Singxer SU-1 Owners
Mar 10, 2017 at 10:49 AM Post #61 of 869
Which 0.3m cable are you using?

I am using a 0.5m Emotiva cable.

I am using the $10 one that is offered on the Kitsune website
 
https://kitsunehifi.com/product/generic-i2s-hdmi-cable-0-3m/
 
Mar 10, 2017 at 1:09 PM Post #62 of 869
I have the same question.  I am actually on the hunt for a quality "shorty HDMI" to add to my inventory.  ([COLOR=0000FF]You can never have too many interconnects![/COLOR])

My SU-1 & Holo are over 1M apart, but for testing purposes, I could temporarily move the two closer together and see if it makes any difference.  


If you do this test, I have a quality 0.5FT Sewell Silverback S6 HDMI cable I can send you at no extra charge. Just cover $2 shipping.

Its the best short HDMI cable I've found available, but at half a foot you literally need to place the Singxer SU-1 right on top of your Holo Audio Spring in the rear.

With a 1-1.5M HDMI cable, I couldn't tell a difference between i2S and AES/EBU. However that changes when using a 0.3-0.5M cable, in favor of i2S. Its hard to tell though. I only hear it through a few specific songs that I use as my benchmark.
 
Mar 10, 2017 at 7:29 PM Post #63 of 869
  I have the same question.  I am actually on the hunt for a quality "shorty HDMI" to add to my inventory.  (You can never have too many interconnects!)
My SU-1 & Holo are over 1M apart, but for testing purposes, I could temporarily move the two closer together and see if it makes any difference.  

Audioquest Carbon - highly recommend. I tested AQ Coffee as well but difference in SQ doesn't exactly match difference in money you have to spend. But yes, is bit better.
 
Mar 11, 2017 at 1:55 AM Post #64 of 869
I bought one of these at 20cm. It fits with the Singxer ontop of the Holo and having the right angle connector.
Have not compared it to the other one. Also have a 0.3m Starlight 7 cable but do not like how it is putting tension on the HDMI socket.
LINK
 
Mar 11, 2017 at 5:54 AM Post #65 of 869
I've sat on the toilet and done some thinking lately. Contemplating life and somewhere in that, HDMI cables came into the picture.

dot dot dot. Awhile back when I had my Holo Audio Spring (now owned by my neighbor), I used a WireWorld Starlight 5.2 & 7 in 0.3M and 0.5M length. To me there was less bending with the 0.5M which was a plus, and when comparing to the shorter 0.3M, I could not tell any audible differences.

Later down the road I picked up a super short 0.5FT cable from Sewell called the "Silverback S6". Due to the shorter length, I had to positioned my Singxer SU-1 right above the Holo Audio Spring next to it's HDMI port otherwise the cable didn't reach. As you may know, I2S is not made for long runs so the theory was that the shorter cable would be better.

0.5FT > 0.3M > 0.5M

After over 6 hours of A/B test with my favorite benchmark songs, that theory (for me), was true, HOWEVER, only when comparing to other cables that were 1 meter or longer. The build quality of the Sewell may not be as great as the WireWorld, but it's shorter distance saved it from jitter. The WireWorld may not be as short the Sewell, but it's improved materials (wire, shielding ,connectors, and design) and build quality saved it from jitter. At the longer length of 1 meter however, nothing can save it. The losses can only be reduced. This was evident when I compared the same cable, a [COLOR=FF0000]WireWorld Starlight 7[/COLOR] but in two different lengths, 1 meter and 0.3 meter. The longer cable had more of a metallic sound to it, the digital edges were sharper and had more glare to it. This was easily noticeable on a R2R Ladder + NOS setup.

After my experimentation, I kept the super short Sewell cable, but still also kept the [COLOR=FF0000]WireWorld Starlight 7[/COLOR] (only 0.3M) as I liked the aluminum connector... I sold the 0.5M one to my neighbor...dot dot dot. Fast forward a couple weeks later. My equipment placement changed and so I could no longer make use of the super short cable (giving to @T Bone
soon)
. What's left was my 0.3M cable. Strange enough, The 0.5M cable I sold to my neighbor was a little long for him, so he went out and bought a shorter cable as he heard from me that shorter was better. Then came the [COLOR=6633CC]WireWorld Ultraviolet 7[/COLOR].. in 0.3 meter.

