Roon Vs jriver

[barrows]

On 3/27/2018 at 8:41 AM, davide256 said:

There's a lot to be said for RAAT vs UPNP/DLNA. It does what its supposed to do, which is eliminate the nuisance network tweaking that UPNP/DLNA requires. Roon without RAAT supporting endpoint doesn't sound any better than UPNP/DLNA

I do not agree with this at all, at least with Sonore Rendu streaming devices.  I find DLNA streaming to require zero tweaking, and find playback via minimserver/bubble or Audirvana + to work perfectly, with excellent sound quality (better than last time I tried ROON, but it has been awhile and I hear ROON has improved since then).

I see no advantage with RAAT personally, at least with my hardware.

One thing I do see as an advantage with ROON is that it can up convert DSD to higher rates, which currently Audirvana does not do: I would like to be able to convert DSD 64 and 128 to DSD 256, for example.  I wish both ROON and Audirvana would add a DSD compatible volume control, right now that is only available in HQPlayer... 

[barrows]

26 minutes ago, firedog said:

https://community.roonlabs.com/t/64-bit-floating-point-volume-control-with-dsd/28810/13

they aren't big fans of the HQP solution

It seems that ROON RAAT runs with five seconds of delay, wow.  It has been awhile since I used ROON, but that seems unnecessary to me.  Certainly DLNA does not do this, I experience no "problems", indeed if one's network is so bad as to require that much buffering perhaps the network issues should be solved (although i do understand that ROON is not really aimed at audiophiles, and this is just another example of that).  Does it really take five seconds for playback to start in ROON after one pushes "play"?

Anyway, a DSD capable volume control is a good feature, and very desirable for users who want to oversample everything to DSD and run a simple DSD DAC like DSC-1 & 2, Playback Designs Merlot, Holo DAC Spring and Cyan, T+A, etc...  Not everyone wants to deal with HQPlayer.

Hopefully Damien will implement a DSD volume control in Audirvana Plus soon, as it has no such long buffer delay as RAAT.  (I would not want it running in the endpoint as this would require considerably more processing in the endpoint, where it should not be).

[barrows]

1 hour ago, ssh said:

5900 albums on my hard drive, hooked up to ST-5, to an UltraRendu to a Yggy. Hit "shuffle play" and no delay for me. No five second delay here.

 

Weird... I wonder why they use that as a reason to not do the volume control then.  Shenanigans I guess...

[barrows]

Well I do not think there is any such buffer in my set up (DLNA with A+) and never have any network hiccups.  Fixing a network problem this way seems back asswords to me.  Of course I would never use WIFI for music distribution as it results in more errors and re-sends (and resultant drop in SQ), if the network is even capable enough to do WIFI with DSD 256, etc.  This is why we do not offer WIFI at all with Sonore products, we are after the best sound quality and stable performance.

 

Sounds like less intelligent design for dummies to use to me, especially if it makes implementing a good DSD volume control impossible.

With all the nonsensical "features" offered in ROON which no one needs, it is hilarious to hear them suggest that a DSD capable volume control is not something they will not consider.

[barrows]

I do not hate ROON.  I do think their resistance to adding a DSD capable volume control  is in error, especially considering all the other features they do offer, which are superfluous for many users.  Why not add one more feature, which would add considerable functionality for many users?  HQPlayer offers a DSD capable volume control which operates on the "core" device when using network streaming, why are the ROON developers so against doing the same?

By all accounts ROON has a very good PCM-DSD oversampling engine, and it offers the capability to oversample lower rate DSD to higher rate DSD.  This makes it a very useful software program for some who might want to run DACs which sound better with high rate DSD input.  There are many users out there who do not want (or need) to use HQPlayer for this, many who might not want to have to set up a very powerful computer to run HQPlayer, and many of the same users may want to take advantage of the increased accuracy and transparency of not having an expensive and distortion adding analog volume control and additional buffer circuitry in their signal path-the obvious solution would be ROON with a good DSD capable volume control. 

Honestly, are you just a ROON fanboy who loves whatever they do, or are there additional features you might like to have?

This is a legitimate feature request, that is all.

[barrows]

@diecaster, you attacked me, suggesting I could not be objective, and that I hate ROON, to me those are the actions of a fanboy singing the praises of ROON regardless of a criticism which has legitimate value.

I like how ROON handles PCM-DSD oversampling and that it allows for lower rate DSD conversion to higher rates, without having to use a full on gaming super computer, but I wish they would add a high quality DSD volume control: this has nothing to do with my ability to "be objective", it is a desire for a feature which would add value to ROON for many users.  HQPlayer can do it (in the core for streaming), so why not in ROON, these are some clever folks, and I am a bit dismayed that they dismiss the value of a DSD capable VC, after all, they have volume control for PCM...

