Article: Sonore microRendu Review, Part 1

[MarkS]

First impression after 10 minutes. Incredibly easy to set up. Seems less grainy than my Mac mini system (tricked out internals with JS-2 PSU and Regen). Not yet sure about bass depth, but I'm using the ifi PSU and PSU had tremendous impact on bass with my Mac mini. Also, it's literally been 10 minutes.

 

I'm not sure about Aurender, but This seems like trouble for Aurelic. I tried their streamer and did not like it on SQ or GUI. Again, This took me less than 10 minutes to get working with Roon, which is class leading library management versus the Lighning App that gave me trouble.

 

Can't make a really critical SQ analysis after 10 minutes, but this thing seems extremely legit to me.

[Evo1668]

First impression after 10 minutes. Incredibly easy to set up. Seems less grainy than my Mac mini system (tricked out internals with JS-2 PSU and Regen). Not yet sure about bass depth, but I'm using the ifi PSU and PSU had tremendous impact on bass with my Mac mini. Also, it's literally been 10 minutes.

 

I'm not sure about Aurender, but This seems like trouble for Aurelic. I tried their streamer and did not like it on SQ or GUI. Again, This took me less than 10 minutes to get working with Roon, which is class leading library management versus the Lighning App that gave me trouble.

 

Can't make a really critical SQ analysis after 10 minutes, but this thing seems extremely legit to me.

 

Hi Mark

how are you using the URendu, as you say it is less grainy than your Mac mini. Does that mean you are no longer running your Mac mini? What are you running Roon on?

[baddog]

And try powering the microRendu with the JS-2 for kicks?

[MarkS]

Running Roon Core on a desktop iMac and using microRendu as a Roon endpoint.

 

Will try the JS-2 soon. 9v or 7v? Anyone know for sure. I think 9v.

 

I've always wanted to try lan based music delivery. This microRendu is part of the way.

 

Vivaldi Upsampler 2.0 coming next week. I hope it's better than the microRendu

 

Continue to think there is a little grain out using the microRendu vs. Mac mini into the DAC usb, but the stage is a bit smaller and bass is not as deep. Will try the JS-2 when I confirm voltage setting for microRendu.

 

Already certain the microRendu is worth more than the money. At worst, I'll use it in my office system with the NuPrime IDA-8.

[MarkS]

Ok, about 4 hours in and 3 immediate takeaways below. Going to let it run for a few days then will listen again.

 

1. PSU matters (JS-2 better than ifi - tighter and deeper bass, more dynamic, more detailed, wider stage, very precise imaging, but more forward presentation so far).

2. It's better than my Mac mini system

3. I think Chris is right

[Gordian]

quote : "Uptone Audio JS-2 linear power supply - use the 7VDC"

from SONORE microRendu

[barrows]

Wow, when I composed my original post I didn't realize I would be causing such an uproar - yours is about the only reply I've seen that is well mannered and not trying to personally attack me based upon asking a few questions - thanks!

 

Regarding DIY, I'm really just talking about getting a Pi, a case, an I2S DAC, putting the DAC onto the PI (no soldering needed for a HAT-compliant board), assembling the case (requires screws and tape), and then flashing a microSD card. Yes, I would not ask my wife to do such a thing, but she wouldn't know how to install/configure any commercial renderer as well.

 

But for those who do have the skills to flash a card and assemble a case, my original post really questioned two things. 1) Can you accomplish the same audio quality with a SBC and "audiophile" linux distro such as Volumio? 2) How future-proof is any commercial renderer option that is based around open source software - how can you be sure that in the future as new open source packages are made available they will run on your proprietary commercial hardware?

 

A friend of mine is a member of a San Francisco audiophile club and they have been meeting up on weekends and doing blind listening tests of equipment, DAC's most recently. There seemed to be widespread agreement that the Schiit Yggdrasil blew away DAC's many times its price. Based upon that, if I was to be in the market for a new USB DAC today, I'd probably choose the Schiit. So I'd be interested in hearing if there have been any similar bake-off of renderers, including one reviewed in this thread, along with an off-the-shelf Raspberry Pi, and later this month the Bryston BD-Pi. One without confirmation bias, which is why I'm not giving weight to many/most of the listening results people on this thread have done.

