microRendu - Sound Quality observations and comparisons

[Antoine]

IFi is responding to these ASR measurements of the iFi(/microrendu). They might prove it's (a) ground loop(s) that messed up these measurements.

"Much Ado About Nothing". It will be a four parts article, first two have been published as of now:
http://www.computeraudiophile.com/f30-ab...ing-28982/

[Confused]

IFi is responding to these ASR measurements of the iFi(/microrendu). They might prove it's (a) ground loop(s) that messed up these measurements.

"Much Ado About Nothing". It will be a four parts article, first two have been published as of now:
http://www.computeraudiophile.com/f30-ab...ing-28982/

Gosh!  I have just caught up with some of the stuff being posted on the Roon forum and well as ASR.  To be honest, I don't know enough to be clear myself who is right or wrong, but there is an absolutely epic amount of controversy regarding some test measurements, almost exclusively written by people who have not actually listened to a microRendu.  The one person who had listened to one and said it sounded good, ended up being trolled and abused in a debate about expectation bias.  Quite unbelievable!  At the time of writing, the Roon moderators have closed the thread whilst they 'decide what to do with it'.

[Confused]

I have had my microRendu for just over a week now.  Last weekend did include a few hours of listening to the thing, but this was far from critical listening, as it involved company and a modest (!?) amount of Macon Rouge.  All very enjoyable, but not ideal for critical listening.  One observation though, the microRendu was absolutely rock solid, not one stutter, dropout or glitch.  No small thing when you think about it, because it is very annoying when you are with company, typically people who think you are a complete lunatic for spending so much money on hifi, if your AIR powered system starts stuttering or whatever.  It's also worth noting that this is an absolute direct comparison with AIR, because the microRendu is running off the same PC, with the same network, identical.  Well, identical except the mR is stable, and AIR is not.  OK, nothing new here...

So what does it sound like?  Even during last weekend's Macon Rouge soaked session, I was coming to the conclusion it was pretty good, I was certainly enjoying listening to it.  Today, I had some free time, so could do some critical listening, as well as some back to back testing with the only real benchmark I have at home, Windows AIR 2.1.3.  The short version of this is that AIR 2.1.3 and the microRendu sound very similar.  Indeed, with many tracks played quick-fire back to back I was genuinely finding difficult to separate the two, so any sound quality observations from here on come with the caveat that we are talking very fine margins here.  However, I did fairly quickly start to think that there was something I preferred about the microRendu, just very slightly more realistic, a fraction more organic than AIR's 'digital edge'.  However, this did worry me.  The margins were so fine that I did have to question how much of this was real, and how much was expectation bias.  Maybe I was expecting the mR to sound more realistic than AIR?  The margins were so small, that yes, it could be imagined.  So I persevered, checked rechecked, many back to back tests, and convinced myself that the mR did have the edge, just!  Later a golden opportunity presented itself, a friend willing to try a blind test.  A small selection of tracks were selected, played randomly between the mR and AIR.  My PC is a reasonable distance behind the 'listening sofa', and out of site.  Sat on the sofa there is no way of knowing which was playing.  The blind tester picked the mR as being preferred pretty consistently, but did make the statement that there was nothing in it.  There was however one track where AIR got the nod.  This was the 2012 remake of Freddy Mercury's Barcelona (remade with a proper orchestra, rather than the electronic trickery of the original).  This was an interesting result for me.  The Barcelona track does have some previous form for me, being one track that put me off the Melco N1Z a bit.  This is a track with utterly thunderous timpani generated bass, and with the Melco, D800 and Blades, I did get the impression that the bass realism was a touch lacking versus the Aurender W20 and also the Cad Cat.  I had been trying this track earlier in the day, and had come to the conclusion that this bass realism was a touch better via AIR.  To be clear, the very deepest bass accuracy is a bit of a KEF Blade strong point, with many alternative and not necessarily cheaper speakers I think this might be difficult to pick up, but it struck me as interesting that this was the only track I had tried earlier where I noticed this slight week point with the mR, and the only track my blind tester picked AIR as being better.  I do wonder if this is in part a quirk of the Devialet USB input.  Why?  Well, at the OAC 'streamer day' they did a back to back test of the Aurender N10, first via USB, then via AES/EBU.  The big improvement for me, and for others in the room, was that the bass definition was better via AES.  Not more bass, far from it, just better defined and more realistic.  The Cad Cat does rather confuse this point, as it appears to reasonably good in this area, but perhaps worth remembering that I auditioned the latest version of the Cad Cat together with it's trick power supply and special grounding box widget, which is one very expensive way to get a USB output.  (Pushing £8k?  I can't quite remember)

One thing though, without going into endless specifics, next time I sit down to listen, I would definitely choose the mR over AIR, it's just that bit nicer to listen too.  So the mR wins, just, and on points, not outright.

That is not the end of the story though......    The microRendu is currently running with the slightly controversial iFi power supply.  Some say this is a giant killing audiophile PSU for budget money, some say it is the work of Beelzebub himself.  For sure, it is not the best thing to power a microRendu with, and both the mR designers and users have stated that the clean power is the route to best performance.  So, in theory, with a better PSU the mR should pull further away from AIR.  Then there is possibility of further improvements of using a USB to AES/EBU converter, such as the re-clocking Mutec MC3+USB.  What I am thinking now is that if a Mutec could mitigate this very slight bass issue, then the mR pulls away or matches AIR in all areas.  It is a very specific issue though, I think the only way to find out would be to try one.  Of course the Mutec should in theory offer improvements in other areas, which together with a more appropriate PSU should be a combination that outperforms AIR comfortably. To be continued.....

