Compatible I2S source devices

I don’t have a directstream dac but I do have a gain cell dac and the singxer su-1 usb to i2s bridge works GREAT! Especially my modded version from kitsune audio. Made such an improvement over direct usb to my pc that I am amazed because I previous had a schiit wyrd and lh labs revive usb fixer and they either did nothing or only made subtle improvement.

I also found a great budget short flat hdmi cable with silver plated conductors and shielding. Hope ebay links are ok, I’m not the owner of apollo av I promise

http://www.ebay.com/itm/Apollo-AV-Lightning-v2-0-5M-HDMI-5-Silver-plated-OFC-v1-4-High-Speed-w-Ethernet-/361615375320?hash=item5431f4bbd8:g:vu8AAMXQVT9TE-oy

I also just got an SU 1. Very noticeable improvement on my DS snr. I currently have a no name HDMI cord feeding I2S and an old IEC power cord I had laying around powering it. Have a Wireworld HDMI on the way 0.3m. As good as this sounds with these no name cords I am eager to see how it will sound with some better cords in the mix. I bought my SU 1 from Kitsune as well. I don’t recall seeing any mods for the SU 1. How was your unit modified? This was definitely worthwhile in my system.

Thanks

Brent

There’s a modded version from kitsune available with changed voltage regulator, power caps, main pcb caps fuse and hook up wire

https://kitsunehifi.com/kte-kitsune-tuned-edition-singxer-su-1-tuned-psu-amongst-other-options-as-a-package-deal/

So, I am thinking of building a dedicated HTPC to send music to my NuWave DSD over the I2S. I have read about the Pink Faun, but apparently this only does PCM, and not DSD? Are there any PCI cards or solutions which can output I2S over HDMI for a PC or for the Raspberry Pi3B? I am not in the mood to write my own drivers or solder anything…

Hi, from my experience I like the idea to use a PC (or HTPC, or NAS) as a server placed somewhere at home but not necessarily closed to the DAC. You can install on the server any software able to manage your musical library and send the audio stream to a small an fanless computer connected to the network and to the DAC in USB or I2S. As this small computer (RPI, NanoPI, Zotac, …) will consume less power than a “normal” computer, you can feed it with a small and good quality PSU for an affordable price. Howether, I don’t know the Pink Faun PCI card.

I take the opportunity now to give a first feedback on the NanoPi. For the moment I hadn’t the time to plug the I2S/HDMI AudioGD module so I use it, in replacement of my RPI3, connected in USB to the DS. It works nicely and did not notice any change in SQ vs the RPI3. A very small box, with power profile configured in “conservative mode”, the four cores seems to never exceed 240MHz even to deal the DSD128 flow. DietPI installed on it with the RoonBridge. Very easy setup, nor more less complicated than the RPI3.IMG_20170404_083258.jpg

Euphonite said

I also just got an SU 1. Very noticeable improvement on my DS snr. I currently have a no name HDMI cord feeding I2S and an old IEC power cord I had laying around powering it. Have a Wireworld HDMI on the way 0.3m. As good as this sounds with these no name cords I am eager to see how it will sound with some better cords in the mix. I bought my SU 1 from Kitsune as well. I don’t recall seeing any mods for the SU 1. How was your unit modified? This was definitely worthwhile in my system.

Thanks

Brent


Hi Brent,

Do you mind if I ask how you are getting on with the SU-1? You still impressed with it and would you recommend it?

Also did the Wireworld cables make a difference?

My music’s on a NAS and I’m wondering if the Sonore Microrendu and the SU-1 would make a great combination to get I2S into my DS.

Would appreciate your thoughts on the SU-1 now you’ve had a bit more time with it.

Cheers,

Alan

Hi Alan,

My SU 1 sounded great out of the box as I stated earlier. I added the wireworld starlight hdmi and a new OCC power cord to it and those are currently “burning in” After adding the new cables it still sounds good but not quite to where it was before right out of the box with the no name cables. I have read that the SU 1 needs some burn in time as well so I have begun running it 24/7 with the new cables and am eagerly awaiting the sound to get back to where it was or better. I will mention that the SU 1 only has one input which is USB so I don’t think you can hook up a microrendu to it as I believe those use a network cable. Kitsune hifi has answered a lot of questions about the SU 1 on their website. I chose the SU 1 because I listen almost exclusively to streaming media and have very few ripped cd’s in my library. As it stands, I would not hesitate to recommend it to anyone trying to clean up their USB connection. The fact that it feeds I2S to the Directstream was icing on the cake for me. Perhaps I am a bit obsessive but I like that USB with all its quirks has the ability to play DSD in higher sample rates than the bridge is currently capable of. I have a Bridge 1 still and it kind of made me leary of network audio. Sounds great once you get it running but there are just too many things that can go wrong with network audio and I no longer have the patience. Bridge 2 sounds like it is easier to deal with but it is limited to DSD 64. With USB you simply install the driver and it works. That’s more my speed. Hope that helps.