 

Regarding 2), I don't believe anybody has commented on that yet. As a technology early adopter, when new release of piCorePlayer, Volumio or Moode comes out, I want to try them. I have a LMS running on one of my Pi's right now, indexing music from my networked drives, thanks to the latest release of piCorePlayer. It seems that it would be hard for any commercial hardware company to validate and support the range of open source audio platforms that keep appearing, and are evolving so rapidly. And at least for me, it's a lot of fun playing with all of these audio-centric Linux distros.

 

And finally, the one thing I really like about the audio DIY world to date is that their communities are full of friendly and helpful people.

My one post here has elicited a set of replies that are for the most part are arrogant and condescending. Too many highly insecure people for me, so I'll be dropping out here - it's not a community I want to be a part of.

 

Your approach will not achieve anywhere near the sound quality of the µRendu into a DAC via USB. A better approach to utilizing a small MoBo unit like Raspi or BeagleBone Black (better) is to add external clocking/reclocking with isolation. The masterclock provided directly on these boards is very high in jitter, and the board itself produces so much noise that getting a good clock onboard would be nearly impossible. you can get very good performance out of such an approach, but only with a more sophisticated implementation and more parts.

the discussion of how to do this better is interesting, but very OT for this thread.

[YashN]

Your approach will not achieve anywhere near the sound quality of the µRendu into a DAC via USB. A better approach to utilizing a small MoBo unit like Raspi or BeagleBone Black (better) is to add external clocking/reclocking with isolation. The masterclock provided directly on these boards is very high in jitter, and the board itself produces so much noise that getting a good clock onboard would be nearly impossible. you can get very good performance out of such an approach, but only with a more sophisticated implementation and more parts.

the discussion of how to do this better is interesting, but very OT for this thread.

 

He would still be using a general-purpose computer rather than one designed end-to-end for the best SQ.

[master]

He would still be using a general-purpose computer rather than one designed end-to-end for the best SQ.

 

Ah, so many ways to skin this (or any) cat.

 

Comes down to how you use it.

 

The way I use it, 100% DLNA, the Pi is no different and certainly no less than the likes of Auralic (both), a PC, Mac Mini, or even a NAS. Unless folks want to sell us cleaners and conditioners for WiFi (that will clean wireless noise) or gold audiophile interconnects that connect wireless devices - I fear all of that may be coming - but let's not go there just yet and for now enjoy the music.

 

Even a general purpose - and even a multi purpose PC - while its multi tasking would sound the same when running DLNA for audio, easily achieved with the UPnP/DLNA Renderer for Foobar.

 

And my money would be on a general purpose PC (obviously an i7 with 16 GB or more RAM) running Bug Head Emperor to destroy just about any other purpose built for audio streamer/server/player.

 

Obviously the last will need head to head or toe to toe DBTs, but I deem it very unlikely that any Aurender, Auralic, Sonore, etc. would compare to the Bug and its myriad of filters that make 16/44.1 redbook sound like full blown vinyl - all coming from the venerable general purpose PC.

[skikirkwood]

Your approach will not achieve anywhere near the sound quality of the µRendu into a DAC via USB. A better approach to utilizing a small MoBo unit like Raspi or BeagleBone Black (better) is to add external clocking/reclocking with isolation. The masterclock provided directly on these boards is very high in jitter, and the board itself produces so much noise that getting a good clock onboard would be nearly impossible. you can get very good performance out of such an approach, but only with a more sophisticated implementation and more parts.

the discussion of how to do this better is interesting, but very OT for this thread.

 

I replaced my beloved HRT MusicStreamer II+ DAC with an IQAudio Pi-DAC+ I2S DAC, so I'm not actually using a USB DAC any longer. That said, while I agree the Pi has a poor master clock and is very noisy, when feeding an asynchronous USB DAC I don't see how any of that matters. If the renderer is feeding the DAC with the right bits, at the approximate right timing, and the DAC's clock is driving the timing, how would the renderer change the sound quality of the DAC's analog output? And yes, I've read John Swenson's articles.

[barrows]

I replaced my beloved HRT MusicStreamer II+ DAC with an IQAudio Pi-DAC+ I2S DAC, so I'm not actually using a USB DAC any longer. That said, while I agree the Pi has a poor master clock and is very noisy, when feeding an asynchronous USB DAC I don't see how any of that matters. If the renderer is feeding the DAC with the right bits, at the approximate right timing, and the DAC's clock is driving the timing, how would the renderer change the sound quality of the DAC's analog output? And yes, I've read John Swenson's articles.

