Best music software?

[Pb Blimp]

I've tried a few, and the best bang for the buck is the Innuos line. Roon Core, enough CPU power to do DSP with Roon, but it won't do HQplayer.

Alex, Any news on what they are bringing to replace the SE?

 

[asiufy]

Paul,

All I can say is that it'll come out soon... I'll likely have it for THE Show in June.


cheers,
Alex

 

[prerich]

I'm still a JRiver user (Jriver 24 64bit - as I also use MathAudio Room EQ), thought about foobar2000, they have a MathAudio Room EQ plugin for free - but alas, I've already paid for the Windows vst version. I love Jriver myself. I guess I love the ability to tweak it to my taste. Now I'm running a separate NAS server located in another room that feeds my HTPC, which feeds my DAC (Burson Vivid op-amps), to my prepro (Denon x4300H). The HTPC is windows based, while the NAS is a FreeBSD based program called Nas4free.

 

[sbo6]

What is general opinion on Roon sound quality? I have used Audirvana+ for 5 years and updated each time a new version came out. I changed DACs 2 weeks ago, and my new DAC is more resolving of my server (Mac Mini with Uptone Audio MMK, LPS, SSD, 8 meg RAM feeding USB to my DAC. It sounds fantastic, but it got me thinking.

So I spent all day on trying Amarra, HQ player, PureMusic, then lastly Roon. Audirvana+ is still ahead of all of these except Roon. This surprised me. I tried Roon about 18 months ago and it was nothing special. That has changed IMO.

I don't upsample, and feed my NOS DAC resident rate. Swopping between A+ and Roon on select tracks, they sound very similar, closer than the rest of the bunch. But A+ has a digital sound to it, a very slight coarseness that clings to the treble region, most noticeable on female vocals for example. This may not apply to all setups, but that is what I am hearing on my system.

And this absolute absence of hash or treble grain brings a more realistic and relaxing musical experience. It is super subtle, and I notice it most on my LCD4s, but it is hard to ignore it. This is great, a very cheap upgrade. Also the Roon interface is first rate, better than all the others I have seen.

Probably most folk on here have already tried Roon or use it already, but if not I would say have a go at it. It has grown up some!

When you say you tried HQplayer, how was it configured? There are many filter, dither, upsample and conversion options which all yield different sound. In my experience, HQplayer solo is better than Roon interface -> HQPlayer which is better than JRiver. Recent SW updates have closed the gap between the Roon interface -> HQPlayer and HQplayer solo but HQplayer solo is still the best IME. I haven't tried Roon solo in a few months but when I did HQPlayer was definately better.

 

[barrows]

A good Ethernet Renderer will blow away ANY "Server" connected directly to the audio system (DAC). The reason for this is that servers, even the best of them, have a lot more going on and make a lot more noise than a good, low power, Ethernet Renderer. Get the big, powerful, music servers out of the audio system and move them far away, and just have a nice small, purpose built, ultra low noise, high end audio product rendering the digital audio for the DAC.

 

[sbo6]

A good Ethernet Renderer will blow away ANY "Server" connected directly to the audio system (DAC). The reason for this is that servers, even the best of them, have a lot more going on and make a lot more noise than a good, low power, Ethernet Renderer. Get the big, powerful, music servers out of the audio system and move them far away, and just have a nice small, purpose built, ultra low noise, high end audio product rendering the digital audio for the DAC.

Sorry but I can't agree, and my stance is not based on comparing servers vs. renderers (BTW - I'm using an ultrarendu) it's based on the lack of evidence and over - simplification of your assessment. Just because servers, "have a lot more going on and make a lot more noise" doesn't mean the complexity limits performance and noise cannot be regulated / filtered. I would bet good money that a SGM will completely destroy my ultrarendu and LPSU.

 

[barrows]

Nope, just not true. You can "believe" whatever you might want. But that does not make it true.

A "music server" directly connected to the audio system presents the following problems:

1. Multiple switching regulators internally which make tons of high frequency noise which is almost impossible to mitigate.
2. A relatively high power processor which makes considerable high frequency noise and contaminates any ground it is referenced to.
3. All the noise generated internally is shared to the other audio components via the AC line connection-the only possible way to avoid this problem is to power the server completely by batteries, which I did with my last ever music server, but batteries come with their own problems, although at least they do provide isolation for the AC line.
4. Music storage devices, whether SSD or HDD, these are also very noisy and require substantial power to drive them, just another source of noise.

