this feature also. Just some questions than ...

 

- What is the best format to convert to, when using the superuniti, ALAC?

- Does this not compromise the quality of the previous rips?

- What sourcefiles does the naim like best (aside from wav which is really to large)?

 

I than can use iTunes as the engine and link it to my NAS for storage and back-up.

 

- What is the best format to convert to, when using the superuniti, ALAC?

- What sourcefiles does the naim like best (aside from wav which is really to large)?

Don't know what SuperUniti or Naim likes best. I don't own a SU or Naim streamer.

 

- Does this not compromise the quality of the previous rips?

No, lossless formats can be converted into each other without loss of data.

 

-

aleg

Posted on: 09 January 2012 by Guido Fawkes
XLD can convert FLAC to ALAC. If you then convert the ALAC to FLAC then you are back to where you started. Nothing is lost. I still can't hear any differences in my setup with WAV, FLAC, ALAC or AIFF. However, there is a definite increase in quality with 24 bit files over 16 bit ones (night and day). I also can hear no advantage of using non iTunes players on a Mac, although bit perfect does optimise the settings and switch sample rates automatically which is very nice.

I cannot hear differences between Western Digital, Hitachi and Seagate disks though so I'll never get a job as a HiFi Critic.

My Naim UQ plays both streamed FLAC and WAV. I couldn't tell which format the file was in just by listening.
Posted on: 09 January 2012 by Simon-in-Suffolk
Aleg, with regard to your question re. Flac and wave decoding. Yes the decoder engine has to unpack the PCM from the wav or flac, irrespective how it's delivered. Therefore if the file is decoded from memory as opposed on the fly with a TCP session I would expect the differences still to be present. Simon
Posted on: 09 January 2012 by Simon-in-Suffolk
Aleg, I have re read your question on a bigger screen :-)
If the FLAC decoder is not faulty, then once decoded into uncompressed PCM and stored in a  WAv, then if it's bit equivalent to the PCM before it was encoded as FLAC there is no difference. There are no other variables associated with the PCM.
I for one store much of my hidef as FLAC but real-time transcode to WAV by my upnp server for my NDX. I can hear no difference at all between native WAv streamed and FLAC transcoded to WAV streamed. ( and I do tend to hear things that many don't notice for my sins)
I have also converted WAV to FLAC and back again using dbpoweramp and performed a bit comparison and there was no difference.
Simon
Posted on: 09 January 2012 by Hook
Hi Simon -

Like you, I could not hear any difference between FLAC and FLAC transcoded to WAV on the fly.  But I left the transcoding on because it makes the UPnP service more universal.   My A/V receiver only understands WAV and MP3.  My Apple devices only understand Apple formats...and WAV.  It's just easier this way, and I no longer have any compatibility issues.  But I retain the ease of management using FLAC tagging.

Hook

PS - Finally got around to upgrading the NDX firmware today....hooray, a timer bar in N-Stream!
Posted on: 10 January 2012 by Aleg
Originally Posted by Simon-in-Suffolk:
Aleg, with regard to your question re. Flac and wave decoding. Yes the decoder engine has to unpack the PCM from the wav or flac, irrespective how it's delivered. Therefore if the file is decoded from memory as opposed on the fly with a TCP session I would expect the differences still to be present. Simon

 

Originally Posted by Simon-in-Suffolk:
Aleg, I have re read your question on a bigger screen :-)
If the FLAC decoder is not faulty, then once decoded into uncompressed PCM and stored in a  WAv, then if it's bit equivalent to the PCM before it was encoded as FLAC there is no difference. There are no other variables associated with the PCM.
I for one store much of my hidef as FLAC but real-time transcode to WAV by my upnp server for my NDX. I can hear no difference at all between native WAv streamed and FLAC transcoded to WAV streamed. ( and I do tend to hear things that many don't notice for my sins)
I have also converted WAV to FLAC and back again using dbpoweramp and performed a bit comparison and there was no difference.
Simon


Hi Simon

 

Thank you for your replies.

 

I think though your first reply was more to the core of the issue than your second reply .

 

Could you explain a little bit more about what would be causing this lingering effect of the decoding fase?

 

Just to recap:

it is about a software player (JPlay) that

- for wav files loads the wav into RAM memory and unpacks it to PCM in RAM for playback;

- for flac files loads the flac into RAM memory, decodes it in RAM memory and plays back the PCM from RAM memory. So no writes to file anymore, everything is handled in RAM memory.