Brent

Euphonite said

Hi Alan,

My SU 1 sounded great out of the box as I stated earlier. I added the wireworld starlight hdmi and a new OCC power cord to it and those are currently “burning in” After adding the new cables it still sounds good but not quite to where it was before right out of the box with the no name cables. I have read that the SU 1 needs some burn in time as well so I have begun running it 24/7 with the new cables and am eagerly awaiting the sound to get back to where it was or better. I will mention that the SU 1 only has one input which is USB so I don’t think you can hook up a microrendu to it as I believe those use a network cable. Kitsune hifi has answered a lot of questions about the SU 1 on their website. I chose the SU 1 because I listen almost exclusively to streaming media and have very few ripped cd’s in my library. As it stands, I would not hesitate to recommend it to anyone trying to clean up their USB connection. The fact that it feeds I2S to the Directstream was icing on the cake for me. Perhaps I am a bit obsessive but I like that USB with all its quirks has the ability to play DSD in higher sample rates than the bridge is currently capable of. I have a Bridge 1 still and it kind of made me leary of network audio. Sounds great once you get it running but there are just too many things that can go wrong with network audio and I no longer have the patience. Bridge 2 sounds like it is easier to deal with but it is limited to DSD 64. With USB you simply install the driver and it works. That’s more my speed. Hope that helps.

Brent

Thanks Brent,

I have the Bridge II but I’m not convinced it’s it’s the pinnacle of what I can hope to get out of the DS in terms of SQ. Perhaps it’s a burn-in issue as the DS & BII are probably at around 200-250 hrs and a few of the upstream wires and stuff are at less than that as I’ve been tinkering a bit.

I believe the MicroRendu supports ethernet input and delivers clean/optimized USB output. My NAS only supports ethernet output. So the plan I am toying with is NAS to MR. MR delivering ‘clean’ USB to SU-1. And the SU-1 then either further cleaning or maintaining a clean signal and converting it to I2S for delivery to the DAC.

I have my concerns about all those conversions. Not to mention having more boxes than an Amazon delivery center on my rack! But at the price point of the MR & SU-1 I’m happy to experiment and just sell anything that’s not adding obvious value.

I’ve already ordered the MR and should received it tomorrow (along with an LPS-1). So I guess I need to audition them properly before adding an SU-1. But I think it’s more than likely that I’ll add the SU-1 as that’s the only way I’ll know for sure if it improves the SQ by feeding the DS an I2S signal.

I’d also looked at a few more expensive options such as the Sonore Signature Rendu (now discontinued) and the Pink Faun AV Streamer. But both are significantly more money (approx $3k) so playing around with the MR and potentially the SU-1 seems an easier path in the short term at least.

If it turns out the BII was better then I’ve had some fun and I guess I’ll be busy on Ebay selling some stuff off

Many thanks for the quick reply and I hope your set-up works well once burned in.

Cheers,

Alan

@Alan, Brent thank you, interested by any feedback/comparaison with the Bridge2, sure :).

I went to the point where I change my mind. My initial project was to build a chain without an USB to I2S conversion component, I mean a chain where the computer delivers the I2S output connected to an I2S/HDMI (I2S in LVDS) module linked to the DSD DAC.
However, I do not master sufficiently how to manage the I2S output on a Raspberry or NanoPI. A HifiBerry DAC + Pro is able to deliver an I2S output and is using a quality clock (not the more rated but a good one). So, a HifiBerry DAC + Pro could be a solution and may be the awaited new Allo Kali board but not yet released.