[barrows]

8 hours ago, firedog said:

Barrows-

Roon is also about zone playback and simultaneous zone playback. HQP isn't. Different goals. Not in direct response to anything you said, but Roon also does DSP to DSD (including for volume levelling) without converting it to PCM, which is a really cool feature that you don't find too often. 

 

I think Roon understand many users have below optimum network setups. I think their POV on who their customer base is and what is most needed for them is simply very different from what you want to do.

 

At the high end of the audio hobby we all have specific likes and dislikes. I know of all sorts of HW and software I'd never buy b/c it doesn't have a specific attribute or feature I find necessary. Obviously whoever produced it decided that feature wasn't absolutely necessary. Doesn't mean it isn't a very good product, it just means it isn't the perfect match for me. 

 

But I use the HW volume control on my system only, so maybe I shouldn't comment on this one. I hate on screen volume controls. Can never get the precision I like with adjustments. 

My point is, ROON offers volume control for PCM, so clearly they see that there is value in having a volume control, so why not for DSD.  I am sure if they wanted to they could figure a way to work it out.

I agree with you concerning how most on screen VCs are configured, but it does not have to be that way: again, sharp folks like the guys at ROON could implement an onscreen volume control with more precision which is easy to use: all you would have to do is have a tap sensitive control which changed only 0.5 dB at a time via a single tap, with a slider for larger adjustments.

[barrows]

12 minutes ago, Abarth said:

Volume control can't be done directly on DSD raw data. It has to be converted to PCM first, and then reconverted to DSD. Anyway, no matter how the wordlength in the digital domain is, volume control in software is a lossy thing. At the end your dac, will be feeded with data words that use only the first few bits at lower volumes in PCM. And for DSD in principle it's the same. For serious listening, software volume control is a no no (also if done in a DAC or network player).

This is not entirely true, HQPlayer does it, and the ESS chips control DSD volume without conversion to PCM rates, that is the entire point here.  While some of this may be semantics, here is one way: say you have DSD 64 at 2.xxx MHz and one bit, then you modulate to 2.xxx MHz at 64 bits, then you control the volume.  This method is what is called "DSD wide" in the pro world and involves no decimation (no change in sample rate, which would create some artifacts, and is how you control DSD volume without degradation).  As long as your modulators are good, and the processing power is enough, this is a transparent process.

@Miska can explain (or not as he chooses) how HQPlayer does it, but it also does not convert to PCM to control volume.

Digital volume control in software is not lossy, if done right, and at 64 bits you have room for 40 bits (120 dB!) of attenuation with no loss, even at 32 bits there is room for 8 bits of attenuation with no loss of 24 bit files (24 dB) of attenuation.  And the fact is that your system, and mine, only really resolve 16 bits at most in room anyway.

The idea that digital volume control is a problem is a myth that needs to be debunked, analog volume controls are indeed lossy, but digital volume controls are transparent as long as one is not using too much attenuation.

[barrows]

1 hour ago, Abarth said:

digital volume control in digital domain done right is not lossy. But at the end of the process, you'll have to feed your eg 24 bit DAC. There is no free lunch. digital volume controls are, for my ears, way more intrusive than anything Roon vs JRiver sound differences.

But its me anyway, which doesn't give much on DSD vs good old PCM DCS with classy mastered source material.

Consider that virtually all recordings done in the last 20 years use digital attenuation in the mixing process...  Apparently your ears are broken, or you just believe mythological nonsense enough to be hearing things which are not there.  This is one of those technical subjects which is not debatable, the actual engineering facts do not lie, and there is nothing "magical" going on.  The idea that digital attenuation is a "problem" is a myth, as long as one is not using too much of it.

[barrows]

 

1.  I understand that there is no difference at all, you are making something up which does not exist.  It does not matter to the 

     DAC what the incoming level is, if it did, then all signals below full scale would have problems.  Unless you are referring to

     the fact that the self noise is not reduced as the volume is lowered; again this is not a real world problem unless one has to

     apply unreasonable levels of attenuation.     

 

2.  Again, it is you who does not understand, you at most 20 bits actual resolution, and volume reduction is reducing the 

     lowest bits which are not information you can hear.  You are not losing actual resolution at all.

 

3.  Agreed, many systems do have gain structure issues: the proper way to handle this to solve it, not by adding additional

     noise and distortion 

     through high levels of analog attenuation.  You seem to believe there are no drawbacks to analog attenuation, this is not 

     the case at all.  Resistors have distortion, which can be measured, and they add thermal noise, the more analog attenuation

     one uses, the more problems you have.  Fix the gain mismatch the correct way, by properly matching your components 

     gain: it is hardly a high end approach to have a massive gain mismatch and then fix it by massive attenuation.

 

4.  I can trot out just as many respectable high end designers who would disagree, and even the late Mr. Hansen, who I was

     lucky to have known, changed his opinion on this making the ESS 9038 chip's digital VC available in the QX-5.