Lush^2 - Share your configuration experiences

[FileMakerDev]

On 8/24/2019 at 5:51 PM, FileMakerDev said:

    microRendu > Lush^2 > ISO-Regen > Lush^2 > Lampizator Big 7

 

Correction:      ultraRendu > Lush^2 > ISO-Regen > Lush^2 > Lampizator Big 7

 

Hours of listening magic w/ TANF again yesterday. Thank you @TheAttorney.

[PeterSt]

This is what I wrote about TANF (A: xWYxxx, B: xWYRxx) after two days:

 

I immediately sense a beautiful piano. The ways this sings is present in all similar sounds. Yes, I am sure this digs out beautiful nuances with the emphasis to “beautiful”.

 

On 8/25/2019 at 2:51 AM, FileMakerDev said:

there is some extra energy in the lower midrange that resonates

 

This is indeed the occurring matter.

 

Very very well done, and again a Great Thanks, @TheAttorney!

 

 

 

 

 

[lmitche]

39 minutes ago, PeterSt said:

This is what I wrote about TANF (A: xWYxxx, B: xWYRxx) after two days:

 

I immediately sense a beautiful piano. The ways this sings is present in all similar sounds. Yes, I am sure this digs out beautiful nuances with the emphasis to “beautiful”.

 

 

This is indeed the occurring matter.

 

Very very well done, and again a Great Thanks, @TheAttorney!

 

 

 

 

 

I am hearing the same on my system with TANF.  With pianos and acoustic guitars, it's like one can finally hear the wood in the soundboards of both of these instruments. Otherwise, it's not a gigantic change from jssg360 squared. Nevertheless I'm keeping TANF in place.

 

The mechanism in play here, kinda like jssg360 2.5 with one end of the shield unconnected is curious. I'm wondering what will happen if the third shield is connected on the 'A' end vs. the 'B' end.  Has anyone tried that?

 

@TheAttorneyMany thanks for this innovation!

 

 

 

 

[TheAttorney]

2 hours ago, lmitche said:

The mechanism in play here, kinda like jssg360 2.5 with one end of the shield unconnected is curious. I'm wondering what will happen if the third shield is connected on the 'A' end vs. the 'B' end.  Has anyone tried that?

 

Yes, I did try this (WYR  WY) at the time, and it didn't sound as good as WY  WYR, but by that stage I was getting testers fatigue and it was all starting to get muddled. So don't let that stop anyone else from trying it.

[FileMakerDev]

9 hours ago, PeterSt said:

On 8/24/2019 at 5:51 PM, FileMakerDev said:

there is some extra energy in the lower midrange that resonates

 

This is indeed the occurring matter.

 

For example, in the slow movements of Werner Thomas & Camille Piazzini's recording of the Saint-Saens Cello Sonatas, the cello resonance is heavenly. Talk about good vibrations. No other Lush^2 config has done this for me... in fact, I didn't think it was possible for a Lush^2 config to do this. I came somewhat close to this lower-midrange magic once before w/ an NOS pair of Mullard rectifiers in my Cary tube amp, but the "cost" was unacceptable (diminished highs, suppressed dynamics). The reason I'm gushing so much about this config is that I get that amazing "buzz" w/o the aforementioned negatives.

[PeterSt]

For myself this is new too. But being at the 3rd day yesterday, I noticed that with the very poor sound of CCR at Woodstock and the somewhat better sound from Hendrix, in both cases the drums sound so very realistic. You can hear the full body (shell) of the tom or snare. It just literally sings. Right through that poor sound quality. Amazing.

[str-1]

On 7/28/2019 at 2:58 PM, TheAttorney said:

An update on my 3-way USB cable comparison – from the point that I had replaced my TQ cable with Lush^2 placed before my ISORegen. On first day, the Lush was a couple of steps behind the TQ. By the second day, Lush had improved, particularly when changing the default config to JSSG360cubed.

 

After 5 days, I put the TQ back. So how would they compare now? Did Lush now match TQ? No it did not. It actually surpassed TQ rather comfortably! Not bad for almost 1/3 of the price.

 

So the Lush stays. I then replaced the USPCB with TQ, placed after the IR. Surprisingly, I preferred the USPCB – despite there being about a trillion $ price difference. The sharper sounding USPCB seemed to be letting more of Lush’s considerable detail through than the more analogue sounding TQ.

Just out of interest, and apologies if you already mentioned this further back than I’ve looked, did you originally prefer the TQ to the USPCB before the ISORegen with the VBUS on the USPCB switched off (assuming it was physically possible for you to use the USPCB in that position)?  Back when I had an ISORegen (and also now with my tX-USBultra) I found things to be cleaner with USPCB VBUS off.  It does of course have to be switched on when going into M Scaler.

[TheAttorney]

59 minutes ago, str-1 said:

...did you originally prefer the TQ to the USPCB before the ISORegen with the VBUS on the USPCB switched off (assuming it was physically possible for you to use the USPCB in that position)?  Back when I had an ISORegen (and also now with my tX-USBultra) I found things to be cleaner with USPCB VBUS off.  It does of course have to be switched on when going into M Scaler.

