Sonore microRendu

[Superdad]

Hey guys:

 

Just wait until Jesus and John have the time to explain all the cool tricks--both h/w and s/w--about this thing. Then people will really be falling all over themselves for one. There is tech stuff in this that really going to surprise…

 

John is taking a trip this weekend to see a family member who lives about 70 miles from me. There is a chance he may have time on Saturday to swing by--and he might just have one of the microRendu pre-production units in his pocket. My DAC is warm and ready!

 

From what I understand, launch date really is drawing near. Quite exciting. Only wish we were going to have the UpTone mystery 1A LPS ready in time. They will be the perfect pair.

 

--Alex C.

[Superdad]

Jesus or John,

Should the microRendu be connected as close as possible to the DAC for best performance, as is recommended for the Regen? Thanks.

 

I'm not Jesus or John, but yes, as close to the DAC as possible is desirable--and part of why they put it in a case only about 25mm longer than a REGEN.

[Superdad]

John, is the circuit (essentially) an improved AMBER Regen?

 

Does not really apply.

 

What made the original "green" REGEN (these were a change to the LED color to differentiate without having to change/mark the case)--of which there were only ever 98 units--all upgraded long ago, and we recently passed the 3000 mark, into an "amber" REGEN were the addition of resistors on the USB input ground and shield lines. They made USB ground line noise take other paths, and for some reason made a remarkable difference in bass detail.

 

But the input to the microRendu is already galvanically isolated Ethernet, so there is no place or point to add the resistors that made the REGEN an "amber."

 

As for what makes the "REGEN circuit" in a microRendu improved over what comes in my USB REGEN product, it is just an even lower noise voltage regulator dedicated to the hub chip.

 

Trust me, from all the inside details I know about the microRendu hardware (Jesus is likely asleep on the east coast when John is up late at night and wants to chat--so he calls me )--including a really big technology surprise yet to be announced--the major musical magic of the microRendu owes relatively little to the "REGEN" hub-chip output circuit.

 

So far I have only heard John rave about what he has been hearing (and he is not one to exaggerate), but in about 29 hours he will be at my place with both the microRendu, and his Bottlehead DAC (along with some test boards for one of our projects together). I can't wait!

I only wish he was going to be able to stay and play longer (he sings in his church choir and has a solo performance Sunday morning--he lives 3+ hours from me).

 

Guess I may be the first with a report about the sound from a pre-production microRendu (powered by a JS-2 of course). I'll have to check with Jesus to find out if he is cool with my telling about what I heard--especially given that there are a couple of small things that might change and advance it even further. But I'm making preparations and pulling out the adult undergarments!

 

--Alex C.

[Superdad]

 

Also, excuse my ignorance, but are Sonore and Uptone 'related' somehow, or is this product a joint venture or something like that?

 

No, it is just that John Swenson designs circuits for both of us. Sonore and UpTone have been fortunate enough to share John's genius for some time. We--separately--mix John's inspirations with our ideas, market knowledge, company goals, and production capabilities to bring forth world-class components.

Certain techniques John discovers along the way remain his IP (intellectual property) and may get incorporated in more than one product. The REGEN circuit being an example. Other more intensive development projects/technologies may stay somewhat exclusive with our respective firms. It is somewhat fluid and John, Jesus, and I work things our among ourselves informally.

 

Hope that clears things up a bit. I'm really just a fan and champion of the microRendu, Sonore, and John Swenson just like everyone else. I've just been privileged to watch is develop behind the scenes over the past year and I hear details as the preamble to my calls with John before we get to the engineering discussions his projects with me (such as the forthcoming FPGA-based LPS and other goodies).

 

--Alex C.

[Superdad]

Also' date=' I own the Intona usb isolator and very much like what it does in my current setup. If I put the Intona between the microRendu and my USB/AES converter, would that create problems or detract from the benefits of the microRendu?[/quote']

 

Hi Blake:

I have an Intona (in need of firmware update but I've been too busy to send it back to Germany and have other solutions anyway), so if there is time at the end of John's 3.5 hour visit this afternoon (we have a LOT of stuff to try in that short period), I might be able to try sticking the Intona after the microRendu proto.

