Mac Mini vs Naim ND5 XS

Hello to all

I would like to know if any of you have had the opportunity to evaluate the differences between Mac Mini and ND5-XS

My configuration: MAC Mini (audirvana-Tidal) connected with USB Asynchronous> DAC V1> NAC 202 (HiCAP + NAPSC)> NAP200.

Currently, as seen from my configuration, I'm using a Mac Mini as a streamer via Audirvana connected to the DAC-V1. I can not complain about the surrender but I'm not totally satisfied with it.

I wanted to understand if taking an ND5-XS instead of the Mac Mini (always connected to the DAC-V1) could bring improvements.

My budjet does not allow me (nor now, nor after) to evaluate other options ... I know that NDX would be better ... and better still the NDS ... and so on ...

My only way and possibility is to evaluate the ND5 and nothing more ...

I wanted to understand if I would improve or not worth it

Thanks in advance

Original Post

Well in the general sense, I suggest one keeps away from computers... computers and high quality audio, especially when digital to analogue conversion is involved are not good bed fellows. The main issue is ground noise and clock stabilityand resultant  intermodulation noise from the clock. A typical computer only needs its clocks to be generally stable... one just doesn’t need to make the clocks extremely stable and as that takes space, adds extra componentry  and adds extra cost, it therefore usually is not done.

With high quality digital audio, the clock(s)  stability including transport clocks  ars one of the most important aspects of design. A stable clock produces less intermodulation components with connected components which in turn lowers the digital and analogue noise floors on these related functions.. which means resultant analogue reconstruction will be more natural and contain less artificial ‘additives’. 

angelodipa posted:

...

Currently, as seen from my configuration, I'm using a Mac Mini as a streamer via Audirvana connected to the DAC-V1. I can not complain about the surrender but I'm not totally satisfied with it.

I wanted to understand if taking an ND5-XS instead of the Mac Mini (always connected to the DAC-V1) could bring improvements.

...

I very much doubt that a ND5-XS would bring significant improvements over the Mac Mini: I understand that the Naim DAC V1 is based on a very decent Audiophilleo USB interface with galvanic isolation and reclocking. But  you could always try it, of course. In this case, please report your findings.

What are you actually missing in your current system? If you could answer this question it would be perhaps easier to suggest paths of improvement. Good luck, nbpf

My system is analogous to yours:  Mac Mini / nDAC / NAC552 and NDX-FM into same.  I am using Audirvana, and a Meridian Explorer USB to TOS between the Mac and nDAC.

Sometimes I think Audirvana sounds as good or better than the NDX; but system automation provided by the NDX is so good, I use the NDX most of the time.  It is great.

Because you use your DAC into a NAC, I recommend you try the ND5XS for the best user experience, if not in hope of any significant improvement in sound quality.  If you can, consider buying used to reduce the risk, in case you do not like it.  That is what I have done with nearly every piece.

Nick

I see the DAC-V1 cannot be controlled by Naim’s system automation the way the nDAC can.  While this is a nice feature, not having it would not spoil the benefit, for me.  You still gain Volume control and source-switching at the NAC, plus the Naim App for music play.

If you route all digital sources throught the ND5 first, including the Mac Mini, then ND5 to the DAC-V1, you will have all the control you could want.

Nick

first of all thanks for your valuable contributions ....

so for your direct experience you are telling me that even the Mac Mini with Audirvana (via asynchronous USB to V1) equals the NDX?

So it comforts me ... of control I do not really care ... with the APP of Audirvana I select the music I want to listen to, and with the remote control of the NAC I adjust the volume ... all this for me at the moment is not a problem.

But if I take a used ND5 instead of the Mac Mini, do you suggest me to connect it to the V1 anyway? would I get the same quality and be able to drive everything (both music and volume) with Naim's APP?

angelodipa posted:

first of all thanks for your valuable contributions ....

so for your direct experience you are telling me that even the Mac Mini with Audirvana (via asynchronous USB to V1) equals the NDX?

No.  They sound slightly different, but equally pleasing, to me, through the Naim DAC/555.  Some might disagree.  Both sound much better to me than Sonos does, through the same DAC, any source.  So, it is not that I am not picky.

. . .

But if I take a used ND5 instead of the Mac Mini, do you suggest me to connect it to the V1 anyway? would I get the same quality and be able to drive everything (both music and volume) with Naim's APP?