[nicoludio]

Confused, thanks for a succinct review. It does help that you have had the benefit of comparing (even if from aural memory) the mR with fairly high end servers.

Perhaps, if you have an opportunity in future, you might want to try the Melco -> USB -> Devialet vs Melco -> ethernet -> mRendu -> USB -> Devialet.

I am using the USB output of the Melco N1A and was wondering how the USB output would compare. Of course, there is also now the SOtM sms-200 to contend with.

[PhilP]

...
 Today, I had some free time, so could do some critical listening, as well as some back to back testing with the only real benchmark I have at home, Windows AIR 2.1.3.  The short version of this is that AIR 2.1.3 and the microRendu sound very similar.  

...
Of course the Mutec should in theory offer improvements in other areas, which together with a more appropriate PSU should be a combination that outperforms AIR comfortably. To be continued.....

Interesting detailed write-up and surprising (though reassuring) that two completely different delivery mechanisms can sound so similar.  

I understand you're a PC-user and so perhaps can't easily try it but AIR 3 offers very significant SQ improvements over AIR 2 which, based on the observations you've made, should put it ahead of MR/iFi.  As you've said, MR can obviously be optimised further with power supplies (and cabling) whereas with AIR tweaking is limited to cabling/router/router PS (which may or may not make a difference   ).

[Confused]

I thought it was worth adding a link to the 'Romaz' review Antoine mentioned a while ago. I stumbled across it this morning, and was particularly interested in his comments regarding the microRendu's bass performance. It looks like we might be hearing similar things?

http://www.head-fi.org/t/787020/review-c...t_12575069

[Confused]

Confused, thanks for a succinct  review.  It does help that you have had the benefit of comparing (even if from aural memory) the mR with fairly high end servers.

Perhaps, if you have an opportunity in future, you might want to try the Melco -> USB -> Devialet vs Melco -> ethernet -> mRendu -> USB -> Devialet.  

I am using the USB output of the Melco N1A and was wondering how the USB output would compare. Of course, there is also now the SOtM sms-200 to contend with.

I can't see how Melco -> Ethernet -> microRendu -> USB -> Devialet could be made to work?   Via Ethernet the Melco basically works as a NAS drive.  To use the mR, you would need to load the mR apps onto the Melco and running the music server software on the Melco?  I am not sure if this can be done, I think you basically need some form of computer to run a mR. I might be wrong?  Can any Melco wizards advise?

[Confused]

One question to those running the microRendu > Mutec > Devialet, any thoughts as to if the Mutec has a positive influence on bass definition / accuracy?

[nicoludio]

@ Confused, the mR can work as a uPnP/DLNA renderer/player. The Melco acts as the server.

Just install MinimServer (I am sure that Twonky works too) and connect to the mR via ethernet, using Kinsky or Lumin as the control app.

[Antoine]

One question to those running the microRendu > Mutec > Devialet, any thoughts as to if the Mutec has a positive influence on bass definition / accuracy?

I believe it improves all aspects of music reproduction. Of course you know, but I'd like to point it out anyway: you're not just evaluating the mR but also the USB input of the Devialet.

Now I'm quite sure the Devialet USB interface is holding back anything connected to it, including the mR. Using the AES/EBU input together with Mutec MC-3+ USB of my Devialet has shown a significant improvement even though in theory it would seem is the less preferred path because of the additional conversion to S/PDIF. However I don't know what Devialet does internally with the output of the XMOS USB transceiver and if a I2S output is used or S/PDIF through that ASRC on the DUET card. In any way I'm sure the Mutec uses better clocks in that process so it could just be that the resulting S/PDIF stream is actually better directly into the AES/EBU input and from there takes a shorter path into the FPGA and DAC sections of the Devialet.

Another theory is that the processing of the USB signal itself, which causes noise and dirties the ground planes, can be better done in a device outside/in full isolation of the DAC. This is also accomplished by using the Mutec. And even it itself also fully isolates the USB input through galvanic isolation. The Devialet does not have galvanic isolation of the USB input.

I do have to confess I have never bothered to listen to the mR directly connected to the Devialet USB input and it's likely I never will. It's been ruled out earlier in the game for me and it's disabled as an input now.

How many hours of playback do you have on it BTW? Many, including the designer, recommend letting it play for at least two days straight (around 50H) before critically evaluating it's performance. Whatever way one thinks about break/burn-in it won't hurt.

Personally I don't know yet if or how much a better PSU will improve SQ as my PH PSU hasn't been delivered yet. Plus I might never be able to clearly tell since in my case the Mutec is in between the mR and the D. I probably will power the Mutec with it as well. I do know that the mR gave me a significant improvement over the DIY music server I used -despite- the Mutec being there also so I will surely be not surprised if the PH PSU improves things further.

PS: I'm also using ethernet to fiber optic converters. Still have to try once without but haven't 'bothered' yet. Romaz you linked to, for example, has tried as well and couldn't discern a clear difference but I do know others did. Much of course also depends on the DAC (e.g. it's ability to reject jitter) and I believe his, the Chord Dave, to be a truly SOTA DAC, way better than our D's.