 

The playback is directly into an audio device attached to the computer and not using UPnP-streaming.

 

Some people report noticeable differences between playback of WAV and of FLAC.

But for both formats the files are being completely decoded and unpacked in RAM memory before playback of the resulting PCM is started. Any differences would than have to carry over from the time of unpacking/decoding to the moment of playback.

 

I was hoping to tap into your engineering background to see if a plausible cause could be pinpointed that would be behind the perceived difference in this specific situation.

 

-

aleg

Posted on: 10 January 2012 by Simon-in-Suffolk

Aleg,

Puting the pressure on now ..

The process of ceating the PCM from unpacking it from a file (FLAC or to a lesser extent WAV) can cause the perturbations. If his is happening at the same time the PCM is beng clocked into SPDIF, I2S etc then these could be side effects that are audble.

Howeer if this decoding happens and completes, and then the PCM is recovered from memory  and packed for playout (SPDIF etc) then I can see no diffrence, and if the PCM is identical and there is no other backgound process that is different or causing perturbations between the two playbacks, then by definition of information theory entropy the two files,  data / information are idenicaland so will sound the same.

Therefore if there is a diference between the two then  there is some other related process ascoiated with the decoding outsde of the PCM itself that is causing the difference.

Simon

 

 

 

Posted on: 10 January 2012 by Simon

Hello world

 

ND5 XS vs DAC

 

4 weeks listening... without an XPS involved the DAC beat the ND5 XS by a margin of 7-8% (probably worth the sacrifice of ND5 XS features for outright performance).  Put an XPS on them in turn - the ND5 XS benefits most, changing the tables at times; when averaged with the availability and quality of content/recordings... "Usefulness"?, the ND5 XS is a bargain.

 

For digital sources I use:

 

Modified Trichord Research CD/Transport and Chord Prodac 1m

MacBook Air 11" (mid 2011) + M2 Tech hiface USB and Chord Prodac 50cm

Homebrew Apple TV2 and Chord Optichord 7m Toslink

Sandisk Cruzer blade 16Gb USB "Drives" with purchased, sample and pwnd digital content

Apple iPhone 4 running latest legit firmware and atypical iTunes content using genuine white USB dock cable

 

USB sticks ultimately sound best with both units, especially with purchased or known source/quality content - even at lowly 16bit 44.1Khz or 48Khz.  I honestly think both devices sound best at 24/192, although the DAC goes upto to 768Khz it doesn't do it for me - sounds compromised/fake, no content available and insane storage rates (currently).

 

M2 Tech is remarkably good, my only option given Naim's choice of USB support and MacBook Air diet costing the digital audio out.

 

I'd love to keep the DAC and ND5 XS.... but, it's the DAC they'll pry from my hands this time round 

Posted on: 11 January 2012 by Aleg
Originally Posted by Simon-in-Suffolk:

Aleg,

Puting the pressure on now ..

The process of ceating the PCM from unpacking it from a file (FLAC or to a lesser extent WAV) can cause the perturbations. If his is happening at the same time the PCM is beng clocked into SPDIF, I2S etc then these could be side effects that are audble.

Howeer if this decoding happens and completes, and then the PCM is recovered from memory  and packed for playout (SPDIF etc) then I can see no diffrence, and if the PCM is identical and there is no other backgound process that is different or causing perturbations between the two playbacks, then by definition of information theory entropy the two files,  data / information are idenicaland so will sound the same.

Therefore if there is a diference between the two then  there is some other related process ascoiated with the decoding outsde of the PCM itself that is causing the difference.

Simon

 

 

 


Hi Simon

 

Thank you for replying again.

I shall release you from this issue (so breath out and relaaaaxxxxx )

 

I think these (if they truely exist) differences will remain some 'ghost' effect that won't be explained anytime soon I guess.

 

But thank you ever so much for thinking aloud along these mysterious lines of digital audio.

 

Regards

 

Aleg

Posted on: 11 January 2012 by Manu

Another test:

 

IMO, as digital sources feeding a nDAC + XPS2, ND5XS+XP5XS and bare NDX are VERY close. ND5XS alone is sligtly more recessed.

 

Simon in S, I have tried your ferrite idea on DC1:

on a bare NDX, I did not like it, it became agressive.

on the ND5XS+XP5XS combo, it brings the very small rythm (faster rises) it was lacking vs the NDX. Like it...

Of course, it depends on the ferrite size and many other factors, so results can vary, it is not a panacea.