 

However, I really doubt it will be of benefit because:

a) The microRendu is already galvaniclly isolated from the rest of the system via Ethernet;

b) The microRendu is extremely quiet from a ground noise perspective; the whole thing uses linear regulators and all superfluous processor circuits and systems are turned off;

c) The USB output of the microRendu is highly optimized for signal integrity and impedance match, and the Intona (with its own technical limitations which I won't get into) is only going to modify and/or degrade that signal into you DAC.

 

But as I said, we'll lend it an open ear/ open mind--if we have time. Otherwise you'll just have to wait and try it for yourself when you get a microRendu. That is always the best way to decide anyway.

[Superdad]

As expected, John Swenson came up to my place in the mountains yesterday for about 4 hours. He brought with him a variety of goodies, and while much of the time was spent fiddling with prototypes of a project we are working on, the highlight of his visit was hooking up and listening to a pre-production unit of the microRendu.

 

I am a lousy reviewer, and Jesus has asked me not to steal the thunder of the planned professional reviews. Plus it was a pre-production piece and John was still tweaking some of the code--turning off PLLs for unused sub-systems, and also loading the code for a totally startling trick which I am not at liberty to reveal.

 

So this is NOT a review. It is a warning.

 

Warning to everyone who has spent lots of time and money building a tricked-out CAPS or other player/streamer:

 

The microRendu is about to change the game! It is that good. And it is SO EASY to use! This was my first exposure to SonicOrbiter OS, and I was blown away by how easy Jesus and Andrew have made activating the device for different modes of operation. And the microRendu just shows up perfectly and automatically. We mostly played it as a NAA for HQ Player, but also tried it as a SqueezeLite endpoint for Logitech Media Server.

 

Please don't ask me questions about power supplies, etc. I won't answer.

 

You have been warned. microRendu is going to be a BIG hit. I am hoping I can get in on the first run.

[Superdad]

Jesus: I'm curious about your recommending Cat 6A over Cat 7 cables. I recently went the other direction because I found that the individually shielded leads made a noticeable difference in actually decreasing any stuttering I was getting from playing HQPlayer at DSD256 through the Sonicorbiter SE. Neither of these are "audiophile" cables, but just well constructed data cables with higher quality terminations. Is there something about 6A that you feel is important in this context?

 

I'm not Jesus (obviously! ), but I'll toss my $0.02 in: First off there are a lot of garbage "Cat 7" cables out there. Bought a set of Kanex on Amazon once that used copper-clad aluminum!

 

There is nothing wrong with good Cat 7 cables (though I really doubt they were what cured the stuttering you had with HQP at high speed unless what you were switching from was junk), but what I don't like is the metal shell at each end which joins the shields and defeats the galvanic isolation inherent in Ethernet (though there is still a transformer in every Ethernet jack/PHY).

 

Even most shielded Cat 6a cables tie the shields at both ends. That may be why my favorite Ethernet cable is the thick BlueJeans/Belden Cat6a. It has the special Belden 10GX high-bandwith construction and is shielded, but the shields are not tied at either end. Did a 5-way shootout with 25-foot cables and the BJC/Belden was audibly better than the rest of what I tried (purposely chosen different constructions). And it is very reasonably priced (sorry, never tried any of the pricy AudioQuest EN cables).

 

--Alex C.

[Superdad]

The endpoint is designed to be lean and mean.

 

I'd rather say "lean and sweet."

[Superdad]

Question; in what mode were you running the microRendu? IOW, what player software was used? Have you tried the microRendu as an NAA (although that seems so overkill!!)?

 

Hi Ted:

 

Why would you think that the microRendu as NNA would be overkill? Seems to me that is possibly the purest and best mode for it. (I'm not a big LMS or DLNA fan). That's how John and I spent most of out time listening to it when he brought it to my place. Was fantastic as NAA. Vastly better than the SQ I got from a Cubox-i4Pro (as NAA with Miska-provided, optimized, text-only Linux microSD card image) last year.

 

Best,

--Alex C.

 

P.S. Although Barrows emphasized the benefit of a really good PS for the microRendu (and I agree), it still sounds great too with lesser power supplies. Lot's of people are going to be VERY happy with this device.

[Superdad]

Are those three things all I need (or in the case of the LPS, might want)?

