Toslink vs USB vs Bridge II


#1

The past few weeks I’ve been testing Toslink vs USB vs Bridge II with my headphones setup.

DS Snr > Oppo HA-1 > Sennheiser HD800-S headphones, with all balanced connections all the way through to the headphones.

All power cables (DC and AC) are unshielded and starquad config. I use one big powerboard which has no filtering or conditioning between outlets, with all devices plugged into this one big board, to help keep impedances between outlets as low as possible.

Just wanted to share my observations. Disclaimer: none of this was done in a scientific way. Software is Roon in bit perfect mode.

So the Toslink source is a Pi2Design 502DAC board on a Pi3, powered by an Uptone LPS-1 (an off the grid power supply, so to speak). Some highlights are a nice ultra low noise LT3042 powering a pair of NDK oscillators which help to lower the output jitter. A Lifatec glass Toslink cable is used (the shortest length they make).

For the USB source, I recently bought my old man (my dad) an iFi DAC2 (NOS DAC) to use with an ultraRendu so he can play with Roon’s DSD up-sampling. He’s been feeling left out and a little jealous about my DS I think, so he’ll be able to use Roon up-sampling to DSD256 with the iFi DAC2 shortly. So I used the ultraRendu to test in my system before I give it to him. This was powered by the Uptone LPS-1 also (off the grid). USB cable used is the High Speed certified 90-ohm spec Supra USB Cable, their shortest 0.7m length.

Now the Bridge II input is FMC isolated but recently I had the switching regulator replaced by a low noise linear regulator added and caps replaced. The DAC side FMC is again powered by the Uptone LPS-1. The router side FMC powered by a Mean Well 5V, nothing fancy. The DAC side ethernet cable used is BJC Cat 6, 1 foot in length from the FMC to the Bridge II.

Before getting the ultraRendu and tricked out FMC’s, Toslink has always sounded better than a direct USB cable from the PC or a direct ethernet cable from router to the Bridge II.

The ultraRendu is the best USB source I’ve ever had in my system (I’ve tried many of the popular USB sources) and for the first time USB sounded as good as Toslink in my system. I don’t know how this comes across, but I consider that a major compliment Ted’s Toslink input and DS’s jitter reduction considering the price of a nice glass Toslink cable vs the price of the ultraRendu. But I also consider it a major complement to Mr John Swenson for a really great USB source (USB signal integrity and all that fun stuff)

But for the first time ever, my tricked out Bridge II input (switching regulator removed and caps replaced and powered by the Uptone LPS-1) has beaten Toslink in my system, to my ears. I had previous tried the stock FMC’s powered by both stock PSU’s and by the LPS-1 but removing the switching regulator has made a key improvement (to my ears).

Now even though the Bridge II module has an I2S output, it is a little computer, like having a RPi3 inside the DS housing. I don’t know how much noise is generated by having this little computer inside the DS and how much extra ‘work’ is needed to power this and if ethernet signal integrity is as much a thing as USB signal integrity. There may be a tradeoff there with the Toslink input and my ears/brain prefer this tradeoff, compared with the (possibly) higher jitter of Toslink. All I have is my ears though.

Considering the cost of a nice glass Toslink cable vs an ultraRendu and vs the Bridge II with tricked out upstream components, the Toslink performance is quite incredible.

It’s been fun trialing but it’s time to pass the ultraRendu onto the old fella and enjoy the isolated Bridge II with Roon.

Even though the Bridge II is now the best sounding in my system to my ears, if someone were buying the DS and asked for my unqualified opinion of the best source, I’d say get a Pi3 with Toslink output and the nice glass Lifatec cable and be done with it. You’ll need to not only spend a lot more to better it (I think) but even then, all the interactions between cables and devices can vary from system to system, that copying one persons setup may not result in better sound in your own system (it could sound worse due to the funny interactions) - that’s the long version of YMMV.

It was a fun experience and something tells me it’s not over yet hehe.


#2

Great review!

I have noticed an amazing increase in performance on my SGCD with the Iso Regen. My Ultrarendu will be on the next shipment; I can’t wait to hear the combination with the IsoRegen.

Sadly I don’t have two LPS-1’s so I’m planning on running Ifi Power/Ultrarendu -> Wireworld Silver Starlight -> IsoRegen + LPS1w/ Sonore Cardas DC cable.

I’ve heard the double ultraR + IR is ridiculous :slight_smile:


#3
cxp said

I’ve heard the double ultraR + IR is ridiculous :slight_smile:


Nice to see a fellow mad man :slight_smile: Unfortunately the old fella will have to make do the ultraRendu on it’s own hehe.