I would try it both ways and decide for myself (if I were curious).  I doubt all would agree.  I already had the Naim DAC 555 when I added the NDX.  I admit I have not tried the NDX on its own, without the nDAC, yet.  Word here is the DAC555 is better than the NDX555.  I take it on faith.  Not as clear a ruling with DAC-V1, if I recall correctly.

Nick

The Naim app will not control volume on the V1. It will, however, control the volume of your 202 if you set up system automation on the NDX or ND5XS, even if you continue to use the V1 with it.

As for which combination sounds best, that will always be subjective to some extent. It will also depend on whether you buy new or used, as new would mean NDX2 or ND5XS2, and old would mean the discontinued models that they replace.

My streaming started with ND5XS. I upgraded it by adding Hugo as an external DAC, the ND5XS then functioning just as a renderer. That was a significant improvement in sound quality.

i then changed my  basic/cheap and very noisy NAS for a Mac Mini as a silent NAS, following which I compared Audirvana on the Mac Mini as a renderer against ND5XS. Initially it was significantly worse, however I hadn’t allowed for the susceptibility of Hugo to RF interference, which was a lot worse from the MM than ND5XS, the latter being designed to minimise it. I added a Gustard U12 isolator between the MM and Hugo, bringing a dramatic change. Redoing the comparison, MM/Audirvana/Gustard vs ND5XS, the MM was better, though only marginally. 

(N.B I cannot say whether the benefit of the MM setup vs ND5XS was due to any difference in te renderer, or if it was the fact that the MM was a combined store+renderer and not streaming my own music collection across a network, therefore with no influence from network cables, switch power supplies etc.)

I cannot stress too much how important is RF isolation - some DACs reject better than others, and my current DAC (Dave) is better than most - but even with that, its designer is finding additional ways to improve, e.g. with ferrites on cables. Dedicated streamers aren’t immune - they are still computers inside, and by their very nature generate RF, however they are designed to minimise, and remove/spisolate before their DAC stages: and that is very possibly where some better streamers score over lesser ones. I think that where different people have reported different experiences with Mac Mini it is very likely due to differences in effectiveness of different approaches to stopping RF, and on teh susceptibility of whatever DAC is used.

At the end of my initially auditioning Dave DAC, which was done with my MM/Audirvana (excluding Gustard on the Dave), I had a comparative listen to the Melco N1A (an audiophile design essentially equivalent, but optimised internally for music including attention to RF minimisation) against my MM - and there was no immediately apparent difference in sound. (System was Dave DAC, Bryston 4Bsst2 power amp, PMC Fact 12 speakers, in dealer’s listening room.)

For reference, my MM is optimised with a dedicated USB bus, no unnesessary functions or connections operating,  and Audirvana is similarly fully optimied, including direct mode. It runs headless - just a small box, without screen or keyboard), controlled by VPN remote control software from any other computer or tablet, both for setup and AUdirvana control, or Audirvana controlled by its own app on an iPad or iPhone. Also, I use SSDs not HDDs, though can’t say whether any difference as I haven’t compared.

I have been supremely happy with the sound of MM/Audirvana, both through Hugo (with Gustard isolator) and now through Dave - however I do find AUdirvana’s library software frustrating and irritating to use (probably mainly a function of  metadata my music and its associated metadata, as other people seem quite happy with it), and unless Audirvana improves that side of it I will readily ‘jump ship’ when I find a suitable alternative.

Frank Yang posted:

I use MM with Audirvana UPnP out, and I am happy with it. However, your V1 is not a streamer so you cannot use the Audirvana UPnP feature.

I don’t understand this optional function of Audirvana (unless you are using it purely as a means of Tidal MQA first stage unfold?) This software’s real strength is its rendering function, which you lose if you stream the music over your network by UPnP, so all you have is its library handling (which with some collections is abysmal).

I gather you are currently running MM/Audirvana into DAC V1, and also from other comments the V1 does have galvanic isoltion as well as being asynchronous. On that basis, and my own experience, and assuming you have your MM setup fully optimised, I would expect ND5XS to be slightly inferior in terms of sound quality - though that may depend on how your network is setup, and questions of network cabling may come into it.

Whether your MM-V1 experience can be improved I don’t know - it is conceivable that additional RF prevention could assist, whether adding (all at low cost) ferrites to the usb cable, or an Audioquest USB regen, but then thetpy might not give audible improvement.

