Audirvana & Sonore Ultrarendu vs direct USB to Auralic VEGA

Dear all,

For some time I have the following observation that I would like to share:

  • My current setup is a Mac Mini which stores audio files (Redbook quality at least). My DAC is an Auralic VEGA. I listen only with headphones (HE-560) to music.
  • Based on almost unanimously good reviews, I bought some time ago a Sonore Ultrarendu to avoid a direct USB connection from my Mac to the VEGA. The ultrarendu is an excellent device which works very well. That is not the question.


  • The connection Audirvana -> Ultrarendu (using MDP/DLNA) -> Auralic VEGA works really well and sounds fantastic.
  • BUT it does not sound (to my ears) significantly better than a direct USB connection Mac-VEGA, which further improves interposing a USB reclocker (Wired4Sound Recovery). It just sounds different, warmer you might say, you lose some “clarity” compared with the direct connection. By the way, the ultrarendu has a SBoost power supply.

How come?

  • As I understand it, Audirvana performs better that other media players by using a different processing technique of the audio file. Could it be that this processing advantage gets lost or diminishes when using DLNA to send the audiostream to the ultrarendu, Audirvana as such reduced to a simple remote control who actually does not process sound?
  • By extension, is it so that when making a direct USB connection, the better audio processing techniques of Audirvana get preserved, leading to better audio quality that the ultrarendu pathway?
  • Finally, this question directed to Auralic VEGA owners: does the VEGA already have pretty decent internal jitter reduction, and does it not really need the ultrarendu?

Any thoughts?

All the best,


1 Like

This post was a long while ago but i’ll reply as my observation has been the same. SMS 200 with Sbooster 12v compared with a mac mini running Audirvana both into a usb/spdif or usb/AES converter and then to my DAC. I slightly prefer the sound from my Mac mini. More pace and maybe a touch less bright. If anything I find the audirvana sound from mac mini direct more involving emotionally. Despite all the hype (which I usually buy into!) I don’t feel the renderer offers any real sonic advantage except functionality for remote streaming. I’ve gone back and forth on headphones and speakers many times. Amir at has measurements of jitter that seem to back up what we are hearing; namely no difference or small difference. On sound quality- as I say- I think I actually prefer the mac mini->usb converter->my DAC, though he would probably say i’m imagining that. I don’t think so of course:) So I will second your experience:)

1 Like

Your impressions seem to be similar to what “romaz” found on AS and WBF. The higher power CPU of the Mac Mini seems to sound superior to the lower power CPU of the SMS 200. More pace and more involving emotionally is typical for higher power CPUs so your impressions do make absolutely sense. Thank you for sharing.


This has nothing to do with the power of the CPU. If anything, the more powerful CPU is noisier and could/should sound worse. If it’s not impacting the sound quality, it’s a good testament to the quality of the USB implementation.

What you have is two slightly different flavours. Which one somebody prefers is a question of personal taste.

Listen to a Taiko Extreme. It has a very powerful CPU and sounds superior. More pace and more involving emotionally are very nice descriptions of a superior sound. This is not a matter of taste.


Whatever characteristic you might observe, it‘s not due to the more powerful CPU.

So you think that the Mac mini has a better USB implementation than the SMS200 and this is responsible for the better sound?


I don’t think that the sound is better. It‘s just more pleasing to you. If somebody else listened he might have reached a different conclusion.

There is a difference how Audirvana delivers signal through USB compared to the whole chain (Audirvana+UPnP+UR+USB), no doubt about it. Hence the small difference in sound, but I wouldn‘t qualify it as necessarily better.

If the power of the CPU was contributing factor, every streamer would have pretty much the most powerful CPU they can put in it for the budget. The reality is quite the opposite. The streamers have just enough CPU power to do the job.

Completely agree!! Exactly!! I’m actually so glad there’s some interest in this discussion (and not just cause we’ve all exhausted anything decent on netflix! — or is it? :slight_smile:

Stumbling across Rodrigo’s post from two years ago felt like validation for me after my comparison and I was actually internally nervous posting my initial reply above! That’s because- after reading all the sotm and rendu stuff on forums over the past few years and owning the sms 200 with Sbooster (12v Hans B approved voltage) for a couple years - I was really biased the other way! I mean- many say sotm and rendu are OMG/no brainer ‘better’ than [insert your fave computer source here] so while doing the comparisons I kept saying to myself ‘why am I consistently preferring the consensus ‘inferior’ way to stream!?
In fact, the best sound I ever had from my system I realized was BEFORE I started playing it through a renderer. Despite cable swaps and different converters etc., I was never able to get my system to sound as good as it used to PR (Pre Renderer) and that’s what prompted me the other night to give my best ‘shot’ at a ‘shootout’ :slight_smile:

I certainly am not claiming a ‘best’ source- just that to me the sotm with sbooster seemed a touch brighter and more ‘etched’ on my system and at a minimum, I definitely didn’t say “wow amazing!” when I switched from my mac mini to the Sotm as many reviewers in the past seem to have. There wasn’t any increase in soundstage depth or a blacker background or more bass or any of the things I typically listen for and yes- with apologies to Amir at ASR - I’m absolutely a man who hears significant differences in USB cables and interconnects etc so I’m generally not a pure ‘objectivist’ and thus open to the ‘subtleties’ and nuances of my own perception, if sonic subtleties there be.
I can see maybe that certain systems - maybe slightly ‘warm’ ones that need a bit O’ brightening- night have a reviewer going ‘yes! so much better!’ etc. I get that. Synergy. But I guess this was the first time in my several years of owning my Sotm - with pandemic kind of time on my hands- that I said to myself ‘you know what?? let’s really see what’s what’ and went and compared it head to head for hours to my mac mini vía converters.
All I say is what I thought would be no contest- (everyone from Darko to Shmarko to Hans says renderers are sooooo much better sounding because simpler or quieter or whatever) was actually VERY close sonically and I ultimately decided in favor of my Mac mini Judge-Judy-style.
Sir! you’re out of order! I’ll have the bailiff escort mr renderer to the door :wink:

Interesting about ‘expectation bias’ (Pah! too lazy to google it! Smithers - google expectation bias at once! :smile:
I actually EXPECTED the renderer would KILL the mac mini !! I put it in the system a few years ago and never (seriously) looked back. I even thought for the first few months ‘yes- maybe it IS a tad better- i like this.”
And I can’t say it won’t be just the thing sonically for some people in some set ups- Just apparently not for me in mine.
Heck- I’ve owned the friggin’ thing for like three years because it was supposed to be “better”: So color me surprised for sure!


Good discussion here, so thanks for that. One thing not mentioned is the quality of the cables being used. Very high quality ethernet and USB cables make a difference (in my system) as does clean power through a PSAudio P12 power regenerator. I will add that the SoTM Ultra Neo (with SoTM power supply) using an Audioquest Diamond ethernet and USB into PS Audio DirectStream DAC changed my system from Bla to Aha!

Post Script:
My system is Windows 10, and I ditched my mac mini’s after both died.

The SoTM sms200 Ultra Neo absolutely blows away the PSAudio DirectStream Network Bridge I & II in my system. That to me was unexpected, since I expected the Network Bridge II to meet or beat the SoTM, which I bought to use with a Denefrips Venus II.

For some time, I played around with Euphony (a player from Euphony Audio). I had an old MacBook Air lying around onto which I installed Euphony, and since I had it in my study, I used it to stream via UPnP to my Raspberry Pi 4 (running RopieeeXL) and then USB to my Devialet. The system was working fine initially, but then UPnP started to cause some problems, and I moved the MacBook into my audio rack so that it could be connected directly via USB.
To my surprise, the sound was extremely good, smooth and dynamic. Just out of curiosity, I switched my Raspberry Pi between the MacBook and the Raspberry Pi. The sound was noticeably brighter with the Raspberry Pi/RopieeeXL inserted between my MacBook and the Devialet!
Although it’s a different system, the observation fits to what has been said in this forum. I am back to Audirvana, and I‘ll re-try the experiment, though on a MacBook Pro. Curious to see if I‘ll observe the same effect.

Just doing the experiment MacBook Pro direct vs RopieeeXL in between!
I had an old MBP lying around, so I mounted Audirvana on it and took it to my audio rack. First I connected the MBP straight to my Devialet with a van den Hul Ultimate 1 metre cable. Nice, detailed sound!
Switching the Raspberry Pi 4B running RopieeeXL between the MBP and the Devialet immediately added a layer of brightness that I did not hear with a straight connection. I should say that the Raspi is powered by a linear power supply.
I am amazed to hear such noticeable difference, especially since many say that separating the computer from the input is beneficial for the sound!

Thanks for sharing.
It seems to be that most people who tried Audirvana sending files via UPnP to an endpoint (RPi based devices, Sonore, SotM etc.) and come back to a direct USB connection prefer direct USB to the endpoint concept.
I use MBP/Audirvana/Qobuz via USB straight to DAC and would never use an endpoint.


Looks like this is commonly accepted here. Though Roon still insists in separating the core and endpoint. I haven’t tried it directly with Roon, there’s no use for me anyway as Audirvana‘s SQ is so much better!

Eventually, I’m going to replace my MBP with, presumably, a new Mac Mini Silicon (M1) and run it headlessly. But that‘ll take a while.

Yes, one reason that Roon is sounding inferior to Audirvana is that there is permanent network activity ongoing with Roon. With Audirvana you can turn network completely off after loading files into RAM.
BTW, there are reports that even the in-built DAC of the new Mac Mini M1 is very good.
It seems to be a really nice device.


Hello Matt

I don’t stream that much, I have a library on an external 2 TB SSD. Will there be any network activity at all, except some occasional traffic checking router availability and such, when I play albums from the external drive? And also, something that was never clear to me, does Audirvana load an entire album into the RAM, or will there be swapping of tracks according to availability of RAM? Say, a 192/24 album will quickly chew up RAM space unless one has 16 GB RAM available.

Br, Paul

Hi Paul,

with Roon you have always network activity either with the NAS or streaming services.
I use Audirvana with Qobuz only, I have no library.
The allocated memory for files loading in my case is 1024Mb only, so only a few tracks are loading into RAM and I can turn wifi off for playback. Sometimes you have to turn wifi on for some seconds at the end of one track to get the next ones for preloading. It is very convenient. A rule to thumb is that you should have about a quarter of the entire memory allocated for tracks preloading.


Oh, I see, streaming exclusively. My MBP is hard wired, no way I‘d be able to set up a stable WiFi network in our house! I use Qobuz exclusively, so HiRes streaming works seamlessly.

Right, my MBP has 8 GB RAM, and I used to allocate 2 GB RAM for buffering. For my next Mac Mini M1 , I‘ll opt for 16 GB RAM, just to be on the safe side.

Br, Paul