Posted on: 11 January 2012 by Manu

I have seen another difference, this one completely objective: the ND5 is 2 time faster at browsing folders in UPnP than the NDX (96\24 version), both in Wireless. May be due to the new streaming board...

Posted on: 11 January 2012 by Lightkeeper

Excellent Simon and Manu, thank you.

Posted on: 11 January 2012 by Simon-in-Suffolk
Hi Manu, interesting comments. The only thing a typical ferrite choke is doing is impeding radio frequenciy currents. For coax its RF currents travelling along the sleeve, and cables like UTP (Ethernet leads) its common mode RF, ie currents passing equally along all conductors. If there is a sonic difference then you have impeded the RF. it it's a negative expierience in terms of preference, then the presence of RF distortion is helping the sound for your ears / set up. RF causes intermodulation and negative feedback distortions in audio frequency circuitry which can cause various unexpected side effects. The fact there is a change shows the presence of RF currents.

Allen, I found adding a choke around the NDX network cable helped when using the NDX native. If the choke is too big, you can loop the cable through the choke, this also increases the inductance and increases the effectiveness of the choke.

Simon
Posted on: 11 January 2012 by totemphile
Originally Posted by Manu:

       

         class="quotedText">
        Another test:

IMO, as digital sources feeding a nDAC + XPS2, ND5XS+XP5XS and bare NDX are VERY close. ND5XS alone is sligtly more recessed.

Simon in S, I have tried your ferrite idea on DC1:
on a bare NDX, I did not like it, it became agressive.
on the ND5XS+XP5XS combo, it brings the very small rythm (faster rises) it was lacking vs the NDX.
Like it...
Of course, it depends on the ferrite size and many other factors, so results can vary, it is not a
panacea.



Manu,
If you don't mind me asking, what's your view on the XP5XS vs. XPS2, when supplying the nDAC as an external PS, is there much in it between the two? In the context of a 282/HC/250 based system, that is.

Thanks
tp
Posted on: 11 January 2012 by Jan-Erik Nordoen
Originally Posted by Simon-in-Suffolk:
Hi Manu, interesting comments. The only thing a typical ferrite choke is doing is impeding radio frequenciy currents. For coax its RF currents travelling along the sleeve, and cables like UTP (Ethernet leads) its common mode RF, ie currents passing equally along all conductors.

Hi Simon,

 

If you are impeding the RF currents then are you not also impeding the RF signal traveling along the coaxial (digital) cable ?

 

Jan

Posted on: 12 January 2012 by Simon-in-Suffolk
Jan, in a word no, unless you have a very poorly designed coaxial SPDIF driver that is trying to drive a balanced signal down an unbalanced line.  If you are using a coaxial cable you are impeding the RF currents flowing through the shield. The SPDIF signal flowing as current in the core is relative to the shield, and so is not impacted.
Of course the presence of RF independent of the required SPDIF information  carried in the SPDIF signal (not in the shield)  is a separate issue. That is known as distortion and causes a change to the shape and phase of the carried digital signal, hence if that signal has only meaning with reference to a clock, you are potentially changing its meaning if you can't recover the original signal. ( signal theory entropy).
Simon
Posted on: 12 January 2012 by NickSeattle
Originally Posted by Guido Fawkes:

. . .

 

iTunes is fine for ripping and the only time it falls short is on damaged CDs and HDCD. Whether it fares worse on a damaged CD is opinion. On HDCD, it does a faithful rip whereas dBPowerAmp decodes the HDCD (as I understand it); this means with iTunes you need a HDCD aware DAC, with dBPowerAmp the NAim DAC would realise the full potential of the HDCD even though it is not HDCD aware.  An HDCD will still sound fine ripped with iTunes and played back through a Naim DAC, you just won'r realise the full potential of the HDCD. (I have very few HDCDs). 

 

. . .

Guy!  This is brilliant!  I have a very few HDCD's myself, but would just as soon get all I can from them.  For nDAC users, dbPoweramp and ripping to 24-bit seems to be the solution, since the nDAC is not HDCD-aware.  Nice!

 

Listiening now to Droplick Murphys, "Sing Loud, Sing Proud" at full HDCD strength via bare Squeezebox Touch as I type this.  (Thanks to Parallels and Windows 7.)  This capability makes dbPweramp indispensible to me, even if I were to run a VortexBox.  Who else competes with a similar feature, I wonder?

 

My nDAC is a week away . . . damn.

 

M a n y   t h a n k s !

 

Nick