Of course, another option for improving sound quality could be changing the DAC- quite a few people who have compared prefer Hugo’s sound to that of Naim’s DACs, typically describing as more natural or analog-like, but not everyone, and I’m not sure I’ve come across a direct comparison with V1. Hearing is obviously the thing to do - but if you do try, be sure to include an isolator otherwise the RF from MM wil have a negative effect on Hugo.

 

Innocent Bystander posted:
Frank Yang posted:

I use MM with Audirvana UPnP out, and I am happy with it. However, your V1 is not a streamer so you cannot use the Audirvana UPnP feature.

I don’t understand this optional function of Audirvana (unless you are using it purely as a means of Tidal MQA first stage unfold?) This software’s real strength is its rendering function, which you lose if you stream the music over your network by UPnP, so all you have is its library handling (which with some collections is abysmal).

Audirvana actually does the rendering, and it outputs  via UPnP instead of USB, or Toslink, i.e. without any cabling. I can play Tidal, Qobuz, and music stored on a NAS.

There are many ways one can improve the PC/MAC -> DAC USB signal integrity. In my case (PC) this greatly benefits the V1 resolution.  I am using the Uptone audio/Sonore products and my setup is here:

https://forums.naimaudio.com/t...of-streaming?page=15

There are many other ideas in the computeraudiophile forums e.g.  in the huge thread "A novel way to massively improve the SQ of computer audio streaming" !

Frank Yang posted:
Innocent Bystander posted:
Frank Yang posted:

I use MM with Audirvana UPnP out, and I am happy with it. However, your V1 is not a streamer so you cannot use the Audirvana UPnP feature.

I don’t understand this optional function of Audirvana (unless you are using it purely as a means of Tidal MQA first stage unfold?) This software’s real strength is its rendering function, which you lose if you stream the music over your network by UPnP, so all you have is its library handling (which with some collections is abysmal).

Audirvana actually does the rendering, and it outputs  via UPnP instead of USB, or Toslink, i.e. without any cabling. I can play Tidal, Qobuz, and music stored on a NAS.

I do not understand either. What do you mean by "outputs via UPnP"? Outputs to what and via which interface? UPnP is a communication protocol whereas USB and Toslink, in this context, denote physical interfaces!

Frank Yang posted:
Innocent Bystander posted:
Frank Yang posted:

I use MM with Audirvana UPnP out, and I am happy with it. However, your V1 is not a streamer so you cannot use the Audirvana UPnP feature.

I don’t understand this optional function of Audirvana (unless you are using it purely as a means of Tidal MQA first stage unfold?) This software’s real strength is its rendering function, which you lose if you stream the music over your network by UPnP, so all you have is its library handling (which with some collections is abysmal).

Audirvana actually does the rendering, and it outputs  via UPnP instead of USB, or Toslink, i.e. without any cabling. I can play Tidal, Qobuz, and music stored on a NAS.

No it doesn’t do the rendering.

But it can or does do some processing of the file before serving across the network, one aspect may be similar to how other UPnP servers can be used to transcode on the fly. 

This from the current Audirvana description:

To send audio files to your HiFi equipment, Audirvana Plus 3.1+ transmits these files over a network using the UPnP / DLNA protocol.

The UPnP / DLNA standard is the protocol used by control points, media servers and media renderers to send and receive audio files.

How does the UPnP / DLNA protocol work?

In a UPnP / DLNA network playback configuration, a control point associated with a media server sends audio data to a media renderer via a network.

The media server offers media content to be used by the media renderer.
The control point coordinates the media server and the media renderer. It controls playback on the media renderer, telling it which media to fetch and play from the media server.
The media renderer (e. g. a network player,…) is an audio device that is controlled by the control point end. The media renderer gets audio files from the media server, converts the digital signal into an analog signal using one or more DACs and reads the received data or sends them to other audio devices (amplifiers, HiFi systems, speakers…).
How does Audirvana Plus 3.1+ work in network playback?

Audirvana Plus software is a control point with an ad-hoc media server. It prepares the audio data in the format that best suits the media renderer, then sends the prepared data in the best possible way to the media renderer using a network.

Please note: Audirvana Plus 3.1 + sends audio files to media renderers that are fully compatible with the Media Renderer UPnP / DLNA standard and whose manufacturers have committed themselves to maintaining this compatibility. Network Reading Partners are identified by the “Works with Audirvana Plus” logo.

