Roon SQ vs DLNA

T38.45 posted:

just pulled the trigger for a dedicated roon server based on i5 to see if the mac is the problem....keep you inform as soon as I'll have it...

Good luck! I doubt that replacing the Roon server will make a significant difference but, plese, let us know how it goes on!

Eloise posted:

Oversimplifying a little bit, there are basically three options: A) server+renderer and DAC in two boxes, B) server and renderer+DAC in two boxes and C) server, renderer and DAC in three boxes.

I understand that taking the DAC out of the renderer (CD player, in those days) was also the original idea behind the Naim DAC, before they started supporting mainly B) type solutions. 

The nDAC was a bit of out outlier for Naim as they were always very anti separating transport and DAC in the days of CD players - until that point they were maybe the only "high end" audio company who had never produced separate transport and DAC.  I would suggest that the ND line are just returning to their previous point of view - that separating the transport and DAC creates more problems than it solves; though with the ND line they have the flexibility by adding the digital inputs.

Right, but they eventually got it right and the nDAC can still be used to convert digital data streams, no matter whether these are generated by UPnP renderers, network players, Roon endpoints or other devices. 

The problem with integrated transport+DAC solutions is that the transport component can easily become obsolete with time. We have witnessed this fact with the current Naim classical streamers, think for instance of the difficulties in integrating Tidal (let alone Qobuz and Roon!) in NDS, 272, etc. This is the main reason why I prefet A) and C) to B), as mentioned in my original post.

The other question is, if one accepts transport and DAC to be integrated in the same box, why not adding storage to that box? This would solve a number of obvious problems and allow the box to be used in wireless networks without limitations in terms of reliability or sound quality.   

nbpf posted:
The problem with integrated transport+DAC solutions is that the transport component can easily become obsolete with time. We have witnessed this fact with the current Naim classical streamers, think for instance of the difficulties in integrating Tidal (let alone Qobuz and Roon!) in NDS, 272, etc. This is the main reason why I prefet A) and C) to B), as mentioned in my original post.
I don’t disagree with you over the issues of the transport + DAC solution: I was more suggesting that if you are looking for that, with a DAC other than the current nDAC then you likely need to look somewhere other than Naim.  In terms of what is coming, I would hope Naim have considered the future in creating their new streaming platform (as I understand it the 1st gen streaming was bought in, the new one is fully developed in house) and it will be more expandable / upgradable without replacing the whole unit.
 
 
The other question is, if one accepts transport and DAC to be integrated in the same box, why not adding storage to that box? This would solve a number of obvious problems and allow the box to be used in wireless networks without limitations in terms of reliability or sound quality.   

Well for all in one systems that’s what Naim has implemented with the Uniti line.  If it carried on in the ND555 remains to be seen but I suspect the streaming functions will be close to identical between Uniti Atom/Star/Nova and ND555.

I was a big fan of dac + digital front ends (aurender, mac, aries, sonore etc) to have max flexibility in terms of music service, formats etc....but I think I changed my view over the last 2-3 weeks. Imho a streamer has big advantages: it has less components- less points of failures- less tuning parameter and usually it's an optimized front/backend (ok, only the very good ones...).

Good, that's all settled then.  A blizzard of confusion to blind us all.  At least we can keep box swapping as we tinker with multiple architectures on the road to streaming nirvana!

Just in case I do need to add a fancy streaming computer (aka a Roon Core), I acquired a monster of a CISCO ethernet switch to replace the lovely simple Netgear GS108T that is now full.  As I look at the Cisco manuals with their acronyms that I don't begin to understand, I do wonder what have I done.... 

Foot tapper posted:

Just in case I do need to add a fancy streaming computer (aka a Roon Core), I acquired a monster of a CISCO ethernet switch to replace the lovely simple Netgear GS108T that is now full.  As I look at the Cisco manuals with their acronyms that I don't begin to understand, I do wonder what have I done.... 

If it has more than 8 ports it is probably fan cooled, I hope you can put it somewhere out of earshot!