I just realised I didn’t mention how the inputs sounded better or worse in my opening post.

To my unqualified mind and ears, I hear more jitter as more analogue sounding - softer?

And I hear the effect of ‘noise’ as some bass being robbed and maybe the top-end also being robbed a little. Among other things.

So after my tricked out isolated Bridge II was setup and tested, I now, for the first time, heard the Toslink input sounding slightly softer compared with the Bridge II input. Slighty - I’m not talking night and day differences here with any of these observations. And at the same time, compared with the ultraRendu + LPS-1 combination, I heard the isolated Bridge II as having now having more bass and more details in the top end. It could be my mind playing tricks on me or it could be the result of higher signal integrity (less noise?) with the fiber isolation + LPS-1 combination - who knows.

I could still easily live with just the Toslink input and RPi3 HAT with a Toslink output though. The cost vs performance is off the charts great.


#4

FYI the DC cable really impacts the LPS1!

The stock LPS1 cable when used with IsoRegen gives it an edgy feel, the Cardas has much improved bass weight and more detailed and natural highs!

You might want to try swapping dc cables on the LPS!

The Sonore dc cable is $140 however, if that is a bit pricey for your taste, audiobacon had great things to say about the Zenwave audio cable which is $70. I did listening with headphones on my SGCD when I switched and couldn’t believe the difference in sound.


#5
cxp said

You might want to try swapping dc cables on the LPS!

Thanks mate. I actually made my own, following John Swenson's directions with his favourite DC power cable - unshielded starquad Canare 4S6

#6

FYI the DC cable really impacts the LPS1!

The stock LPS1 cable when used with IsoRegen gives it an edgy feel, the Cardas has much improved bass weight and more detailed and natural highs!

You might want to try swapping dc cables on the LPS!

The Sonore dc cable is $140 however, if that is a bit pricey for your taste, audiobacon had great things to say about the Zenwave audio cable which is $70. I did listening with headphones on my SGCD when I switched and couldn’t believe the difference in sound.


#7
Mi2016 said
cxp said You might want to try swapping dc cables on the LPS!

Thanks mate. I actually made my own, following John Swenson’s directions with his favourite DC power cable - unshielded starquad Canare 4S6


Did you notice a positive change ?

#8

Definitely. I’ve had the Sonore cable before also and other expensive shielded cables. It’s always nice when the cheapest ends up sounding the best!


#9

I just received a balanced digital cable from Blue Jeans Cable today because I’d never tried the AES/EBU input on the DS and this 502DAC board has an isolated balanced 1/4" TRS output.

The cable from Blue Jeans is the very nice Belden 1800F ,1/4" TRS to 3-pin XLR , with a nice Canare TRS compatible plug at one end and Neutrik NC3-B series XLR connectors with gold-plated pins at the other end.

It sounds very very nice indeed, balanced from the source right through to the cans. I don’t think it’s as good as the ultraRendu I tested or the modified fiber converters into the Bridge II but again we’re not talking drastic differences at all. I feel the Toslink input is as good as the AES/EBU input in my system, comparing the same source, the 502DAC powered by the Uptone LPS-1.

I’ll be demo-ing a friends Iso Regen later this week. That will be interesting.


#10

For anyone interested, I’ve got the borrowed Iso Regen running in the setup. My mate hasn’t used it much so I’ll let the system run for 24 hours before listening.

I’ve got a Corning USB3.0 optical feeding the Iso Regen and then one of John Swenson’s 90-ohm USPCB adapters between the Regen and the DS USB input.

I was a little worried the Corning wouldn’t work because it supplies very little power at it’s female end and the Regen needs a little bit of power to power it’s input side (20mA I think). Luckily there is music playing so all is well in terms of hand-shaking with the USB hubs.

Chain is:

Surface Pro 4 & Roon > 10m Corning Optical USB > Uptone USPCB > Uptone Iso Regen & LPS-1 & Canare 4S6 unshieded StarQuad DC Cable > USPCB > DS

The DS 24/192k bit perfect test passed, as I hoped.

@yacheah the Iso Regen is only warm to touch. I definitely wouldn’t say it’s hot in my setup, even after a couple hours of playing.


#11

The unit has been playing for ~12 hours while I’ve been out. I couldn’t wait any longer - I had to have a peek at the sound.

Wow… I’ll do more comparisons with the modified Bridge II chain (modified FMC’s) and the ultraRendu which I gave the old fella, over the weekend hopefully.

It’s too early to say if it’s the best sound I’ve ever heard with my DS but it’s got to be right up there.