There is nothing "magical" about SGM or Innuous, etc. which makes them immune to most, or all, of these problems. At Sonore we used to make custom music servers, long ago! Until we discovered that the best solution was removing all commercial grade computer gear from the vicinity of the audio system was the only way to eliminate its noise. Ethernet makes this possible. It is amusing to me now to see "new" music servers hit the market at ever increasing price points, using the approaches Sonore did in our servers years ago, but we have moved on.

 

[Empirical Audio]

A good Ethernet Renderer will blow away ANY "Server" connected directly to the audio system (DAC). The reason for this is that servers, even the best of them, have a lot more going on and make a lot more noise than a good, low power, Ethernet Renderer. Get the big, powerful, music servers out of the audio system and move them far away, and just have a nice small, purpose built, ultra low noise, high end audio product rendering the digital audio for the DAC.

My experience has been the same and since I designed two virtually identical interfaces, one isolated XMOS USB and the other Ethernet, I have a valid comparison:

https://www.audiocircle.com/index.php?topic=155232.0

Not that you don't need to optimize things for Ethernet, because you do. It's just different things, including router/switch mods, cabling and playback software. These are easier and cheaper IMO than doing the same for USB.

I still need USB in order to stream Amazon Prime music and other apps that use the computer audio settings.

Steve N
Empirical Audio

 

[Empirical Audio]

Paul,

All I can say is that it'll come out soon... I'll likely have it for THE Show in June.


cheers,
Alex

I found Linn Kinsky/Minimserver/BubbleUPnP to beat all others including Roon (without HQplayer), Amarra, Jriver, Audirvana, Foobar2K and others. It works with DLNA devices.

Steve N.
Empirical Audio

 

[MtnHam]

No one seems to have commented on Roon's own dedicated servers, the Nucleus and Nucleus+.
Comments?

 

[sbo6]

Nope, just not true. You can "believe" whatever you might want. But that does not make it true.

A "music server" directly connected to the audio system presents the following problems:

1. Multiple switching regulators internally which make tons of high frequency noise which is almost impossible to mitigate.
2. A relatively high power processor which makes considerable high frequency noise and contaminates any ground it is referenced to.
3. All the noise generated internally is shared to the other audio components via the AC line connection-the only possible way to avoid this problem is to power the server completely by batteries, which I did with my last ever music server, but batteries come with their own problems, although at least they do provide isolation for the AC line.
4. Music storage devices, whether SSD or HDD, these are also very noisy and require substantial power to drive them, just another source of noise.

There is nothing "magical" about SGM or Innuous, etc. which makes them immune to most, or all, of these problems. At Sonore we used to make custom music servers, long ago! Until we discovered that the best solution was removing all commercial grade computer gear from the vicinity of the audio system was the only way to eliminate its noise. Ethernet makes this possible. It is amusing to me now to see "new" music servers hit the market at ever increasing price points, using the approaches Sonore did in our servers years ago, but we have moved on.

You mean, the same way you "believe" whatever you want? I love my ultrarendu but I'm not so naive to think it's better than any computer - based solution. Can you point to comparisons against high end music servers like the SGM?

 

[barrows]

"You mean, the same way you "believe" whatever you want? I love my ultrarendu but I'm not so naive to think it's better than any computer - based solution."

My understanding is based on experience, technical understanding of the problems at hand (I am a product design consultant with Sonore), and measurements, not just a whimsical belief.

Sometimes I wonder if this forum might better be called the "What's Expensive" forum. Just because a certain approach might cost more, does not always mean it is better. The fact is, one can throw huge amounts of money at the server problem, and with huge amounts of money one can solve some (not all) of the problems inherent in this approach, or one can move to Ethernet distributed audio and spend a bunch of the money saved on better speakers.

Buying an off the shelf motherboard, and putting it in a nice case with some good power supplies and a better operating system does not solve the inherent noise issues that complete servers face.

 

[BlueFox]

Sometimes I wonder if this forum might better be called the "What's Expensive" forum.

LOL. I have said the same thing. A thread starts about a piece of gear, and it soon degenerates into the “What’s Most Expensive”.

Anyway, the Lumin S1, and possibly it’s replacment, the X1, is the correct answer in this thread.

 

[sbo6]

"You mean, the same way you "believe" whatever you want? I love my ultrarendu but I'm not so naive to think it's better than any computer - based solution."

My understanding is based on experience, technical understanding of the problems at hand (I am a product design consultant with Sonore), and measurements, not just a whimsical belief.