Emme posted:
Simon-in-Suffolk posted:

Well in the general sense, I suggest one keeps away from computers...

I totally agree with you

Agree or not, computers can be effective as renderers, especially when dedicated to the task provided there is effective RF blocking unless the DAC has effective blocking built in, and assuming an asynchronous DAC. 

That doesn’t mean it may not be possible to better a computer source, as I am sure the best dedicated audiophile designs can, with minimisation of potential interferences at all stages - but certainly a good computer implementation more than match the rendering stage of the likes of ND5XS, and possibly match the Melco N1A (my qualification is because my comparison was brief, so I could only say no obvious differences). 

And I suspect that with even audiophile renderers, the effectiveness of RF blocking and susceptibility of different DACs may be the main cause of differences in sound. 

Innocent Bystander posted:
Frank Yang posted:
Innocent Bystander posted:
Frank Yang posted:

I use MM with Audirvana UPnP out, and I am happy with it. However, your V1 is not a streamer so you cannot use the Audirvana UPnP feature.

I don’t understand this optional function of Audirvana (unless you are using it purely as a means of Tidal MQA first stage unfold?) This software’s real strength is its rendering function, which you lose if you stream the music over your network by UPnP, so all you have is its library handling (which with some collections is abysmal).

Audirvana actually does the rendering, and it outputs  via UPnP instead of USB, or Toslink, i.e. without any cabling. I can play Tidal, Qobuz, and music stored on a NAS.

No it doesn’t do the rendering.

But it can or does do some processing of the file before serving across the network, one aspect may be similar to how other UPnP servers can be used to transcode on the fly. 

This from the current Audirvana description:

To send audio files to your HiFi equipment, Audirvana Plus 3.1+ transmits these files over a network using the UPnP / DLNA protocol.

The UPnP / DLNA standard is the protocol used by control points, media servers and media renderers to send and receive audio files.

How does the UPnP / DLNA protocol work?

In a UPnP / DLNA network playback configuration, a control point associated with a media server sends audio data to a media renderer via a network.

The media server offers media content to be used by the media renderer.
The control point coordinates the media server and the media renderer. It controls playback on the media renderer, telling it which media to fetch and play from the media server.
The media renderer (e. g. a network player,…) is an audio device that is controlled by the control point end. The media renderer gets audio files from the media server, converts the digital signal into an analog signal using one or more DACs and reads the received data or sends them to other audio devices (amplifiers, HiFi systems, speakers…).
How does Audirvana Plus 3.1+ work in network playback?

Audirvana Plus software is a control point with an ad-hoc media server. It prepares the audio data in the format that best suits the media renderer, then sends the prepared data in the best possible way to the media renderer using a network.

Please note: Audirvana Plus 3.1 + sends audio files to media renderers that are fully compatible with the Media Renderer UPnP / DLNA standard and whose manufacturers have committed themselves to maintaining this compatibility. Network Reading Partners are identified by the “Works with Audirvana Plus” logo.

As far as I am concerned, Audirvana streams the same digital signals to a DAC regardless whether the output channel is USB, Toslink, ethernet /wifi network  (in the case of UPnP). 

How can it be different? Unless the Audirvana developer takes an exception approach with respect to sending a different audio signals over the UPnP protocol?

Frank Yang posted:

As far as I am concerned, Audirvana streams the same digital signals to a DAC regardless whether the output channel is USB, Toslink, ethernet /wifi network  (in the case of UPnP). 

How can it be different? Unless the Audirvana developer takes an exception approach with respect to sending a different audio signals over the UPnP protocol?

Yes, the same digital signal is sent via USB or Toslink (though subject to different causes of potential degradation in the different transfer processes and paths).

However, what DACs have UPnP inputs? If it is going to a streamer via UPnP, it is being rendered by the streamer’s renderer not Audirvana, unlike the direct feeds. 

Innocent Bystander posted:
Frank Yang posted:

As far as I am concerned, Audirvana streams the same digital signals to a DAC regardless whether the output channel is USB, Toslink, ethernet /wifi network  (in the case of UPnP). 

How can it be different? Unless the Audirvana developer takes an exception approach with respect to sending a different audio signals over the UPnP protocol?

Yes, the same digital signal is sent via USB or Toslink (though subject to different causes of potential degradation in the different transfer processes and paths).