T38.45 posted:

I was a big fan of dac + digital front ends (aurender, mac, aries, sonore etc) to have max flexibility in terms of music service, formats etc....but I think I changed my view over the last 2-3 weeks. Imho a streamer has big advantages: it has less components- less points of failures- less tuning parameter and usually it's an optimized front/backend (ok, only the very good ones...).

You are right to a certain degree a streamer with built in dac can be simpler if you just plug it in and go without expensive psu's. I used to have a UQ 2 and then an ND5XS and loved them both but now I have a SOtM sms-100 streaming renderer connected to a Chord 2Qute with bog standard cat 5 ethernet and USB cable  and that is it no switches no psu's no nas drive just a hired wired usb drive that can be disconnected when in not in use. The main beauty of this set up is that when used Qutest units start to appear on the used market I can upgrade to one if I wish.

I do not know if it would sound better with a cisco switch and all that but my experience with expensive cables has lead me to believe that I simply do not have the golden ears that some on possess and so I content myself with standard now.

 

I‘m going for a Auralic G2 now, seems that rendu is not the perfect solution for roon in my setup(!!!) . It is fantastic with lms squeezbox and upnp but lacks air , soundstage and colour with roon. I know this sounds banana but even my wife noticed this after 10sec listen to the same recording played over dlna and roon. I don‘t know why....maybe rendu needs more horsepower or has more jitter or too many components with roon....no clue... keep you informed folks!

I think the differences people hear with bit perfect playback are down to differences in distortion and noise introduced by hardware\components. Cables, computers, hifi gear, etc all introduce distortion and/or noise at varying levels. For example, if your dac does not offer some form of isolation on its input then electrical noise from the Mac mini will enter it and compromise sound quality. The harder your Mac mini cpu hdd etc is working with fans spinning then the more noise is likely introduced to your dac. I run a battery powered macbook (fanless entry level macbook with zero mechanical noise from hdd etc) connected to DAC which offers galvanic isolation on usb so noise is a non issue for me as it's so low and would be barely measureable inside the DAC. I'm unable to hear differences between various bit perfect players of cd rips (apple lossless) such as iTunes, Audirvana, and Roon.

Halloween Man posted:

I think the differences people hear with bit perfect playback are down to differences in distortion and noise introduced by hardware\components. Cables, computers, hifi gear, etc all introduce distortion and/or noise at varying levels. For example, if your dac does not offer some form of isolation on its input then electrical noise from the Mac mini will enter it and compromise sound quality. The harder your Mac mini cpu hdd etc is working with fans spinning then the more noise is likely introduced to your dac. I run a battery powered macbook (fanless entry level macbook with zero mechanical noise from hdd etc) connected to DAC which offers galvanic isolation on usb so noise is a non issue for me as it's so low and would be barely measureable inside the DAC. I'm unable to hear differences between various bit perfect players of cd rips (apple lossless) such as iTunes, Audirvana, and Roon.

Battery power - or more specifically complete isolation from the mains - does help avoid ground plane modulation, which I understand is one potential cause of sound deterioration in the DAC, however the computer itself inherently is still a significant source of RFI so adequate isolation is indeed still necessary. 

Incidentally, for clarity with reference to “fans spinning” in the Mac Mini, it  has just a single low powered, low speed fan (inaudible from more than a foot or two away from the unit in a quiet room), and using ssd file storage and running Audirvana doesn’t seem to excite it to run any faster than idle speed (though in my own setup I don’t have Audiv doing anything extra like upsampling).

But more significantly, for all of us what differences we hear between different players will depend on DAC, rest of system, room (that often forgotten critical part of the system) and our own ears and how we listen. 

T38.45, have you tried Roon 1.3 and DSP options when you found UPNP superior? Like you I find the UI of Roon irresistible and will eventually move there once I settle on a suitable server. 

As you pulled the trigger for an i5 based server I suspect it will not be ideal for Roon DSP (for example sonicorbiter suggests the I7 for Roon DAP and INnuos warn that their models are not suited for Roon DSP ).

I am curious whether Roon  DSP would allow Roon to equal or best UPNP...

Hi, it’s worth remembering that upsampling will cause a degree of digital (mathematical) distortion and so is to be avoided if possible for optimum sound quality. This should not be confused with zero-value-sample oversampling which is mathematically benign, and is what Naim  use for example  in their reconstruction filters.