 

Experience dictates otherwise. I have lots of experience using DACs which "re-clock" and "eliminate jitter", and in every case they still respond to a lower jitter input stream with better sound quality.

[Doak]

I replaced my beloved HRT MusicStreamer II+ DAC with an IQAudio Pi-DAC+ I2S DAC, so I'm not actually using a USB DAC any longer. That said, while I agree the Pi has a poor master clock and is very noisy, when feeding an asynchronous USB DAC I don't see how any of that matters. If the renderer is feeding the DAC with the right bits, at the approximate right timing, and the DAC's clock is driving the timing, how would the renderer change the sound quality of the DAC's analog output? And yes, I've read John Swenson's articles.

OMG... knew it would eventually get to the (very, very) old and monotonous "bits is bits" thing.

Please take this elsewhere, if you must take it anywhere. This stuff is OT and NW (not welcome) on this thread.

[Distinctive]

Not to beat a dead horse, but even IF the DIY solution is better (which it isn't), why this compulsory need of telling ordinary customers wanting a simple and holistic solution that they should buy a batch of PCB's etc and assemble this into something useful. If you DIY guys have seen the light in order to save a few bucks, just sit back and enjoy your ultimate technological superior solution.

[firedog]

There is a weakness in the story of all these USB devices (and I also own them): not a single manufacturer has published measurements showing that the output from his device is superior to the output without it.

 

I also subscribe to the "only trust my ears" argument, because in the end, even if I'm fooling myself, I enjoy listening more with the device installed.

 

But it would be nice if someone could actually show improvement that these devices bring.

[hifi_swlon]

There is a weakness in the story of all these USB devices (and I also own them): not a single manufacturer has published measurements showing that the output from his device is superior to the output without it.

 

I also subscribe to the "only trust my ears" argument, because in the end, even if I'm fooling myself, I enjoy listening more with the device installed.

 

But it would be nice if someone could actually show improvement that these devices bring.

 

It sure would.

 

It would also help with forums like WBF, which although I find utterly unpleasant and in general dislike the tone and manner of many contributors, I do think its fair to pose the question about a products effectiveness and ask for some sort of proof. After all, if manufacturers are making claims about weaknesses in USB digital data, noise on ground planes, noise in PHYs etc etc and making statements that such things exists in general in most or all devices, then surely they must have done a huge amount of research and measurement to determine this, as well as to evaluate the effectiveness of their prototypes etc, and it doesn't seem unreasonable to ask to see some if it. I can see there are various reasons to not do so - from trade secrets and competitive advantage, right through to there not being any because its based on pure hypothesis.

 

I found the Regen didn't bring any change I could detect in my system, and I was intrigued by the discussions. It's a shame the main discussion was on WBF really as I would have quite liked to follow it.

 

Let's see what the microRendu does when it finally arrives..... like you say, for the average home user, all we have are our ears and they aren't calibrated measurement instruments as such and are part of a system that can be easily influenced, albeit we can justify the results to ourselves....

[alubis]

I highly value HiFi News effort to show measurements on several transports such as Aurender N10 and Melco.

[YashN]

The way I use it, 100% DLNA, the Pi is no different and certainly no less than the likes of Auralic (both), a PC, Mac Mini, or even a NAS.

 

How is the inevitable crunch in bandwidth because of the combined Ethernet/USB controller in the Raspi no different than Auralic, a PC, Mac Mini, especially for high-rate DSD?

 

Even a general purpose - and even a multi purpose PC - while its multi tasking would sound the same when running DLNA for audio

 

It never will sound the same as a motherboard, a set of components, interfaces and power distribution designed from the ground up to do one thing and one thing really well: computer for the best SQ.

 

This device is not in the same league as a general-purpose SBC at all. One is a Lamborghini Countach, the other is a Lada. Both can make you travel from point A to B faster than by foot and you can sure work on a Lada to improve it...

[YashN]

It would also help with forums like WBF

 

I think we can forget about that for a while until things settle there (check my thread on the microrendu there to see what I mean...).

 

I believe some independent measurements with eye pattern were posted on the pinkfishmedia forum by another John (unless it was that other forum with an animal name like hifipig?, memory fails).

[hifi_swlon]

I think we can forget about that for a while until things settle there (check my thread on the microrendu there to see what I mean...).