Instead, I purchased an Armature Hecate board http://www.audiophonics.fr/fr/interfaces/armature-hecate-interface-xmos-xcore-208-usb-spdif-asynchrone-version-diy-p-11413.html

This board presents 3 interesting points:

  • Use an XMOS 208 processor to manage the USB flow and its multiple formats (DSD included) up to 384Khz and DSD256 native or DoP
  • Built around two low jitter Crystek clocks
  • Has an galvanic isolation on all of its outpus
  • SPDIF (RCA) output and I2S output
  • the Hecate board is powered by the USB port (so the NanoPI ou RPI3 in my case)

The only concern is that the I2S output is not in LVDS so the IS2/HDMI module (from AudioGD) is required for our famous and beloved PSaudio DACs.
The galvanic isolation seems to me a very good point, because even if I use a well estimated iFi Audio iPower PSU to power my NanoPI (or RaspberryPI3), I’m thinking that the PSU quality does not need to be at the quality level of a LPS-1 for exemple, because of the isolation offered by the Hecate board. And when I’ll add/plug the I2S/HDMI module on the Hecate’s I2S, I could choose a LPS-1 to power the HDMI module which consume only 30mA, thus far away from the 1A or more required by the PI + the XMOS208/Hecate board. My purpose is to build a very low power SuperCap PSU for 50mA or 100mA max which just sufficient for the I2S/HDMI module. So the PSU should require only few SuperCaps capacity at a low voltage (3,3V in nominal).

Now the sound quality 30_gif Connected for the moment on my NanoPI with DietPI and Roon, I only tested the SPDIF output. Amazing how good I gained in transparency and precision everywhere. The soundstage is fantastic. All is more real.

What surprised me is to obtain such result on the SPDIF input of the DSD senior while I believed that the best option was either the USB or 2S inputs. For the moment the Hecate board is floating in the Air as I still not installed it in an aluminum case, but very happy by the change. I can play DSD in DoP up to DSD64, I configured Roon to downsample in consequence all higher frequencies because even if the XMOS 208 processor is able to deal higher frequencies we are limited by the SPDIF specification on the DAC.

Alan, I have the uR/LPS-1too but no SU-1, I’d be interested in what it brings. I removed my LANRover as it didn’t improve SQ for me (but had a big improvement with my stock PC).

Hi Patrick - yeah I’d read some of your previous posts in this thread and it sounds a great project.

I’m personally not sure that I’m technically skilled enough to pull off a custom build but I admire the effort all the same. Glad it’s working well for you.

Hi Roger - sure - I will report back with either my observations on SQ or a “help I’m stuck!” confused depending on how I get on.

Cheers,

Alan

Hi Alan,

Your journey seems very similar to mine. Many many options available. Which is best? aye aye aye. If you end up incorporating the su 1 with the MR I would be curious to know what you discover in your system. My gut tells me that one or the other would most likely be redundant from a sound quality standpoint. Although adding the su 1 does get you the I2S input you are looking for. I don’t remember if the MR offers isolation or not. If not then I would expect the su 1 to be an improvement. When I was considering the MR one thing that steered me toward the su 1 was all the unknowns when in it comes down to which power supply to use for the MR. Again many options which can get quite expensive and I was unable to discern which option would offer the best sound at the best price. That’s why I opted for the SU 1 with its “built in” power supply. Perhaps less “tweakable” but at least I knew what I was getting pretty much. In the past year I have upgraded most every component in my system save the amp and I am at a point where I just wish to let everything settle in, listen and enjoy without obsessing (for as long as is possible for an audiophool) with what needs tweaking next. Please do keep us abreast of your findings.

Brent

@bigalmc thank you Alan, not sure that the technical skill to build a custom setup with a good PSU, a RaspberryPI and the Hecate board is finally not so difficult. The Hecate board exists in a another version with a box, you just have to plug it if you don’t need an I2S output. It is a close solution to the SU-1 (same processor, same clock) the only difference, if I’m not wrong, is the galvanic isolation you have only on the Hecate.If you discard this difference, the SU-1 has a better ratio price/features. But don’t know how much the galvanic isolation can impact the sound.

But I’m wondering if the galvanic isolation is not a mean to reduce the cost of the whole solution, wouldn’t it permit to invest less in the PSU ? If the USB converter (USB to I2S or USB to SPDIF) is not connected electically to your DAC, is it really interresting to use a LPS-1 PSU which isolates the component from the earth ground ?