Innocent Bystander posted:
Halloween Man posted:

I think the differences people hear with bit perfect playback are down to differences in distortion and noise introduced by hardware\components. Cables, computers, hifi gear, etc all introduce distortion and/or noise at varying levels. For example, if your dac does not offer some form of isolation on its input then electrical noise from the Mac mini will enter it and compromise sound quality. The harder your Mac mini cpu hdd etc is working with fans spinning then the more noise is likely introduced to your dac. I run a battery powered macbook (fanless entry level macbook with zero mechanical noise from hdd etc) connected to DAC which offers galvanic isolation on usb so noise is a non issue for me as it's so low and would be barely measureable inside the DAC. I'm unable to hear differences between various bit perfect players of cd rips (apple lossless) such as iTunes, Audirvana, and Roon.

Battery power - or more specifically complete isolation from the mains - does help avoid ground plane modulation, which I understand is one potential cause of sound deterioration in the DAC, however the computer itself inherently is still a significant source of RFI so adequate isolation is indeed still necessary. 

Incidentally, for clarity with reference to “fans spinning” in the Mac Mini, it  has just a single low powered, low speed fan (inaudible from more than a foot or two away from the unit in a quiet room), and using ssd file storage and running Audirvana doesn’t seem to excite it to run any faster than idle speed (though in my own setup I don’t have Audiv doing anything extra like upsampling).

But more significantly, for all of us what differences we hear between different players will depend on DAC, rest of system, room (that often forgotten critical part of the system) and our own ears and how we listen. 

I would imagine modern silent and fanless energy efficient laptops such as entry level macbook or mobile devices such as ipad are much less a source of RFI than perhaps big powerful noisy desktop PCs used to be. My guess is DAC design is of far more importance these days for sound quality and tackling RFI.

Halloween Man posted:

I would imagine modern silent and fanless energy efficient laptops such as entry level macbook or mobile devices such as ipad are much less a source of RFI than perhaps big powerful noisy desktop PCs used to be. My guess is DAC design is of far more importance these days for sound quality and tackling RFI.

RFI interference from a computer is that it need have nothing to do with fans, but is inherent to the very design/architecture of computers, primary sources being various components of the computer itself: All computers have an internal oscillator running at radio frequency as the clock for the CPU. The characteristic waveform of the clock signals is square-wave, which tend to result in a lot of harmonics not just the clock frequency. Then the video card may have its own generator, as may other components like the monitor and keyboard (latter n/a in headless Mac Mini). These clocks are usually sub-divided into a number of other frequencies within the computer for various other components that run at different bus speeds, like RAM, adding additional sources. These all of course operate regardless of how powerful a computer may be, or whether it is running on battery power. In addition, even on battery power there it is possible that there may be switching regulators changing voltages within the motherboard, making more sources of RFI.

Careful design of the computer’s circuitry and layout can minimise the RF content of the outputs, whether USB or SPDIF - but that is unlikely to be a consideration in a computer whose primary focus is on being compact, as in a notebook, whereas it could be in one designed for high quality audio, as Aries Mini, SOtM SMS‑200, Melco, Innuos Zenith, Uniti Core, various streamers etc. It doesn’t make computers useless, but does demand attention to removal/isolation unless the DAC is good at doing that.

Hello IB, yes I didn't mean to imply fans are a source of RFI, I have no idea. I was implying modern laptops such as entry level macbook use very little power and probably generate not much more RFI than an iPhone or iPad, and are completely isolated from mains when running on batteries. Nothing like the noisy mains computers of old. Doesn't the designer your much loved Chord Dave use a humble modern laptop as his source? Surely, if he considered RFI was an issue then I'm pretty sure he wouldn't use a laptop given he has dedicated much of his work to eliminating RFI from the DAC. This suggests DAC design is of more importance than worrying about RFI from a laptop.

Halloween Man posted:

Hello IB, yes I didn't mean to imply fans are a source of RFI, I have no idea. I was implying modern laptops such as entry level macbook use very little power and probably generate not much more RFI than an iPhone or iPad, and are completely isolated from mains when running on batteries. Nothing like the noisy mains computers of old. Doesn't the designer your much loved Chord Dave use a humble modern laptop as his source? Surely, if he considered RFI was an issue then I'm pretty sure he wouldn't use a laptop given he has dedicated much of his work to eliminating RFI from the DAC. This suggests DAC design is of more importance than worrying about RFI from a laptop.