Sitting there beside him I thought to myself... If the WireWorld Starlight 7 0.3M & 0.5M sounded the same to my ears, wouldn't the Ultraviolet 7 as well? but wait..  the material and design is not on the same level as the Starlight 7 so in theory it should sound worse, or perhaps at this short length it shouldn't matter?

After another 6 or so hours of A/B testing with my buddy, we found the two to sounded exactly the same. Because of that reason, I will be departing my Starlight 7 0.3M and he him, his Starlight 0.5M. From now on we'll both use the [COLOR=6633CC]WireWorld Ultraviolet 7[/COLOR] (0.3M)

BUT WAIT.. THERE'S MORE.

HDMI transmission was made for audio and video. Television interlinking. Designers of audio circuits such as PS Audio (among many others) simply made the HDMI port as a standard for sending and receiving I2S data as they felt it was a good type of connection. It's more than good actually.. The HDMI connection is superior as it offers a balanced way to transmit data. 

In I2S internal protocol, these are the main bits that are transferred over.
  • DATA (Serial Data) also known as SDATA
  • LRCLK (L/R Clock, Word Clock, or Frame Clock) - also known as LRCK, WCK, or WCLK..
  • BCLK (Bit Clock or Serial Clock) - also known as SCLK, SCK, or BCK.
  • MCLK (Master Clock) - also known as MCK

In HDMI, each of these paths are paired with its own independent ground (return) thus becomes less susceptible to interruptions. There are other I2S connectors used in audio such as the RJ45 (or DIN, D-Sub, BNC, etc). Some do it better than others. Some do NOT provide their data and clocks with it's own independent ground at the input/output thus hurts signal fidelity during transfer. For susceptible paths like I2S, this increases jitter. Think of it like balanced vs single ended design. 



Unbalanced I2S, signal conductor becomes susceptible to noise (which induces jitter).




Balanced I2S, signal conductor becomes less susceptible to noise due the noise being canceled out (through polarity inversion).

Now when it comes to the I2S port (HDMI) offered by Holo Audio Spring & Singxer SU-1, I wanted a cable that was worth the money and offered the least amount of diminishing returns. This was why i settled on a $6.95 super short cable from Sewell in the first place. It provided the same performance for a fraction of the price. Sure one could throw money in and buy the best cable out there and no doubt it'll perform well, but that's like buying a pricey condom for no reason. You end up with a Magnum XL, plenty of extra "headroom", but none of which you can actually use.

Anyways.. dot dot dot..  I wanted to understand why the [COLOR=6633CC]WireWorld Ultraviolet 7[/COLOR] sounded as good as the more expensive [COLOR=FF0000]WireWorld Starlight 7[/COLOR]. Was it because at 0.3M length material science didn't matter as much? What could it be? What is the inner workings? What sorcery is this!

The Starlight 7 was clearly better ON PAPER.
  • 23 GA wire vs 24GA wire (thicker wire)
  • 24 conductors vs 16 conductors (twice as much wire for the important stuff)

I had to think broader than just reading off specifications. To truly understand how well an HDMI cable can work for audio use (I2S in particular), one must understand how it's pins are configured. Below is the configuration for Holo Audio Spring and Singxer SU-1.




If you pay close attention, the main pins used are 1, 3, 4, 6, 7, 9, 10, and 12. This carries the I2S data and clocks.

Now if you look at a conventional HDMI cable, the grouping of the wires appear so:



As you can see, The vital wires (or pins) are placed in the correct group brackets and not left floating, unshielded, or bunched together with what could be noisy. The problem with this conventional HDMI cable is those wires are not isolated enough and can suffer from cross-talk. It's group wires are only shielded from one another with the use of mylar aluminum foil. WireWorld offers both the foil as well as a copper braid over each grouping (A, B, C, and D). By improving on signal fidelity one improves on jitter (theoretically).