 

Dave, who says yes he needs an Ethernet cable to attach from the router to the microRendu and a USB cable to attach the microRendu to the DAC but those are trivial in terms of setup

 

[EDIT: I started writing the below reply about 4 hours ago, but then got distracted with work in the shop. So while my comments may not seem to follow the direct of the past few hours in the thread, they are still correct and perhaps helpful for those that have a simple network.]

 

You got it Dave! And you will be very surprised at how easy it will be to set up. SonicOrbiter OS (at least with the microRendu; I've never used SO OS before) has a web browser interface. So on your regular computer you first type in an address to a Sonore server, then it looks for a microRendu on your network. Upon finding it, it opens another browser window that is pointed at the IP address of your microRendu and then you have a simple interface to choose what mode you want the microRendu to operate in (NAA, SqueezeLite, DLNA, Shairport, etc.). AFAIK, after that first discovery of the IP address of the microRendu on your network, internet connection is no longer needed (so chill any privacy paranoia).

 

I may not be remembering the above entirely correctly (I'm sure Jesus will pop in) as John was only here with the unit for a few hours (much of that time spent on other things), but I was struck by how easy and reliable it seemed. We were plugging and unplugging the microRendu from my network (we were experimenting with it as NAA and also trying to chase down a stutter that turned out to be an HQ Player issue), and the microRendu always showed right back up on the network with the same IP address (no need for the Sonore internet server discovery step). And hot swapping the DAC on its output worked flawlessly as well.

[Superdad]

Yes, iPeng or any other squeeze controller. I'm currently using two different android phones, LMS web interface or an SBT as a controller for my prototype microRendu. Yes the touchscreen of an SBT can be setup to control other squeeze players! (including the microRendu)

 

John S.

 

What about an Apple Watch? Can I browse my entire music library on my wrist? Can I, can I Mr. Wizard!! I'll buy 100 microRendus right now if it can. Or maybe I'll strap a DAC/headphone to my other wrist, and the microRendu to one ankle and the battery supply for everything to the other. Oh wait, I need a network connection. WiFi mesh network arriving next month… ;)

[Superdad]

John,

 

Is this true for USB DACs like the Ayre QB-9 that is opto-isolated from USB and runs Rankin's Streamlength Asychronous code and does not rely on the computers clocks?

 

Nick

 

Hi Nick:

 

Virtually all USB DACs use async-USB inputs these days. Almost nobody sells a DAC with adaptive-USB anymore.

 

With regards the opto-isolators. All DACs that use digital isolators (be they RF type, opto-type, capacitive, or John's favorite magneto-resistive), place them AFTER the USB input PHY and processor (and generally reclock as they all add jitter). But by the time the signal becomes I2S headed towards the DAC chip, much of what happened at the input stage (ground plane noise, jitter, etc.) is somewhat baked-in. That's because those isolators isolate just the ground planes--the signal has to pass.

 

This is why the REGEN is effective even with all the "galvanically isolated" DACs--much to the chagrin of some of the designers. So yes, that applies to the Ayre QB-9 as well.

 

Hope that helps.

 

--Alex C.

[Superdad]

With all due respect, this is just false. If it is possible that noise/jitter is actually baked into the USB signal, and if you are implying that jitter is additive, then why do the intermediate conversion from Ethernet -> USB -> I2S? You can't be suggesting that the USB step actually removes noise? Certainly not.

 

There is no such thing as "baked in" noise/jitter etc unless the bits themselves are altered. Likewise the "isolators" are free (and strongly encouraged) to use more sophisticated techniques than merely isolating the ground plane.

 

There are different strategies that DAC can use to "clean" an I2S/DSD signal, and it is pure utter hogwash to suggest that such cleaning is better applied at the USB stage.

 

Hi Jonathan:

 

First off, I apologize to Jesus and all for going off topic. I did not even pay attention to which thread I was in when making the above post.

 

I am not the engineer here, nor do I pretend to be--and it is clear that I was inelegant enough in phrasing my comments to the extent that you misunderstood my points. So let's see if just stating a few non-controversial facts clears things up a little:

a) All digital isolators add jitter, typically a few hundred picoseconds worth;

b) All USB processing chips (be they programmed FPGA, XMOS, Cmedia, etc.) add a similar amount of jitter as well;

c) Reclocking flops are most often used after isolators to knock down some of that jitter.