However, what DACs have UPnP inputs? If it is going to a streamer via UPnP, it is being rendered by the streamer’s renderer not Audirvana, unlike the direct feeds. 

Is a streamer not a DAC? It is a DAC with a network interface.

Frank Yang posted:

A DAC is a renderer, that is why a Naim DAC / NDS / NDX  sound different to a Chord DAC

I think I will put a stop here.

A DAC is a digital to analog converter. A renderer takes a music file and converts it into a digital music stream that the DAC converts.

DACs can and do sound different because of the way they reconstruct an analog signal.

Renderers can sound different if their conversion is not bit perfect, or if they create electrical noise such as RF that enters the DAC and in some way modifies the analog output.

Been reading this post with interest. I have gone the other way- ND5 to NDX and then to mac mini with the NDac Xps2 . Works great although both the ND5 and NDX are still great streamers. 

The cable between the man and the Dac made a fair bit of difference for me. The Vertere double DFi was the one i settled with. 

I would probably look to sell the V1 and get a decent SPdif  Converter with a used NDac. Power supply can be done later

The other thing I have found is Tidal Is worse v local files on Audirvana however the Tidal Masters ones sound even better

Innocent Bystander posted:

A streamer is a renderer plus a DAC.

Hence you can have an Nd5XS or NDX etc acting as renderer feeding an external DAC. Or you can have a separate renderer like Audirvana, microRendu etc. Or one combined with a music store like in Melco, Innuos etc.

 

IB, are you saying (implying) that Audirvana streams different audio digital info to a streamer as opposed to a DAC via USB or toslink cables? Interesting!

banzai posted:
Innocent Bystander posted:

A streamer is a renderer plus a DAC.

Hence you can have an Nd5XS or NDX etc acting as renderer feeding an external DAC. Or you can have a separate renderer like Audirvana, microRendu etc. Or one combined with a music store like in Melco, Innuos etc.

 

IB, are you saying (implying) that Audirvana streams different audio digital info to a streamer as opposed to a DAC via USB or toslink cables? Interesting!

Ieven if both rourptes are using the same DAC I can’t see how it can be other than so - of course whether and how different it may sound will depend on the renderer in the streamer, and on many other factors such as the RF isolation of the streamer to DAC compared to the RF isolation of MM output to DAC, any effect of the different cables, and any effect of other network components (e.g. switches) or setup.

And as already mentioned Toslink and USB are unlikely to sound the same, the effectiveness of RF isolation possibly being different (relevant if the DAC is susceptible), and in particular if by Toslink this means the MM’s inbuilt optical output as opposed to a USB to Toslink converter from a dedicated USB bus output the Toslink would be limited by the MM’s hardware and driver software.

Innocent Bystander posted:
banzai posted:
Innocent Bystander posted:

A streamer is a renderer plus a DAC.

Hence you can have an Nd5XS or NDX etc acting as renderer feeding an external DAC. Or you can have a separate renderer like Audirvana, microRendu etc. Or one combined with a music store like in Melco, Innuos etc.

 

IB, are you saying (implying) that Audirvana streams different audio digital info to a streamer as opposed to a DAC via USB or toslink cables? Interesting!

Ieven if both rourptes are using the same DAC I can’t see how it can be other than so - of course whether and how different it may sound will depend on the renderer in the streamer, and on many other factors such as the RF isolation of the streamer to DAC compared to the RF isolation of MM output to DAC, any effect of the different cables, and any effect of other network components (e.g. switches) or setup.

And as already mentioned Toslink and USB are unlikely to sound the same, the effectiveness of RF isolation possibly being different (relevant if the DAC is susceptible), and in particular if by Toslink this means the MM’s inbuilt optical output as opposed to a USB to Toslink converter from a dedicated USB bus output the Toslink would be limited by the MM’s hardware and driver software.

Thanks IB.

I am not really concerned about how a DAC / Streamer deals with its input in this particular case, I just would like some clarifications about the Audirvana output, does it send the same digital data (musical info bit by bit) regardless whichever its output channel.

banzai posted:
Innocent Bystander posted:
banzai posted:
Innocent Bystander posted:

A streamer is a renderer plus a DAC.

Hence you can have an Nd5XS or NDX etc acting as renderer feeding an external DAC. Or you can have a separate renderer like Audirvana, microRendu etc. Or one combined with a music store like in Melco, Innuos etc.