Now what does the [COLOR=FF0000]WireWorld Starlight 7[/COLOR] achieve?
Well for starters it's 8 core wires that carries I2S data is now twice the size (now 20 gauge). It's "DNA HELIX" design correlates to its core wires being twisted in pairs. Now what does this translate to?




A thicker wire means more power handling and greater signal fidelity (less resistance).
The intended purpose of twisted pairs is to cancel out electromagnetic interference and reducing cross-talk.
All important things in maintaining a pure signal for I2S. But where does this leave us?

My theory is that at the recommended length for I2S , the thickness in wire does NOT matter ENOUGH.
Improvements gained in a thicker cable with twisted pair, it loses in length. Length in which is extremely vital in I2S transmission where every millimeter counts. Why? The fact is that twisted or braided wires use MORE wire than a single wire at the same length. Not only that, but now there are two wires open to picking up noise, not 1. Don't forget the Starlight 7 is a thicker cable and harder to bend.

The circuit traces that connects to the input or output of those HDMI connectors are nowhere near 24 gauge. A standard HDMI cable has 28 gauge, a high speed one has 24 gauge, but that is measured in long length usages. Bottom line is, WireWorld HDMI cables are overkill for I2S transmission.. Even the affordable Chroma 7 should be more than sufficient for the task and I have doubts of it sounding any worse than both the [COLOR=FF0000]WireWorld Starlight 7[/COLOR] &[COLOR=FF0000] [/COLOR][COLOR=6633CC]WireWorld Ultraviolet 7.[/COLOR]

Why am I not getting a Chroma then? Well I don't like yellow. Over time the inner materials may break down the copper thus decrease its conductivity. I prefer having a silver plating over my conductors. Last but not least.. the metal shell looks nice and offers extra shielding.
 
Mar 11, 2017 at 11:14 AM Post #66 of 869
@Energy you must spend a lot of time on the toilet or you are fast thinker!  THANK YOU for a wonderful and educational post.  I sure as heck learned a lot and may have saved some money and time experimenting with different HDMI cables.  You are a rockstar on that toilet.  Oh by the way, some of use can actually use the "headroom" of a Magnun XL!  Thats what I tell myself anyway!  LOL!
 
Mar 11, 2017 at 11:55 AM Post #67 of 869
@Energy
 
Thanks for your experiences with different HDMI cables. I was going to order the Starlight 7 HDMI cable for the Singxer being I have the Starlight 7 USB for the microRendu (which I haven't used yet due to waiting for the Uptone LPS-1). 
 
I'm now seriously considering an alternative like the Island 7 (0.3m) HDMI.
 
Mar 11, 2017 at 7:42 PM Post #68 of 869
@Tboooe - It's a pleasure. After I'm done with "the business" I usually just sit there contemplating life. It's a vicious cycle. "how am I making money right now? What am I doing with my money? Oh i need to buy an HDMI cable."

Haha I wanted to make that analogy. If your setup is longer than mine, the increased wire, guage, and twisting of the Starlight 7 might help :] . Just pull that cable down to the base even if it can't go no more. It'll make you happy you're using all of it muhaha.

@joseph69 - I recommend going for at least a Chroma 7 in 0.3M.



The island is 28 gauge, whereas those pricier ones, like the chroma, is 23-24 guage. Just in case any signal loss occurs during transmission, I wouldn't recommend anything smaller than 24 guage.

Category cables are known to work for i2S quite well, the higher the number (CAT5, CAT6, CAT7), the better the quality (twist, shielding, gauge). I mention category cables due to some designers using RJ45 Ethernet port for i2S. This entails the use of category cables.

Even the lower end CAT5 is at least 24 gauge. But to be real honest, based on our use of 0.3M length, i doubt the Island 7 will sound any different than the Starlight 7 or Ultraviolet 7.
 