I understand that the said designer does indeed, at least on his frequent travels - I’m not sure that he has clarified re home listening. And what he has done, certainly with Dave and I believe also with Hugo TT that you use, but definitely not with plain Hugo - is build in very effective RF blocking in the DAC - though even he with his battery powered laptop has stated that he finds that a USB Regen adds that tiny bit more isolation as to be worthwhile.

And that is a critical point: not all DACs have effective RF isolation, and for any that don’t, if the source is riddled with RF (as even laptops can be) then the sound can be degraded, presumably all the more so as RF contamination increases. So, for DACs without exemplary RF rejection, using general purpose computers without including effective isolation between computer and DAC risks sound degradation, and even with sources designed for audio, differences in effectiveness of RF rejection might explain differences in sound quality, varying with by DAC.

 

As an aside, how do you use your macbook in practice? I assume it is dedicated to music play, at least when in use for that purpose, screen closed, and no other applications running? Regarding recharging, presumably in dedicated and screen off mode consumption is low and the batteries last a while - how long? Do you plug it in after each listening session, or may e weekly, or are you using it at other times of the day connected to mains?

Simon-in-Suffolk posted:

Hi, it’s worth remembering that upsampling will cause a degree of digital (mathematical) distortion and so is to be avoided if possible for optimum sound quality. This should not be confused with zero-value-sample oversampling which is mathematically benign, and is what Naim  use for example  in their reconstruction filters.

Thanks Simon, I was somehow wondering if that were the case indeed so good to have this point clarified.

Also Roon DSP puts a high demand on resources which in turn takes its toll on noise... Innuos for example  are apparently reluctant to increase the resources to fully support Roon 1.3 DSP and are honestly warning about this in their homepage.

Innocent Bystander posted:
Halloween Man posted:

Hello IB, yes I didn't mean to imply fans are a source of RFI, I have no idea. I was implying modern laptops such as entry level macbook use very little power and probably generate not much more RFI than an iPhone or iPad, and are completely isolated from mains when running on batteries. Nothing like the noisy mains computers of old. Doesn't the designer your much loved Chord Dave use a humble modern laptop as his source? Surely, if he considered RFI was an issue then I'm pretty sure he wouldn't use a laptop given he has dedicated much of his work to eliminating RFI from the DAC. This suggests DAC design is of more importance than worrying about RFI from a laptop.

I understand that the said designer does indeed, at least on his frequent travels - I’m not sure that he has clarified re home listening. And what he has done, certainly with Dave and I believe also with Hugo TT that you use, but definitely not with plain Hugo - is build in very effective RF blocking in the DAC - though even he with his battery powered laptop has stated that he finds that a USB Regen adds that tiny bit more isolation as to be worthwhile.

And that is a critical point: not all DACs have effective RF isolation, and for any that don’t, if the source is riddled with RF (as even laptops can be) then the sound can be degraded, presumably all the more so as RF contamination increases. So, for DACs without exemplary RF rejection, using general purpose computers without including effective isolation between computer and DAC risks sound degradation, and even with sources designed for audio, differences in effectiveness of RF rejection might explain differences in sound quality, varying with by DAC.

 

As an aside, how do you use your macbook in practice? I assume it is dedicated to music play, at least when in use for that purpose, screen closed, and no other applications running? Regarding recharging, presumably in dedicated and screen off mode consumption is low and the batteries last a while - how long? Do you plug it in after each listening session, or may e weekly, or are you using it at other times of the day connected to mains?

Hi IB, I think it came as a surprise to Rob Watts that with plain Hugo 2 (no galvanic isolation on USB) he found that he could hear little, if any, difference between optical (fully isolated) and USB from his mains powered laptop. His assertion was that the DAC design of Hugo 2 has sufficient RF rejection for modern mobile devices. I think we are saying the same thing really, that the DAC design is all important and if done properly then modern laptops or mobile devices, especially if battery powered, shouldn't pose a problem or degrade sound quality.

