This can be a concern on some classical servers - in order to prevent "server noise" from entering the data stream they become slow when used with large music libraries. It is why I think that a two "machine" approach could be ideal - one manages the database, fetches data and created the music stream, another machine "re-formats" and "cleans" it . Some people think that the renderer should perform this last function and be independent of the data feeder.
A new star in 2019 : Innuos Statement Server
- Thread starter CKKeung
- Start date
This can be a concern on some classical servers - in order to prevent "server noise" from entering the data stream they become slow when used with large music libraries. It is why I think that a two "machine" approach could be ideal - one manages the database, fetches data and created the music stream, another machine "re-formats" and "cleans" it . Some people think that the renderer should perform this last function and be independent of the data feeder.
This completely depends on the amount of post data processing you are trying to do. If you pass the data bit perfect (as is bit depth) I see a typical CPU using only 10% of it's capacity at most. It is nonsense to say everyone needs the kind of power you are describing.
If we post process PCM to DSD 512, then yes, that is a bit task. But that is a specialist sector, many don't need that kind of file manipulation.
The other thing IMO is more power generally means more heat and potentially more noise as well. Two servers can only add to this negative situation.
IMO most servers need a current generation CPU power, excellent power supply and separation from the data lines, care with isolation and stripped down system functions to just server music duties. These requirements can be accomplished with a 3K box at most if well designed. I am currently testing a few at the moment, and will report back on the Roon Nucleus+.
If we post process PCM to DSD 512, then yes, that is a bit task. But that is a specialist sector, many don't need that kind of file manipulation.
The other thing IMO is more power generally means more heat and potentially more noise as well. Two servers can only add to this negative situation.
IMO most servers need a current generation CPU power, excellent power supply and separation from the data lines, care with isolation and stripped down system functions to just server music duties. These requirements can be accomplished with a 3K box at most if well designed. I am currently testing a few at the moment, and will report back on the Roon Nucleus+.
Are these always enabled by default? Are there any other steps to improvde SQ right out of the box that you would suggest. (thanks BTW, didn't know this one)
Yes. Last week I sent one to my Father. He has finally decided to go to streaming and was looking at a lot of things and called me almost daily to talk through them. I ended up sending him a NUC with Roon OS preinstalled that I have used here in the audio room, LPS, NAS etc and got him all setup.
Well suffice to say he loved as I knew he would, but he couldn't stand the noise of the fan in the NUC. So he opted to go with the Nucleus+ and as a Christmas gift split the cost on one for him since I gave him all the other hardware. Tested it and it sounds the same as the NUC, minus the fan noise. Since he lives in CA I bought it here and shipped to test and save him on Sales Tax
Definitely not better than a Linux or Windows PC fully done up. Even just having a USB card like Sotm or Jcat will sound superior to the Roon Nucleus which doesn't have a PCIE expansion to allow for it.
Why I didn't build him a CAPS or suggest something different to him was because he is a VERY old school audiophile and computers, drivers and troubleshooting are not his forte. Started him out on a Raspberry PI a year ago and he gave up on it one it stopped working just one time and went back to CD's.
Even Windows Server once it doesn't shut down properly one time can boot to a sub menu and you need to hit the F1 key to bypass it. So for him it was a must to be completely headless (no monitor, mouse or keyboard) and never run into an issue. This is where the trade off is perfect for those who don't want to run Roon Core on top Linux or Windows and guarantee to run headless flawlessly.
Sonically it is more compressed. Doesn't have as much dynamic range or tonal weight. Music sounds a bit fatter by comparison in a colored way
Basically the Kassandra wants data as is, so no messing. Many DS DACs or ones that prefer DSD need a LOT of processing in Room or HQ Player first to negate the horrid filter and oversampling issues. No such problem with the Kass.
Thats why I was happy to see Alex bring that disabling feature up. Have not played with Roon as much a Jriver and dont use any of the DSP features on either, but didnt know about features turned on by default that needed to be shutoff to improve sq with the AC DAC
This is not only an affair of 10 or 20%, it is also related to the type of processing - people have analyzed the noise due to different types of processing and saw clear patterns depending on function. I run Roon with no DSP, but as soon as your library becomes large (greater than 1 Tbyte) it becomes slow on older CPUs. Yes, sometimes more powerful CPUs generate more noise - when I refer to two CPUs I am not referring to two servers.
If things were simple we would have had a perfect server since long ...
Joshua,
Should've bought your DAC a Innuos. Give me a call if you want to fix that situation and make your dad happy
Unfortunately DSP is always on by default, in every installation, for every device you configure (for the first time). And again unfortunately, they have changed the way DSP looks and works in 1.6, so the path I gave is no longer valid.
Now you need to right-click on the device name, select DSP, and click on all the filters there, and remove one by one. And, of course, make sure everything is set to "Disabled".