Mar 14, 2017 at 9:31 AM Post #69 of 869
thanks for all the info on i2s cables, I decided to get a bluejeanscable 1ft hdmi for my su-1 kitsune edition(don't have it yet).
maybe we should merge threads with this other thread?  http://www.head-fi.org/t/820897/singxer-su-1/45
 
Mar 15, 2017 at 5:35 AM Post #70 of 869
 
Thanks! My main reason for getting the Singxer SU-1 over the F-1 is that I want to use AES/EBU as that is supposed to be the Yggdrasil's best interface for sound quality. But I am also hedging my bets for the future. A USB DDC that supports AES/EBU, S/PDIF Coax (both BNC and RCA), and I2S is a flexible device indeed. The main drawback to the SU-1 was the power supply. Now that is no longer an issue so I pulled the trigger. It will take a week or two to get the special order built and then a week or two get here. But I am not in a hurry. What I have works extremely well.


Quoting myself....what is the world coming to?
 
I cancelled my Singxer SU-1 order. The main guy at Singxer emailed me today and told me the "the warranty is more difficult” and that the voltage fed to it had to be precisely 5vdc. 6vdc would kill the device. Plugging the power in with the power on would kill it. There were just too many issues for me to order it. On top of that, it sounded like he hadn't even started work on it so who knows when I would have received it.
 
I guess I am going to stick with the Singxer F-1 and use SPDIF coax until something better comes along at a reasonable price. The Mutec MC3+USB is an obvious option but the price is just too high to justify. I guess I could buy a used Singxer SU-1 and modify it myself. I will be trying out a UpTone Audio ISO Regen soon any way so having some extra cash is a good thing!
 
Mar 15, 2017 at 7:09 PM Post #71 of 869
Quoting myself....what is the world coming to?
 
I cancelled my Singxer SU-1 order. The main guy at Singxer emailed me today and told me the "the warranty is more difficult” and that the voltage fed to it had to be precisely 5vdc. 6vdc would kill the device. Plugging the power in with the power on would kill it. There were just too many issues for me to order it. On top of that, it sounded like he hadn't even started work on it so who knows when I would have received it.
 
I guess I am going to stick with the Singxer F-1 and use SPDIF coax until something better comes along at a reasonable price. The Mutec MC3+USB is an obvious option but the price is just too high to justify. I guess I could buy a used Singxer SU-1 and modify it myself. I will be trying out a UpTone Audio ISO Regen soon any way so having some extra cash is a good thing!

 
Aw no! I was so excited to see your work. It's that's the case then, I'd recommend sticking to the F-1 for now and wait until something with better isolation and supply comes out.
 
I'm definitely interested in the ISO Regen but at this rate with all the transitions into linear supplies, regenerative signals, isolation, reclocking, and rebalancing. I think it's a wiser choice to wait a little longer for an all in one box solution. That way the signal length is much shorter (for the cables) and only PCB to PCB distance. I feel that tweaking to perfection is easier than incorporating large groups of devices. The Singer SU-1 is indeed a good unit, but still has more flaws than which I'm comfortable with. None severe, but for end game transport, I think it's better to wait and mod something else that's better. 
 
Would be nice if they can finally offer full galvanic isolation at high speeds, clean 5V for the USB input, cleaner 5V for the reclocking circuit, better clocks, and regeneration. I don't have RJ45, but it would be nice if they added that in also for those who need one. If they offered all the previously mentioned items, the PCB would be larger and should have room to accommodate.
 
Mar 15, 2017 at 8:26 PM Post #72 of 869
   
Aw no! I was so excited to see your work. It's that's the case then, I'd recommend sticking to the F-1 for now and wait until something with better isolation and supply comes out.
 
I'm definitely interested in the ISO Regen but at this rate with all the transitions into linear supplies, regenerative signals, isolation, reclocking, and rebalancing. I think it's a wiser choice to wait a little longer for an all in one box solution. That way the signal length is much shorter (for the cables) and only PCB to PCB distance. I feel that tweaking to perfection is easier than incorporating large groups of devices. The Singer SU-1 is indeed a good unit, but still has more flaws than which I'm comfortable with. None severe, but for end game transport, I think it's better to wait and mod something else that's better. 
 
Would be nice if they can finally offer full galvanic isolation at high speeds, clean 5V for the USB input, cleaner 5V for the reclocking circuit, better clocks, and regeneration. I don't have RJ45, but it would be nice if they added that in also for those who need one. If they offered all the previously mentioned items, the PCB would be larger and should have room to accommodate.