If you thought the original Regen was something - just wow.

I have a feeling this evening is gonna be a late night.


#12

I’m gonna play devils advocate and guess you like it so much because you are using Roon which to my ears has a slightly mechanical - not exactly organic flow to it. For me the IR dramatically increases tonal density so it makes sense it may smooth everything over and give you an impression of fullness. But… as you know… the UltraRendu is way more transparent so that is my preference.

The question is… will things change when you can get your hands on Audirvana/Windows this winter?

Also, is that Corning optical a new addition since you tested the ultraR?


#13

Hi mate

I’m looking forward to trying Audirvana just because I’ve tried every software player since Winamp (high school days) and I like trying different things to see/hear their pro’s and con’s but I don’t really understand comments about Roon’s sound - but just because I don’t understand them doesn’t mean they’re not valid. But when you properly isolate from the computer (optical + LPS-1 or Intona or Iso Regen etc), the software’s sound should be less of a factor. I say should but perhaps there’s something else.

In addition, if someone listened to Roon’s sound last year only, the sound may very well be different today and may be different a year from now. They’re constantly improving.

I’ve had the Corning optical a while. It’s 10m in length so let’s me connect the PC to the DS. The cable going into the IR matters much less than the cable going out - because the IR isolates from everything upstream of the hub. I used the Corning upstream of the IR and didn’t use it with the ultraRendu - it was better to stick with the short 90-ohm certified Supra USB cable for the ultraR output.

It’s very difficult to compare single components between different systems and try to draw conclusions. You have a different DAC (the new Stellar?), different power cables, different digital cables, different analogue cables, different amp, different speakers and more. And all of that is even before we get into how we like our music to sound. So it’s difficult (impossible) to accurately predict how inserting the same component into two different systems will affect the overall system sound.

Right now, without my modified FMC’s on the Bridge 2 I have to say the IR really betters the Bridge 2, as does the ultraRendu also (to my ears, in my system). The modified FMC’s + LPS-1 really do help both the Bridge 2 and the ultraRendu.

I’m looking forward to more listening this Sunday afternoon !

Cheers


#14
Mi2016 said . . . when you properly isolate from the computer (optical + LPS-1 or Intona or Iso Regen etc), the software's sound should be less of a factor.
Why? These are two unrelated variables.

#15
Elk said Why? These are two unrelated variables.
I followed that with "I say should but perhaps there’s something else" because I'm definitely no expert.

Let me ask Ted if he’s happy to share @tedsmith

If you have two programs that are properly setup and working in exclusive mode using the ASIO driver (I say properly assuming no system sounds interfere).

If one of those programs is severely CPU intensive and pushes the CPU to 100% usage and you have another program that only uses 2%, is it possible that complete isolation from the PC could help with noise generated internally?

Assuming we’re using the PC USB output direct to a USB DAC.

Or does this increased internal noise only manifest itself as jitter, which becomes a negligle issue with async USB?


#16

Got it. Speculation grounded in an extreme scenario. confused But nothing wrong with this, speculating and wondering can be fun.

I suggest that if better isolation results in more transparent playback, the playback software’s inherent sound will now be more of a factor, not less. That is, one is now able to better hear the differences between various players.


#17

Comparing Roon via IR and UltraR.

Using pc USB you have ASIO and exclusive mode but using UltraR should be using DLNA/MPD bypassing all of those drivers. At least the DLNA part is how Audirvana is doing it.

Does Roon decompress before it goes out?


#18
Elk said Got it. Speculation grounded in an extreme scenario.
An extreme example only to demonstrate and perhaps have a better understanding, that perhaps these variables are not as unrelated as you stated as a fact. Perhaps they are unrelated but I don't think we can say with your earlier absolute certainty.

With proper isolation and with bit perfect playback and assuming proper exclusive mode settings are both setup and working, it becomes quite complicated to understand how software can sound different.

But our ears and brain are pretty special things and should never be discounted so if people hear the differences there must be other things.


#19

@Mi2016 give the ISO Regen a try with the 9V iFi iPower and post back about the heat and the sonics please vs the LPS 1?

Thanks


#20
cxp said

Comparing Roon via IR and UltraR.

Using pc USB you have ASIO and exclusive mode but using UltraR should be using DLNA/MPD bypassing all of those drivers. At least the DLNA part is how Audirvana is doing it.

Does Roon decompress before it goes out?


Eep now this is getting into the mechanisms of networked audio protocols and Roon uses RAAT not DLNA/MPD.

The differences between protocols is complicated stuff and may (or not) be where the software sound differences come from. But this is really way over my head. It’s complicated stuff.