 

IB, are you saying (implying) that Audirvana streams different audio digital info to a streamer as opposed to a DAC via USB or toslink cables? Interesting!

Ieven if both rourptes are using the same DAC I can’t see how it can be other than so - of course whether and how different it may sound will depend on the renderer in the streamer, and on many other factors such as the RF isolation of the streamer to DAC compared to the RF isolation of MM output to DAC, any effect of the different cables, and any effect of other network components (e.g. switches) or setup.

And as already mentioned Toslink and USB are unlikely to sound the same, the effectiveness of RF isolation possibly being different (relevant if the DAC is susceptible), and in particular if by Toslink this means the MM’s inbuilt optical output as opposed to a USB to Toslink converter from a dedicated USB bus output the Toslink would be limited by the MM’s hardware and driver software.

Thanks IB.

I am not really concerned about how a DAC / Streamer deals with its input in this particular case, I just would like some clarifications about the Audirvana output, does it send the same digital data (musical info bit by bit) regardless whichever its output channel.

As far as I can assess the answer is no: leaving aside anything to do with RF, Toslink implementation, different DACs etc:

1) in its internal mode Audirvana renders the file into a digital music stream that goes direct to the DAC to be converted into an analog music stream.

2) in its UPnP mode Audirvana streams  the music file to a remote renderer, which renders it into a digital music stream that is fed into the attached DAC to convert into an analog music stream.

in respect of 2), if set up to so so, I understand that Audirvana may upconvert, or apply MQA first stage unfold, or  apply any chosen digital effects through Apple’s Aidio Units capability, then sending to the DAC a digital audio stream. I understand that when transmitting as in 1), the same digital effects can be applied, though here I am  hazy as to how any changes are made to the file before sending it to the remote renderer.

Thanks again IB.

The primary 

Innocent Bystander posted:
banzai posted:
Innocent Bystander posted:
banzai posted:
Innocent Bystander posted:

A streamer is a renderer plus a DAC.

Hence you can have an Nd5XS or NDX etc acting as renderer feeding an external DAC. Or you can have a separate renderer like Audirvana, microRendu etc. Or one combined with a music store like in Melco, Innuos etc.

 

IB, are you saying (implying) that Audirvana streams different audio digital info to a streamer as opposed to a DAC via USB or toslink cables? Interesting!

Ieven if both rourptes are using the same DAC I can’t see how it can be other than so - of course whether and how different it may sound will depend on the renderer in the streamer, and on many other factors such as the RF isolation of the streamer to DAC compared to the RF isolation of MM output to DAC, any effect of the different cables, and any effect of other network components (e.g. switches) or setup.

And as already mentioned Toslink and USB are unlikely to sound the same, the effectiveness of RF isolation possibly being different (relevant if the DAC is susceptible), and in particular if by Toslink this means the MM’s inbuilt optical output as opposed to a USB to Toslink converter from a dedicated USB bus output the Toslink would be limited by the MM’s hardware and driver software.

Thanks IB.

I am not really concerned about how a DAC / Streamer deals with its input in this particular case, I just would like some clarifications about the Audirvana output, does it send the same digital data (musical info bit by bit) regardless whichever its output channel.

As far as I can assess the answer is no: leaving aside anything to do with RF, Toslink implementation, different DACs etc:

1) in its internal mode Audirvana renders the file into a digital music stream that goes direct to the DAC to be converted into an analog music stream.

2) in its UPnP mode Audirvana streams  the music file to a remote renderer, which renders it into a digital music stream that is fed into the attached DAC to convert into an analog music stream.

in respect of 2), if set up to so so, I understand that Audirvana may upconvert, or apply MQA first stage unfold, or  apply any chosen digital effects through Apple’s Aidio Units capability, then sending to the DAC a digital audio stream. I understand that when transmitting as in 1), the same digital effects can be applied, though here I am  hazy as to how any changes are made to the file before sending it to the remote renderer.

Thanks again IB and apologies to the OP that I am diverting from his original message.

The primary reasons why I am interested in the streaming / UPnP are that:

  • No funny effects from RF, jitters, cable limitations such as 24/92 
  • Clean setup because of no USB isolator involved
  • Ability to listen to Tidal Master, Qobuz because they are not supported by Naim
  • No need to use Asset to stream music from a NAS because Audirvana provides a better solution, and hopefully a better sound. In fact, I think it does give a better sound
Innocent Bystander posted:

Renderers can sound different if their conversion is not bit perfect, or if they create electrical noise such as RF that enters the DAC and in some way modifies the analog output.