 

So as to a perfect signal coming out of the isolators, that is not the case and it is not their job. (And my understanding of them pretty much stops there.)

 

But in no way was I suggesting that "cleaning of an I2S/DSD signal is better applied at the USB stage." My point was simply that isolators and reclocking after the USB stage do not completely immunize a DAC from what comes before. That's why people can still hear the difference between USB cables and why the REGEN works (making the DAC PHY's job easier such that it does not generate as much noise inside the DAC--noise that is pernicious enough to make it through isolation and to affect the DAC master clock).

 

Getting back to the microRendu:

To the extent that Ethernet's galvanic isolation keeps noise out of the device that is good. And while the Ethernet PHY does generate packet-noise, that is all before the entire iMX6 processor and John has done an outstanding job of separating power domains (you'll have to ask him how many there are in the microRendu), and moreover the clocking scheme and USB output sections of of the product are extraordinary (Jesus and John will reveal more about that in due time).

So I am not suggesting that the USB output is "cleaning" the I2S. Frankly I am not even sure if/where a distinct I2S signal is even generated in the microRendu as the USB sub-system is part of the processing module itself--which in turn feeds a hub chip for final REGEN-like signal-integrity/impedance match, again to help the DAC that follows it.

=========

 

And just to flesh out the above with words from the person I learn from, here is an excerpt from a post of John Swenson's in response to a question about why DACs with isolators (always after the USB input stage) still are audibly susceptible to upstream variables:

 

The isolation helps but is not nearly the panacea many people think. Lets travel through the system and look at both the power and signal and what happens to them as we go through the system.

 

So lets start with a USB receiver with bursts of high frequency noise on both the power and ground planes. This PG (power/ground) noise will modulate the data being sent to the isolators. It will slightly increase jitter and the amplitude of the pulses will vary with the noise.

 

This noisy power also goes into the driver side of the isolator. The signal going across the barrier (light, EM waves, magnetic field etc) gets modulated by this PG noise as well. The PG noise also changes the threshold of the input receivers, adding jitter to the signal.

 

On the other side of the barrier we have a couple things happening, the varying signal level, caused by the PG noise in the driver, also causes the receiver current to change, even with no signal applied. Thus the receiver causes PG noise on the "clean side" directly related to the PG noise on the "dirty side" It is definitely attenuated, but not by nearly as much as most people expect. Then we also have traditional logic noise caused by the fact the output is a normal logic signal, every time the output changes it creates noise on the PG planes on the clean side. The jitter on the signal created by the PG noise on the dirty side is still there PLUs jitter introduced by the isolation scheme. This jitter changes the spectrum of the logic noise on the PG planes on the clean side.

 

So then we feed the signal through a reclocking flop, which is supposed to get rid of all that jitter on the input. Well it helps, but no reclocking flops are completely effective. The PG noise at the flop still causes jitter to show up on it's output, PG noise changes the threshold where the flop detects the "switch" of the clock, thus increasing jitter on the output.

 

The result of this chain is that PG noise on the "dirty side" can still make it through to the "clean" side. It IS attenuated, but not completely gone.

 

Cascading such stages can theoretically help, but in order for that to work the reclocking clock has to get fed back through the isolators which significantly degrades the clock so it turns out cascading doesn't help much. (a two stage cascade does make things better, but not by a huge amount).

 

 

Again, very sorry for the off-topic.

 

--Alex C.

[Superdad]

If all you want to do is have the microRendu find your music, then hang the music off any computer in your network and throw the free Minimserver app on that computer, pointing it to your music. The microRendu will find it and play it.

 

Or if you want to enjoy the wonderful SRC filters and/or SD modulators of HQ Player, then just choose NAA in the microRendu's web-interface menu and then NAA as back-end/device in HQ Player preferences (on your desktop/laptop) and you are good to go! Drag albums, open playlists, or use HQP's library function (I never have) to play music from any local or network-accessible drive.

[Superdad]

Right Firedog.

And the setting of the microRendu's output mode and any other configuration changes is via a web interface. That is, the microRendu hosts is own local web page. Once you know its IP address on your network you just go to its page. And Sonore has made the discovery of the device and its IP address really easy: You just point your computer browser to some internet web page and it finds your microRendu and then opens a new window for you right to the control panel. Elegant genius really. Designed for dummies like me!