@Euphonite Totally agree with the big forest of options offered on the market and in discussions. I finally choose few requirements to drive me ahead:

  • Having a SBC as small as possible but able to deliver at least DSD128 with a Roon,Upnp or Squeezelite software

  • No ground loop. For that I really observed the benefit of such isolation, so very prone to believe in the efficacity of the LPS-1 from Uptone.

  • The possibility to power the last component connected electrically to my DAC with a descent PSU

  • Usage of oscillators having the lowest jitter specifications (typically the Crystek one)

All these requirements are based on many many researches found on Internet on this forum and so on… with a personnal interpretation. So, I strongly advice to not believe all my asumptions, sorry angel_gif.

This is the prototype of my new streamer, based on RPi3, Hifiberry Dac+ Pro as reclocker, hdmi/I2s output board (with galvanic isolation) designed and made by me and my friend Antonio. We are also planning to make one with my NMR board modded with Crystek femtoclock.

http://i63.tinypic.com/2qluxck.jpg

http://i67.tinypic.com/16948ci.jpg

It works perfectly with my DS, but hooked up to his Gustard X20 Pro it plays only 24 bit files. 16 bit doesn’t work.

Anybody knows why?

Hi Carletto, very nice work 41_gif

Did you experienced a very noticeable improvement with and without the galvanic isolator ?
Which kind isolator are you using and what is the purpose of the colored switches ?
Did you place the galvanic isolation on the I2S line before the LVDS conversion or after the LVDS/HDMI stage (so the isolator is the closest component viewed directly by the DAC) ?

I don’t know the gustard frown

Pat

Hi Patrick, I listened to the new streamer only once at home. I should make a better comparison vs the old one, switching between two I2s inputs.

It sounds very, very clear and transparent to me, better than old one, but it may be due to the expectations :slight_smile:

I’m too busy at work in this period, so I have no time for good testings.

Microswitches are to use different dacs with different I2s pinout. My friend has the Gustard X20 Pro.

I use an IL715 NVE for isolation, before the LVDS chip. The isolation cleans up the signals but can also increase jitter.

Got to compare with and without isolation…

happy to report that picoreplayer 3.20 fixed the noise problem on DSD!

without further hardware ;o)

great job Clive!

Hi Linvincible,

Just tried and for me stiil poping noises as usual, I’m doing something wrong ? Which Audio output are you using ? help…

I just read about the new Cocktail Audio X50 Digital Music Server, not sure if anyone has some experience…

http://www.cocktailaudio.it/cocktail-audio-x50/ X50_F.png

X50-B.jpg

  • DSD64, DSD128, and DSD256 formats are supported through hardware conversion from DSD to I²S (available on both RJ-45 and HDMI connection)
  • Support 24bit / 192Khz WAV, FLAC formats
I assume its DFF/DSF, but not DSD ISO...

So I’ve been doing quite a bit more reading about the “best” ways to feed our PS Audio DACs. I currently use a laptop USB out direct into my NuWave DSD, which sounds fine, but I hate hearing the fan on the clunky old laptop!

So, I have been considering various solutions:

  1. Build my own low end silent HTPC for around $300 bucks https://mediaexperience.com/htpc-builds/entry-htpc/

  2. Get a Ethernet to USB converter, like the Microrendu for $700 with Ifi power supply http://www.sonore.us/microRendu.html

  3. PS Audio Lan Rover instead of MicroRendu, to connect computer to the DAC but keep in separate rooms, and “regenerate” the USB… for $599

  4. Cook up a Raspberry Pi 3B with the Audio-GD HDMI output, with or without a Kali or HiFiberry DAC+ Pro as a “reclocker” first, for $150-200 total after all parts included. I have already tried RPi3 over USB, it sounded like a cheap cassette tape. Hifiberry Digi+Pro sounded very nice, but I don’t like downsampling DSD to 88.2, considering I already spent the extra $$ for a nice DAC like the Nuwave DSD.

  5. Chuck the whole thing (and time spent figuring it all out) and trade in my Nuwave for the Directstream Jr and use the built in Bridge II with my current network, laptop running JRiver in another room, and call it a day. I would have to work many extra hours for this, but no soldering, fighting Linux configurations, custom power supply wiring, etc.

Since I already have lots of music, much in SACD iso format on my NAS, I am not interested in any music servers which have hard drives, etc. I need JRiver to make it all work, since it effortlessly handles SACD iso.

Thoughts?

1 Like