 

I believe some independent measurements with eye pattern were posted on the pinkfishmedia forum by another John (unless it was that other forum with an animal name like hifipig?, memory fails).

 

I did, and it reminded me why I never go there Actually thats not fully fair as like I say being inquisitive and questioning is good for all of us, but it's just so, er, DULL, repetitive and ego central over there.

 

Yes, the eye patterns for the Regen were John Westlake via pfm (and not Wigwam? - not aware of an animal one!), but again some disputed whether that was indicative of an audible improvement. I don't have the kit, interest or inclination to do measurements, or the background to decipher them, but I do appreciate others doing it - whatever the conclusion - providing there's no predetermined agenda. As far as I'm aware nothing ever came from the manufacturers, unless I missed something....

[kennyb123]

There is a weakness in the story of all these USB devices (and I also own them): not a single manufacturer has published measurements showing that the output from his device is superior to the output without it.

 

I also subscribe to the "only trust my ears" argument, because in the end, even if I'm fooling myself, I enjoy listening more with the device installed.

 

But it would be nice if someone could actually show improvement that these devices bring.

 

But would sales increase if they published measurements? I think not.

 

There's obviously enough business to be had amongst those who don't really need measurements to convince them to buy. A smart company is going to apply scarce resources to where they are most needed (supporting existing customers, developing new products, bringing in new customers).

 

I've found that most (but not all) of those seeking measurements aren't easily satisfied. As soon as measurements are shown, they make it their mission to discredit those measurements. "Where's the proof that can be heard in a controlled blind test", they'll ask. Why would a company waste precious resources trying to satisfy such people when they could instead be applying those resources to taking even better care of their actual customers?

 

A company also has to be mindful of their competition. A company that publishes measurements could actually end up giving away their competitive advantage and killing their own sales.

 

Just my two cents.

[Doak]

I did, and it reminded me why I never go there Actually thats not fully fair as like I say being inquisitive and questioning is good for all of us, but it's just so, er, DULL, repetitive and ego central over there.

<SNIP>

Well put. They're off my "go to" list. Not worth the trouble.

[firedog]

FWIW, I just compared my microR NAA upsampling to DSD 64 (that's what I can do with the mR and my DAC) to a direct connection from the same server to the DAC with a CAPS PC (Windows), which allows me to do upsampling to DSD 128 to my DAC.

 

To my taste, the mR sounded better. The only words I can describe it with are that the direct to PC playback had what I'd call sort of a "hollowness" - it was more upfront, but not as solid. The mR sound was a bit more recessed, but had more "solidity". To my taste, plus for the mR. That may mean the mR sounds better, or that it is adding some kind of pleasant distortion. Doesn't matter to me.

 

Of course someone else might have had an opposite reaction, but I can only go on what I hear.

[kennyb123]

l

To my taste, the mR sounded better. The only words I can describe it with are that the direct to PC playback had what I'd call sort of a "hollowness" - it was more upfront, but not as solid. The mR sound was a bit more recessed, but had more "solidity". To my taste, plus for the mR. That may mean the mR sounds better, or that it is adding some kind of pleasant distortion. Doesn't matter to me.

 

Well said. I hear this too. My take is that there is some kind of noise present that creates that hollow sound. Strip that away and you end up with greater solidity and coherence. I heard the Aries do this relative to a Mac. And I now hear the microRendu doing this relative to the Aries.

[skikirkwood]

Not to beat a dead horse, but even IF the DIY solution is better (which it isn't), why this compulsory need of telling ordinary customers wanting a simple and holistic solution that they should buy a batch of PCB's etc and assemble this into something useful. If you DIY guys have seen the light in order to save a few bucks, just sit back and enjoy your ultimate technological superior solution.

 

Well, let's keep on beating that dead horse. First of all, I didn't choose a DIY solution to save a few bucks. I had a Squeezebox Touch and after several years was curious as to what might provide a better streaming component for my stereo. So investing $50 was a low cost and low risk experiment to see what could be done as a possible replacement. I'm not naturally a DIY kind of guy - but the risk/reward ratio here seemed worthwhile to try it.