[Superdad]

Thank you, firedog. So, for example, LMS runs on a networked device somewhere (computer, NAS, or server) and Squeezelite runs on the micro Rendu that is next to the DAC. Correct?

 

Yup. Exactly.

[Superdad]

We try to keep all our tax lawyers fully employed with byzantine and highly complex rules.

 

Yes Larry, but at least we don't have sales tax rates that approach or exceed 20%. As you know, VAT in Europe can run from 15-23%, whereas here in the states we typically pay 8-9%. And some customers in certain Asian or Eastern European countries have told me that import duties can run as high as 100% of declared value! Ouch...

[Superdad]

I have a Bryston Mps-2 power supply for my Bryston preamp,can I use that power supply to power the microrendu and if I can where do I find a cable for that?

 

That supply will most likely not be appropriate for the microRendu. The MPS-2 has a 12V/350mA output. Too high a voltage and too low on the current capability.

[Superdad]

Can you claim back the sale tax at the airport when you leave the country with the mRendu (like in the case of VAT/GST)?

 

No you can't, but unless you are having the unit shipped to an address in Florida I don't think Sonore will charge sales tax in the first place. For mail order-only businesses in the USA, it is only required that the seller collect sales tax from purchasers located in states where the seller has operations. That includes warehouse locations, so Amazon and other big sellers who have shipping operations in many states collect sales tax on everything they themselves ship.

But small firms who ship from just one location only have to collect tax from buyers in their own state. So UpTone collects sales tax just from buyers with California shipping destinations, and Sonore does so on sales made in their home state of Florida. The law says that BUYERS (not the sellers) who purchase goods from out-of-state are supposed to report their own purchases on their annual state income tax form and pay the sales tax on what they bought that year. But nobody does that!

 

I can see it now, audiophiles are going to ask Sonore to sell the microRendu in airport duty-free shops!

Good grief...

[Superdad]

Uptone Audio JS-2, pretty darned close to $1k.

I suspect this will be the reference supply for the uRendu.

 

Well, while the J(ohn)S(wenson)-2 choke-filtered, power-factor-corrected, dual-output, isolated ground, 5-7A LPS is terrific (but parts-costly and labor intensive--and mostly in short supply), it is perhaps a bit overkill for the low current requirements of the microRendu.

 

Yet the other Swenson masterwork (delayed a little due to his focus on finishing microRendu hardware) may be the ultimate match for the new Sonore piece. 3.3V/5V/7V at 1amp and 100% isolated from the AC mains, with radical, yet-to-be-revealed tech requiring an FPGA (and even a microSD card for potential firmware updates), we are hoping this new piece will land at under $300. And the small 11cm x 11cm case even comes from the same enclosure series as both the microRendu and the REGEN.

 

Apologies to Jesus for plugging a product in the off-topic discussion of power supplies. I won't be offended if you delete this semi-shameless post. It's just that there are a lot of Sonore/Swenson/UpTone fans here and I find myself getting more excited about our new LPS for use with the microRendu than I am for it to be used for REGENs, Squeezbox Touch, USB cards, and SSD.

 

Please take any questions about our "mystery" supply over to its ongoing thread:

http://www.computeraudiophile.com/f27-uptone-audio-sponsored/uptone-audio-regen-power-supply-add-24963/index22.html#post512063

 

Cheers,

 

--Alex C.

[Superdad]

I can't wait to see what cleaver wording reviewers use to compare it to devices costing thousands of dollars more.

 

No doubt something such as "Size doesn't matter."

[Superdad]

There are no upsampling/filtering options with the Squeezelite output mode.

 

Can't LMS itself run the SoX SRC engine? I recall doing this at my place with John once (that's how we tuned the filter for the BottleHead DAC; he wrote a routine to output the final coefficients and then another to load them into the DAC's FPGA), and I thought we used SoX on the server side.

The high-horsepower server side is where the heavy lifting of upsampling ought to take place anyway. I'm not much of an LMS guy, being in love with the filters and SDM of HQ Player (but not its user interface), so when I buy a microRendu I'll use it as an NAA.

[Superdad]

Hi microrendu design team! Did you guys also put something like this into the microrendu?