I use my laptop primarily as a dedicated music source, the only thing connected to it is the Hugo TT DAC via USB. When playing (using iTunes) I run on batteries and leave the screen open and on so I can see what is playing (lowest brightness), no other apps running. I control it using my iPhone using the Apple Remote App. I almost exclusively play bit perfect CD rips 16/44 apple lossless, occasionally digital radio. Never felt the need for hi-res or streaming, I like to keep the music I buy ) Batteries last around 8 hours, if running low I'll charge overnight at the end of a listening session. I tend to charge it perhaps every week or two depending upon use. I do take the laptop away with me for occasional light business use when I'm away from home.

ok, so at the end we must agree that the source first rule still applies- especially for digital systems. :-)

I think that DARKO wrote that a high-end renderer (Innuous SE beast) and a midrange DAC (Hugo) gives you more  than a mid-range renderer and a high-end DAC.

Halloween Man posted:

Hi IB, I think it came as a surprise to Rob Watts that with plain Hugo 2 (no galvanic isolation on USB) he found that he could hear little, if any, difference between optical (fully isolated) and USB from his mains powered laptop. His assertion was that the DAC design of Hugo 2 has sufficient RF rejection for modern mobile devices. I think we are saying the same thing really, that the DAC design is all important and if done properly then modern laptops or mobile devices, especially if battery powered, shouldn't pose a problem or degrade sound quality.

I use my laptop primarily as a dedicated music source, the only thing connected to it is the Hugo TT DAC via USB. When playing (using iTunes) I run on batteries and leave the screen open and on so I can see what is playing (lowest brightness), no other apps running. I control it using my iPhone using the Apple Remote App. I almost exclusively play bit perfect CD rips 16/44 apple lossless, occasionally digital radio. Never felt the need for hi-res or streaming, I like to keep the music I buy ) Batteries last around 8 hours, if running low I'll charge overnight at the end of a listening session. I tend to charge it perhaps every week or two depending upon use. I do take the laptop away with me for occasional light business use when I'm away from home.

I think there is a little confusion, true RF radiation will permeate audio electronics, but often I see the term RF interference mentioned when I don’t think that is what is being referred to. Galvanic isolation will have zero affect on RF frequencies... galvanic isolation is there to stop dc loops and VLF (very low frequency) loops. What can also happen is ground plane modulation due to poorly designed and implemented decoupling capacitors and PCB layout .. and yes here the Chord devices along with Naim devices have much thought here.

Now the other area where interference will be induced in the DAC is from digital transport jitter, which will produce intermodulation components in the receiving electronics... this in itself can also produce ground plane modulation. Now Rob Watts told me on the higher end Chord DACs the ground plane modulation is not measurable with the equipment he is using, however it’s not to say it’s not there... so that is why with with every device I have come across they will sound at their best with accurate and ultra stable transport clocks.

 

Simon-in-Suffolk posted:
Halloween Man posted:

Hi IB, I think it came as a surprise to Rob Watts that ...

...

Now the other area where interference will be induced in the DAC is from digital transport jitter, which will produce intermodulation components in the receiving electronics... this in itself can also produce ground plane modulation. Now Rob Watts told me on the higher end Chord DACs the ground plane modulation is not measurable with the equipment he is using, however it’s not to say it’s not there... so that is why with with every device I have come across they will sound at their best with accurate and ultra stable transport clocks.

I thought that at least the Naim DAC under normal operations would override the transport clocks with its own internal clocks. Whether this is for the better or for worse I do not know, it proably depends on the accuracy of the transport clocks and of the nDAC internal clocks. But, according to https://www.naimaudio.com/site..._dac_august_2009.pdf, overriding takes place. This should make the accuracy of the transport clock invisible to the digital to analogue section of the nDAC. I understand that other DACs might work quite differently, of course.

nbpf posted:
Simon-in-Suffolk posted:
Halloween Man posted:

Hi IB, I think it came as a surprise to Rob Watts that ...

...