Should've bought your DAC a Innuos. Give me a call if you want to fix that situation and make your dad happy
Unfortunately DSP is always on by default, in every installation, for every device you configure (for the first time). And again unfortunately, they have changed the way DSP looks and works in 1.6, so the path I gave is no longer valid.
Now you need to right-click on the device name, select DSP, and click on all the filters there, and remove one by one. And, of course, make sure everything is set to "Disabled".
Joshua,
Should've bought your DAC a Innuos. Give me a call if you want to fix that situation and make your dad happy
Unfortunately DSP is always on by default, in every installation, for every device you configure (for the first time). And again unfortunately, they have changed the way DSP looks and works in 1.6, so the path I gave is no longer valid.
Now you need to right-click on the device name, select DSP, and click on all the filters there, and remove one by one. And, of course, make sure everything is set to "Disabled".
Well he is just getting his feet wet, but I can say that I have never seen him more happy which is a very rewarding feeling. He may go the full enchilada with an Innuos some day. Thanks of the advice
asiufy: have you compared the sound of the Statement using both iPeng and Roon (DSP off)? While I am waiting to hear back from my Innuos dealer, (just asked a short time ago) I think this is a question worthy of a public forum answer.
The Statement, a demo, is arriving at my dealer on Monday. It will be installed Wednesday while i am away and ready for my arrival Friday night.
While the questions about other severs are interesting I believe this isn't the thread for them This thread is about the Statement!
popspin
The Statement, a demo, is arriving at my dealer on Monday. It will be installed Wednesday while i am away and ready for my arrival Friday night.
While the questions about other severs are interesting I believe this isn't the thread for them This thread is about the Statement!
popspin
Last edited:
I have been a-b ing iPeng v. Roon since I first got the Zenith. Whenever there is a new Roon update I anticipate an improvement but so far there is never any doubt that iPeng sounds better to me. I always disable all DSP in Roon before listening and I can usually pickup the Roon shortcomings after the third music file. Whether I am streaming (Tidal or Quboz) or playing files I downloaded or ripped from cd iPeng is always on top. I haven’t revisited the new Roon unpn option simce the last 1.6 update, but I did try it on the older (1.5v.) and still much prefered Roon out of the loop. Hope this helps.
,
Can you describe in detail how Ipeng sounds different ( better) than Roon 1.6.....
and briefly, is iPeng a similar service as Roon Labs......
Thanks....
Hi Number9.
Unless I'm missing something from the explanation of the iPeng website, this is a checkpoint. That is, the software that you install on your Ipad, Iphone ... etc, to control the server software that is installed on your server hardware. This software, if the transmission protocol is UpnP can be Towonky, Minimserver, Jriver ..... etc. In case the transmission mode is RAAT, both the server software and the one installed at the control point is called ROON (there is no other, because RAAT is a proprietary transmission protocol). Therefore besides that Innuos has pre-installed (although not activated by default) ROON, what I would like to know is that server software has pre-installed and that is controlled with iPeng.
I do not know if this software is from third parties like Twonky or JRiver ..... or it owns Innuos, as is its InnuosOS, a UNIX-based operating system.
Regards
Unless I'm missing something from the explanation of the iPeng website, this is a checkpoint. That is, the software that you install on your Ipad, Iphone ... etc, to control the server software that is installed on your server hardware. This software, if the transmission protocol is UpnP can be Towonky, Minimserver, Jriver ..... etc. In case the transmission mode is RAAT, both the server software and the one installed at the control point is called ROON (there is no other, because RAAT is a proprietary transmission protocol). Therefore besides that Innuos has pre-installed (although not activated by default) ROON, what I would like to know is that server software has pre-installed and that is controlled with iPeng.
I do not know if this software is from third parties like Twonky or JRiver ..... or it owns Innuos, as is its InnuosOS, a UNIX-based operating system.
Regards
Innuos is Windows based so you can play music with any kind of software you like.
But that is far away from the thread theme.
You are entering on a never ending one.
I'm getting lost.....
But that is far away from the thread theme.
You are entering on a never ending one.
I'm getting lost.....
Asiufy:
Yes. It was my understanding that Innuos didn’t provide uPNP functionality out of the box. And that for Apple users iPeng was a good option.
I have read many divergent opinions about sound quality of iPeng vs Roon.
So I’m looking to get a better handle on these options.
Thanks.
popspin
Innuos is Windows based so you can play music with any kind of software you like.
But that is far away from the thread theme.
You are entering on a never ending one.
I'm getting lost.....
No, it's not Windows based. No wonder you're getting lost
popspin,
No, Innuos provides UPNP, Squeezelite (iPeng) and Roon out of the box.
The problem is that iPeng (Squeezelite) only outputs through the USB port, which I have no interest in. So I wasn't able to test that.
After your question last night, I began to compare Roon x UPNP on the Statement, but my first tests were inconclusive. I'll have to set up our UPNP infrastructure all over again, as I haven't used that stuff in many months. I'll do some more comparing later.
Similar threads