All possible.. however not being bit perfect is unlikely unless resampling or gain adjustment is used. Electrical noise is also possible but not that likely .. other than with ground plane isolation and modulation issues with cheaper equipment. The most likely cause of perceived sound differences from renderers is the system coupling with the DAC where the side effects of the intermodulation frequencies produces from a transport clock’s  stability. Any serialised data stream (USB, SPDIF, Ethernet etc) requires a transport data synchronisation clock.. and this clock itself can generate by products or artefacts which can couple via cross talk into other systems. And specifically analogue signal reconstruction circuitry.

Simon-in-Suffolk posted:
Innocent Bystander posted:

Renderers can sound different if their conversion is not bit perfect, or if they create electrical noise such as RF that enters the DAC and in some way modifies the analog output.

All possible.. however not being bit perfect is unlikely unless resampling or gain adjustment is used. Electrical noise is also possible but not that likely .. other than with ground plane isolation and modulation issues with cheaper equipment. The most likely cause of perceived sound differences from renderers is the system coupling with the DAC where the side effects of the intermodulation frequencies produces from a transport clock’s  stability. Any serialised data stream (USB, SPDIF, Ethernet etc) requires a transport data synchronisation clock.. and this clock itself can generate by products or artefacts which can couple via cross talk into other systems. And specifically analogue signal reconstruction circuitry.

I agree with your general remarks but, according to https://www.naimaudio.com/site...dac_august_2009.pdf, the Naim DAC overrides the implicit clocking of incoming S/PDIF streams with one of its internal clocks before the data enter the DAC stage. This suggests that, under normal operations (stable selection of the nDAC's internal clock), the amount of jitter associated with the data entering the DAC stage of the nDAC solely depends on the quality of the nDAC's internal clocks. In this case, the differences that we hear when we listen to different transports connected to the same nDAC (assuming that they all realize bit perfect streams) would in fact be due to electrical noise alone. Am I missing something here? Of course, other DACs might work very differently and be more sensitive to the clocking/jitter of the incoming streams.

banzai posted:

Thanks again IB.

The primary 

Innocent Bystander posted:
banzai posted:
Innocent Bystander posted:
banzai posted:
Innocent Bystander posted:

A streamer is a renderer plus a DAC.

Hence you can have an Nd5XS or NDX etc acting as renderer feeding an external DAC. Or you can have a separate renderer like Audirvana, microRendu etc. Or one combined with a music store like in Melco, Innuos etc.

 

IB, are you saying (implying) that Audirvana streams different audio digital info to a streamer as opposed to a DAC via USB or toslink cables? Interesting!

Ieven if both rourptes are using the same DAC I can’t see how it can be other than so - of course whether and how different it may sound will depend on the renderer in the streamer, and on many other factors such as the RF isolation of the streamer to DAC compared to the RF isolation of MM output to DAC, any effect of the different cables, and any effect of other network components (e.g. switches) or setup.

And as already mentioned Toslink and USB are unlikely to sound the same, the effectiveness of RF isolation possibly being different (relevant if the DAC is susceptible), and in particular if by Toslink this means the MM’s inbuilt optical output as opposed to a USB to Toslink converter from a dedicated USB bus output the Toslink would be limited by the MM’s hardware and driver software.

Thanks IB.

I am not really concerned about how a DAC / Streamer deals with its input in this particular case, I just would like some clarifications about the Audirvana output, does it send the same digital data (musical info bit by bit) regardless whichever its output channel.

As far as I can assess the answer is no: leaving aside anything to do with RF, Toslink implementation, different DACs etc:

1) in its internal mode Audirvana renders the file into a digital music stream that goes direct to the DAC to be converted into an analog music stream.

2) in its UPnP mode Audirvana streams  the music file to a remote renderer, which renders it into a digital music stream that is fed into the attached DAC to convert into an analog music stream.

in respect of 2), if set up to so so, I understand that Audirvana may upconvert, or apply MQA first stage unfold, or  apply any chosen digital effects through Apple’s Aidio Units capability, then sending to the DAC a digital audio stream. I understand that when transmitting as in 1), the same digital effects can be applied, though here I am  hazy as to how any changes are made to the file before sending it to the remote renderer.