It's clear that it helps to isolate and regenerate the USB signal before the DDC. Even DDCs like the Singxer which offer some level of isolation benefit. I doubt we are going to see a DDC with an isolator/regnerator built in. They will continue to take in the crappy USB signal and use that to convert to some other digital format. The added expense would price them right out of the market for most of their potential customers. I don't think there are that many of us out there that are willing to pay for both a DDC AND isolation/regeneration in the same box. The Mutec MC3+USB would probably cost $700 more than it does now. I don't mind two devices. It gives us more choices.
 
Mar 15, 2017 at 9:29 PM Post #73 of 869
It's clear that it helps to isolate and regenerate the USB signal before the DDC. Even DDCs like the Singxer which offer some level of isolation benefit. I doubt we are going to see a DDC with an isolator/regnerator built in. They will continue to take in the crappy USB signal and use that to convert to some other digital format. The added expense would price them right out of the market for most of their potential customers. I don't think there are that many of us out there that are willing to pay for both a DDC AND isolation/regeneration in the same box. The Mutec MC3+USB would probably cost $700 more than it does now. I don't mind two devices. It gives us more choices.

 
I don't doubt the performance of what the isolator or regenerator does when it comes to added benefits. Just had wishful thinking that there would be a single box fix, but after hearing you out, I would totally agree. A single solution like that would probably arise more problems than anything. Increasing heavily in price and diverging them from their target price bracket. Two devices and swapping out the decliner at this point seems like a better idea as long as the cable connecting the two devices is kept short.
 
I've been reading some information on the ISO REGEN over at Computer Audiophile (http://www.computeraudiophile.com/f27-uptone-audio-sponsored/iso-regen-31769/)
If they can create something truly galvanically isolated and regenerative at the same time, wouldn't the signal going into the Singxer's isolation circuitry only add more noise and jitter? If that's the likely case, It would be better to hook the signal directly to the reclocking platform. This is a hard thing to do as it will require extensive knowledge and the desoldering of SOIC's. In this case, wouldn't it be better to get a different reclocker entirely?
 
Mar 16, 2017 at 1:11 AM Post #74 of 869
  I've been reading some information on the ISO REGEN over at Computer Audiophile (http://www.computeraudiophile.com/f27-uptone-audio-sponsored/iso-regen-31769/)
If they can create something truly galvanically isolated and regenerative at the same time, wouldn't the signal going into the Singxer's isolation circuitry only add more noise and jitter? If that's the likely case, It would be better to hook the signal directly to the reclocking platform. This is a hard thing to do as it will require extensive knowledge and the desoldering of SOIC's. In this case, wouldn't it be better to get a different reclocker entirely?

 
My goal with the ISO Regen is to completely separate the computer equipment from the audio equipment and provide the DDC or DAC with the best USB packets possible. I also think the ISO Regen will allow the Singxer F-1 to do a better job since it won't have to deal with poorly formed USB packets and all the power related issues such as AC leakage, ground loops, and noise.
 
My DAC, the Yggdrasil, is supposed to have a pretty good USB implementation. So I will be testing the ISO Regen with and without the Singxer F-1. The Singxer F-1 is supposed to have very low jitter numbers. I am still learning how to listen to source material critically so I can better evaluate changes made in the computer and audio systems and how they affect sound quality.
 
I have no problem modifying an inexpensive piece of hardware. I am less inclined to risk modifying multiple pieces of hard since that adds up fast. I would not be willing to modify my Yggdrasil!
 
Mar 16, 2017 at 1:10 PM Post #75 of 869
 
I cancelled my Singxer SU-1 order. The main guy at Singxer emailed me today and told me the "the warranty is more difficult” and that the voltage fed to it had to be precisely 5vdc. 6vdc would kill the device. Plugging the power in with the power on would kill it. 

I don't see your point. It is obvious when you modding your stuff you lose warranty. It is also obvious XMOS has to be supplied by 5V.
What you expected??
 

Users who are viewing this thread

Back
Top