Sometimes I wonder if this forum might better be called the "What's Expensive" forum. Just because a certain approach might cost more, does not always mean it is better. The fact is, one can throw huge amounts of money at the server problem, and with huge amounts of money one can solve some (not all) of the problems inherent in this approach, or one can move to Ethernet distributed audio and spend a bunch of the money saved on better speakers.

Buying an off the shelf motherboard, and putting it in a nice case with some good power supplies and a better operating system does not solve the inherent noise issues that complete servers face.

barrows,

If the utrarendu is as good or better than and SGM, that's awesome. I will also add that I went down the path of continually improving my Media PC (SOTM USB card, LPSU, SSD, etc.) and when I jumped to the ultrarendu + LPS-1 powered by an HDplex LPSU powered by a PSAudio power regenerator, it was levels better than the PC. All I am saying is - as in most things in audio and especially digital - almost nothing is definite. I would still love to read about a shootout between the ultrarendu and an SGM on a highly resolving system.

 

[morricab]

Have you tried a proper server, instead of a Mac Mini?
Even an old Auralic Aries will sound better than a Mac Mini, LPS notwithstanding...

We tried an Aries with Fenton clock and it was inferior to our inexpensive blusound node. More synthetic sound and lightweight in the bass. We returned the Aries and kept the bluesound. Note, neither had an LPS or battery power.

 

[bonzo75]

Auralic Aries is quite poor, both dac and streamer.

 

[morricab]

Auralic Aries is quite poor, both dac and streamer.

Yes it was.

 

[Taiko Audio]

barrows,

If the utrarendu is as good or better than and SGM, that's awesome. I will also add that I went down the path of continually improving my Media PC (SOTM USB card, LPSU, SSD, etc.) and when I jumped to the ultrarendu + LPS-1 powered by an HDplex LPSU powered by a PSAudio power regenerator, it was levels better than the PC. All I am saying is - as in most things in audio and especially digital - almost nothing is definite. I would still love to read about a shootout between the ultrarendu and an SGM on a highly resolving system.

I can only add from client feedback that the ultrarendu does not improve upon the sgm direct usb output.

 

[barrows]

I can only add from client feedback that the ultrarendu does not improve upon the sgm direct usb output.

The problem with statements like this is that there is no context. With an Ethernet renderer such as the ultra Rendu there are an infinite number of ways of setting things up which matter. I would not expect an ultra Rendu to outperform a (what, ten times higher cost?) SGM server without good set up and an equal playing field in terms of playback software, etc... With the SGM the set up is pretty much already done.

A better comparison would be the Sonore Signature Rendu SE (with an internal power supply at least we can be assured the power supply set up is correct). Then use the same playback software (Windows/HQ Player) in the computer serving the files to the Rendu, with the Rendu in NAA mode, and the same oversampling/filter settings i HQP as used in the SGM. Now I am very confident that in this set up (and taking care of a few network details as well which I can spell out if need be) the Signature Rendu and SGM will be compared in an apples to apples comparison, and the the Rendu will handily outperform any server. Oh yeah, one more detail, the server has to be completely powered down and and unplugged from the wall and any other components when making the comparison, otherwise its noise will invalidate the results.

So, as we can see, accurate, valid, comparisons are difficult to do, and I would hesitate to take away any value from them without knowing the details.

With the SGM Server, you get something, for a lot more money, which is a bit more plug and play, and with the Rendu's you have something which requires a little more in the way of careful set up. But ultimately, the lower noise footprint of the Rendu will win out.

 

[Empirical Audio]

You mean, the same way you "believe" whatever you want? I love my ultrarendu but I'm not so naive to think it's better than any computer - based solution. Can you point to comparisons against high end music servers like the SGM?

I cannot determine what outputs this SGM server has. USB and I2S on HDMI??

For USB, the issue is the USB interface inside the DAC because this is the Master Clock. These are usually disappointing. Too much jitter.

If it's HDMI, then you are relying on the clock inside the SGM. The only clock they talk about in the description is the CPU clock. No jitter specs at all.

To achieve the best SQ, the main thing is jitter. In order to compete with the best renderers I2S and S/PDIF outputs, the Master Clock jitter must be in the <10psec range, like this renderer:

https://www.audiocircle.com/index.php?topic=156409.msg1690615#msg1690615

Plus, there is no guarantee that Roon/HQ player will beat freeware like Linn Kinsky/Minimserver/BubbleUPnP.

Steve N.
Empirical Audio