 

https://www.we-online.de/web/en/electronic_components/produkte_pb/application_notes/usbeple.php

 

I sure hope not! I personally don't like the sound of common-mode chokes on the USB data lines. That's what they do in the Jitterbug and I think that their interaction with USB cables of varying impedance is what some people enjoy the Jitterbug effect while others hate it.

 

John might also chime in with his trepidations about heavy ESD filters on data lines (Wurth uses a diode network for a portion of that).

 

But I can assure you, the magic of the microRendu's USB output goes way beyond what filtering devices are doing--and even beyond what its built-in "REGEN" hub-chip circuit does for nice signal integrity and impedance match. One of the breakthroughs (of the overall product) has to do with how the microRendu implements the USB subsystem of the processor. Jesus will get mad at me if I say more than that.

 

Give them some time, let the product get out there, and then I am sure some (but not all) secrets will be revealed.

 

 

[EDIT: 4:31 p.m. Clearly John is a faster typist than I am! But I did somewhat correctly anticipate his response before ever seeing it. ]

[Superdad]

If the mR equalises the sound (as has been implied but not confirmed) then choosing Roon over A+ is likely for me - what then will become of our beloved audio engines. I sort of want this to happen but I massively admire what Damien (and others) have done and would be sad to see their efforts come to naught.

 

Hi Andy:

 

I was an A+ user for a long time (all pre-2.x versions) precisely because of Damien's excellent implementation of integer and direct modes--he really seemed to cut through CoreAudio's clouds--and most especially because of the full iZotope Advanced SRC engine and parameter controls. I spent a long time tuning a filter that made my NOS PCM1704K DAC sound miraculous.

 

And then I took HQ Player for a spin and discovered the wonders of Miska's poly-sinc filter family, as well as his excellent dithering algorithms. After that there was no looking back. Sonically I can not create a filter in A+ that can rival what HQ Player can do.

 

This all relates directly to your interest in the microRendu and your conundrum regarding where to have the SRC engine:

 

a) One of the microRendu's mode options--and the one I intend to use exclusively--is as an NAA for HQ Player. So run HQ Player on your desktop or laptop (or on the DAC-attached machine that the microRendu displaces--that's my plan), do all the SRC and SDM you want--to rates at the limits of your DAC, and that gets sent to the microRendu/NAA.

 

b) Although I have yet to try it, I understand that Signalyst and Roon have worked together to allow Roon to function as a front end to HQ Player (whose own extremely primitive UI could only be loved by an audiophile). So if you embrace HQ Player for SQ, you can also have the joys of Roon.

 

As for your other pondering about the most appropriate place--computer or DAC--for SRC filters, that has been discussed at length elsewhere. The nutshell of it being if you find a DAC whose--by definition resource-constrained filters--please you as much as what can be accomplished with really excellent algorithms (see HQP) running on a powerful computer. I have yet to hear it in my own system, but I am told the Schiit mega Burrito filter might be one of the likeliest candidates for a DAC-based filter to live with.

 

Good times, good choices...

[Superdad]

...just tried again HQplayer on AIFF files : I get more air than A+, but also a thinner sound. I can't say I prefer HQplayer over A+ , it's just...different. More warmth/natural with A+, more precision/attacks with HQplayer. But I never tried all the Hqplayer filters/upsampling (I never liked upsampling in A+ or with any player, so...). At this time, I don't see why I should move to HQplayer...

 

Sorry for straying off topic--my last post on this s/w matter I promise:

If you are not going to do any s/w upsampling (which just means you like the sound of the upsampling filters and delta-sigma modulators in your DAC), then there is no purpose to using HQ Player.

 

And I agree that, on a relatively stock installation of OS X, A+ does a bit better job self-optimizing the system and cutting through the sonic weaknesses of CoreAudio. In the past I compared HQ Player versus A+ (early 2015 versions of both) with all upsampling turned off (I used 24/176KHz source material so as not to be bothered by the aliasing artifacts from my NOS PCM1704K DAC when fed Redbook), so I think I did a fair comparison.

 

But the two are VERY close if you run a highly-slimmed OS X (72 processes/under 300 threads at idle) booted from an SD card.

As mentioned if I don't upsample in s/w then my NOS/filterless DAC passes everything--and Redbook aliasing is not a pretty sound to my ears any more, and in that department HQ Player rules.