 

A friend of mine replaced his HRT MusicStreamer II+ with a $6500 Playback Designs MPD-3, and let me have the HRT DAC on permanent loan. Comparing the sound of the Squeezebox (with its built-in DAC) to the Raspberry Pi B+ and MusicStreamer, I heard a significantly better sound on my Bryston/B&W system. As part of that experiment, I tried out various Linux distros and found them hard to work with and configure for audio - until I stumbled upon Volumio. It's a highly tuned Linux distro designed for a headless SBC acting as an audio streamer.

 

Then I started reading about the advantages of I2S DAC's, and thought for $45 it would be fun to try one and compare the sound to the HRT DAC, so I bought an IQAudio Pi-DAC+. IQAudio is as far as I can tell a one man company from Glasgow who recently moved to England. I like supporting small companies with innovative products, and everything I read about his DAC indicated it was pretty remarkable. When I received the DAC it took 5 minutes to add it on top of the Pi, power it up, and use Volumio's Web-based UI to choose it as the audio output device, as well as enable the hardware volume control it supports. The sound was stunning. No soldering involved, which is good, since I've never soldered anything in my life.

 

I was blown away how good this little DAC sounded. And I've had my friend's Playback Designs DAC in my music room. Perhaps the MPD-3 had a bit better soundstage and depth, but not a very big difference. And that's before I discovered I could change the default digital interpolation filter on the DAC to one which I thought sounded even better.

 

After that I forked up another $15 and bought a case for the Pi/DAC combo - that was the hardest part of the whole DIY effort - taping the corners of the case components before screwing them in.

 

Since then I bought another Pi, the 3, and another IQAudio DAC, along with a bunch of microSD cards, having tried out Moode and Rune Audio distributions. I'm currently running Volumio on one, and piCorePlayer on the other. I then added the open source Spotify Connect module to Volumio, so even though Volumio supports Spotify now I can use the Pi as a Spotify Connect device, which is a big improvement.

 

None of this was to "save money". It was due to my curiosity, and I think it's a lot of fun to play around with the various "audiophile" linux distros.

 

I'm also not saying the solutions I stumbled upon are "superior" to commercial solutions such as the microRendu. Different people have different requirements and weights as to what provides the superior solution to them. I am, however, questioning whether the "purpose built" approach that is being advocated by Sonore provides any better audio quality. I don't believe it does. Apparently the engineers at Bryston agree with me, with their approach of using the Raspberry Pi 2 and HiFiBerry Digi+ board as the core of their new BD-Pi.

 

I am also questioning the longer term approach of packaging a bunch of open source modules with a proprietary hardware product like this and how future-proof it will be. Actually, I'm questioning how usable is the current software packaged with the Sonore unit compared to integrated Linux distros such as Volumio and Moode. Yes, Squeezelite is one option for the microRendu, but the piCorePlayer platform incorporates the latest releases of Squeezelite with other options and software (e.g. Shairplay), providing what I am guessing (since I haven't used the Sonore software) is a more integrated, functionally complete and easy to use interface. Third party apps that control the Linux MPD player (an option for the microRendu) are pretty weak, whereas Volumio, Moode and Rune offer highly usable UI's that run in standard web browsers but with CSS tuned to smart phones and tablets above and beyond PC's.

 

And regarding Doak's comment that I'm not welcome here, this is, after all, a comment thread on a review of the microRendu. So in any critical review, and the comment thread to follow, I believe the basic design approach and "benefits" of the product should be questioned and debated. I also believe alternatives should be compared. Hence, my motivation for my original posting.

 

Now, to end here by adding more fuel to the fire, after reading up on digital audio over the last year I decided to post my first article on Medium about what I found. Here it is:

 

https://medium.com/@skikirkwood/truth-lies-and-fraud-in-the-audiophile-world-a365e56c97c4#.nhblwvrgi

 

And I followed that with a post on how I evolved from an early CD player to my current gear:

 

https://medium.com/@skikirkwood/how-to-play-2000-cds-without-a-cd-player-d6f231057971#.xikszxv5l

 

I'm sure the above two Medium postings will provide a bunch of you with a lot more ammo for snarky comments. Bring it on.

[Doak]

Well, let's keep on beating that dead horse..........

 

And regarding Doak's comment that I'm not welcome here, this is, after all, a comment thread on a review of the microRendu. So in any critical review, and the comment thread to follow, I believe the basic design approach and "benefits" of the product should be questioned and debated. I also believe alternatives should be compared. Hence, my motivation for my original posting.

<SNIP>

Despite your attempt at justification, I (still) see no valid relation between your posts and the topic of this thread.