Now the other area where interference will be induced in the DAC is from digital transport jitter, which will produce intermodulation components in the receiving electronics... this in itself can also produce ground plane modulation. Now Rob Watts told me on the higher end Chord DACs the ground plane modulation is not measurable with the equipment he is using, however it’s not to say it’s not there... so that is why with with every device I have come across they will sound at their best with accurate and ultra stable transport clocks.

I thought that at least the Naim DAC under normal operations would override the transport clocks with its own internal clocks. Whether this is for the better or for worse I do not know, it proably depends on the accuracy of the transport clocks and of the nDAC internal clocks. But, according to https://www.naimaudio.com/site..._dac_august_2009.pdf, overriding takes place. This should make the accuracy of the transport clock invisible to the digital to analogue section of the nDAC. I understand that other DACs might work quite differently, of course.

And indeed the Chord DACs, certainly Hugo, TT and Dave, do the same, so jitter is not an issue - however I think Simon is suggesting that jitter at an earlier stage produces intermodulation that then gets carried through, not being removed by reclocking or RF blocking, to have an adverse effect? 

Regarding residual ground plane modulation, even with Dave’s very extreme attention to isolation, and using a battery powered source isolated from any connection to ground, Rob Watts does note detectable, though tiny, improvement with a USB Regen on the input cable 

 

Hi Simon, there is no doubt of RF residual coupling via the USB of mobile devices such as laptops. Rob Watts did state though elsewhere that the galvanic isolation on his dacs together with battery operation effectively means almost perfect RF noise isolation from the source. I believe TT DAC PCB ground plane is protected from external RF by the aluminium case.

If Rob watts and chord electronics are unable to measure minute ground plane modulation then I suspect no-one can.

Maybe we should just all go optical and have done with this nonsense!

Installed Auralic G2 few minutes ago...enabled roon on mac....sounds very very good, soundstage, musical flow, colour...this Auralic is a beast...everything is there...and that out of the box stoned cold....wow

next stop: i will replace mac roon with i5 roon if i get it to work ;-)

ok, this is it! MUCH darker backround with i5 running roon...great, simply great. Never thought that could be such a difference btw mac and i5 dedicated server. Migration of roon is a bit tricky but sound is better now. ...I'm there:-))

So my tip: if you start with roon please use a dedicated server and if it doesn't sound as good as upnp/dlna go for a better streamer! this sounds crazy because roon handles metadata only but for me this set-up worked ...

 

T38.45 posted:

ok, this is it! MUCH darker backround with i5 running roon...great, simply great. Never thought that could be such a difference btw mac and i5 dedicated server. Migration of roon is a bit tricky but sound is better now. ...I'm there:-))

So my tip: if you start with roon please use a dedicated server and if it doesn't sound as good as upnp/dlna go for a better streamer! this sounds crazy because roon handles metadata only but for me this set-up worked ...

 

Hi @T38.45, Could you describe/list your playback path now, please - from File Storage, through to Ampflication re, Roon Core and Endpoint.

Thanks, Simon.

Pev posted:

@T38.45 which QNAP do you have? - I am running Roon on my QNAP TS251 and it's totally fine if you don't want to use a lot of DSP. Just hang a 120gb SSD off the back and you're good to go.

Any DSP will change thecsound (isn’t that the point?), so comparison of Roon with direct from NAS is only valid if all DSP is avoided. Could this possibly be the issue? Is Roon doing some DSP, maybe even not known to the OP?

T38.45 posted:

ok, this is it! MUCH darker backround with i5 running roon...great, simply great. Never thought that could be such a difference btw mac and i5 dedicated server. Migration of roon is a bit tricky but sound is better now. ...I'm there:-))

So my tip: if you start with roon please use a dedicated server and if it doesn't sound as good as upnp/dlna go for a better streamer! this sounds crazy because roon handles metadata only but for me this set-up worked ...

 

What has broght the most significant improvement? Replacing the microRendu with the Auralic G2 or replacing the mac mini with the i5?

Auralic G2 sounds  more fluid than sonore+sbooster, no doubt. but i5 was much better than mac....so I would really start with an i5+sonore+good power supply if I had to do it again! Guess if you can handle  roon on NAS you may have other results ... but give it a try! Maybe you don’t need i5.

Likes (0)
×
×
×
×