Thanks again IB and apologies to the OP that I am diverting from his original message.

The primary reasons why I am interested in the streaming / UPnP are that:

  • No funny effects from RF, jitters, cable limitations such as 24/92 
  • Clean setup because of no USB isolator involved
  • Ability to listen to Tidal Master, Qobuz because they are not supported by Naim
  • No need to use Asset to stream music from a NAS because Audirvana provides a better solution, and hopefully a better sound. In fact, I think it does give a better sound

I forgot to mention the most important reason /driver why I decided to use the Audirvana UPnP solution is that there is absolutely no "drop off", I think it is because we can afford to allocate a big inbound buffer on a Mac computer.

nbpf posted:

I agree with your general remarks but, according to https://www.naimaudio.com/site...dac_august_2009.pdf, the Naim DAC overrides the implicit clocking of incoming S/PDIF streams with one of its internal clocks before the data enter the DAC stage. This suggests that, under normal operations (stable selection of the nDAC's internal clock), the amount of jitter associated with the data&nbsp requires a transport data synchronisation clock.. and this clock itself can generate by products or artefacts which can couple via cross talk into other systems. And specifically analogue signal reconstruction circuitry.

Careful, you are mixing things up. Naim are talking about sample clock recovery from the transport stream. Sample clock recovery has not really been done this way since the early days of SPDIF , and so most devices these days work similar to the NDAC, and streamers and synchronise the serial stream to their own clocks thereby mitigating transport clock jitter directly from the sample clock.

i was referring to system cross talk. A serial stream clocked a clock that is modulated with noise will create frequency components related to the noise in connected systems through indirect and direct coupling. This is noise caused by crosstalk. This noise is added into the receiving system in various ways. 

It seems that many do confuse these two things. The former removes directly related transport clock jitter into the sample clock, but the latter introduces noise artefacts into connected systems independent in a way not directly related to the transport clock mapping to sample clock. That is why although a system can be said to reduce sample clock jitter, different transports of differeing clock stability can still sound different... what your hearing is system coupling and cross talk.

This is why in my opinion different Ethernet switches can sound different... it boils down to cross talk with connected devices from PHY layer synchronisation serial stream clock used on the Ethernet interfaces... and has little bearing on the data frame timing itself.

Simon-in-Suffolk posted:
nbpf posted:

I agree with your general remarks but, according to https://www.naimaudio.com/site...dac_august_2009.pdf, the Naim DAC overrides the implicit clocking of incoming S/PDIF streams with one of its internal clocks before the data enter the DAC stage. This suggests that, under normal operations (stable selection of the nDAC's internal clock), the amount of jitter associated with the data&nbsp requires a transport data synchronisation clock.. and this clock itself can generate by products or artefacts which can couple via cross talk into other systems. And specifically analogue signal reconstruction circuitry.

Careful, you are mixing things up. Naim are talking about sample clock recovery from the transport stream. Sample clock recovery has not really been done this way since the early days of SPDIF , and so most devices these days work similar to the NDAC, and streamers and synchronise the serial stream to their own clocks thereby mitigating transport clock jitter directly from the sample clock.

i was referring to system cross talk. A serial stream clocked a clock that is modulated with noise will create frequency components related to the noise in connected systems through indirect and direct coupling. This is noise caused by crosstalk. This noise is added into the receiving system in various ways. 

It seems that many do confuse these two things. The former removes directly related transport clock jitter into the sample clock, but the latter introduces noise artefacts into connected systems independent in a way not directly related to the transport clock mapping to sample clock. That is why although a system can be said to reduce sample clock jitter, different transports of differeing clock stability can still sound different... what your hearing is system coupling and cross talk.

This is why in my opinion different Ethernet switches can sound different... it boils down to cross talk with connected devices from PHY layer synchronisation serial stream clock used on the Ethernet interfaces... and has little bearing on the data frame timing itself.

Thanks for the clarification, I'll have to ckeck the notions of "cross talk" and "system coupling", I do not know what they mean precisely. But if the clocking artifacts of streams can have detrimental effects on sound quality in spite of being reclocked internally by the nDAC, then it is conceivable that streams generated by ultra accurate clocks could also have a negative effect on sound quality if their accuracy is not matched by the accuracy of the nDAC's internal clocks!

Add Reply

×
×
×
×