 

From memory, I've never got VBUS=off to work, either before or after my IR, with or without HMS inserted before my DAVE DAC.

 

So I've never had the opportunity to see how VBUS isolation affects SQ. But even with USPCB not set to its best potential, the fact that I only slightly prefered the massively more expensive TQ shows that the USPCB can punch well above its weight - or that the TQ was underperforming in this particular setup.

[str-1]

1 hour ago, TheAttorney said:

 

From memory, I've never got VBUS=off to work, either before or after my IR, with or without HMS inserted before my DAVE DAC.

 

So I've never had the opportunity to see how VBUS isolation affects SQ. But even with USPCB not set to its best potential, the fact that I only slightly prefered the massively more expensive TQ shows that the USPCB can punch well above its weight - or that the TQ was underperforming in this particular setup.

Thanks.  I’ve just looked back in the ISORegen impressions thread and now think I might have been mistaken about my IR working with VBUS off.  It is nearly two years since I last used the IR (sold it some time ago) and I guess I just assumed it worked with VBUS off because my tX-U does, and it definitely sounds better that way.

[Superdad]

1 hour ago, str-1 said:

... and I guess I just assumed it worked with VBUS off because my tX-U does, and it definitely sounds better that way.

 

That’s because the ISO REGEN uses 5VBUS on its input (which we dedicate an LT3042 reg to turn into 3.3V) to power the upstream side of its Silanna high-speed galvanic isolator chip.

The SOtM tx-USB units do not offer galvanic isolation, so they do not require VBUS. That lack of isolation may also be why they sound better when VBUS is turned off.

 

As far as VBUS on the output side, that is entirely a function of the requirement of a given DAC’s USB input stage. 

[FIndingit]

Has the quest ended already?

[robocop]

Quote

Has the quest ended already?

 

Not at all I'm still getting use to my new Lush^2. I've tried 3 settings but unwilling to comment yet. It does appear most have settled on a setting for the time being but always another option.

 

Robert

[ray-dude]

This was originally discussed in the "Novel" thread, but it seemed more appropriate to post some initial impressions here.  Based on some posts by @Bricki and others, a couple weeks ago I added a 4th shield layer to my two Lush^2 cables, and I've been experimenting with various configurations since there.  Some belated listening impressions (with thanks to @Bricki for the inspiration).

 

For the below, here is my labeling convention for the various shield configurations:

 

0 (Lush^2 Black wire - USB ground)

1 (Lush^2 White wire - inner shield)

2 (Lush^2 Yellow wire - middle shield)

3 (Lush^2 Red wire - outer shield)

4 (new wire - added outer shield)

 

Before adding the 4th shield layer, my Lush^2 configuration of choice was JSSG360^3 (A: 1-2-3 B: 1-2-3).  Before that I was using JSSG360^2.5 (TANF) (A: 1-2, 3 B: 1-2-3)  With standard Lush^2, I preferred the JSSG360^3 configuration: a bit darker and more relaxed, without a loss in detail.  

 

I added braided shielding to both my Lush^2 cables, attached a solid core wire at both ends, and then insulted and wrapped the wires.  Pro tip: if you use an appropriate gauge solid core wire, it slides nicely into the jumpers on the end, making it easy to experiment with the Lush^2 connection strip.

 

First, starting with the JSSG360x2 (A: 1-2, 3-4     B: 1-2, 3-4).

 

My initial impression of JSSG360x2 was very strong, and I preferred it to JSSG360^3. However, after some extended listening tests, I was finding the extra vividness to to be a bit too much on some more challenging tracks.  That was motivation this weekend to step through some additional configurations to see what patterns I could see.  

 

Alas, in a fit of unusual efficiency, I had added a 4th shield layer to both my Lush^2 cables at the same time (one from NUC 7i7 to IsoREGEN, and one from ISO Regen to Matrix Audio XSPDIF2, which then connects to my Hugo mScaler by Toslink).  This meant I didn't have an easy way to compare the 4 shield configurations with the usual 3 shield Lush^2 configurations (DOH!)  Below notes that compare to a standard Lush^2 are from memory, and should be consider a plea to someone with more foresight than me to test a 4 layer config against a standard 3 layer config.

 

 

JSSG360x2bullseye    A: 1-4, 2-3   B: 1-4, 2-3

A JSSG360 pair nested between a JSSG360 pair.  This definitely got my attention: Cleaner/darker sound, less high energy, more relaxed presentation.  It retained the air and resolution of JSSG360x2, but cleaned up some of the fatigue I was sensing on some more challenging tracks.  I preferred it to JSSG360x2.  The new top dog.

 

JSSG360x2link      A: 1-3, 2-4      B: 1-3, 2-4

Clean, but more high end energy than the bullseye config.  Still seemed a bit better than 360x2, but still a bit fatiguing on the more challenging tracks.  Bullseye still in the pole position.

JSSG360^3+       A: 1-2-3, 4       B: 1-2-3, 4

Same as my original favorite JSSG360^3, but with the 4th shield layer floating above it.  Seems brighter than the 360^3 that I remembered, and more fatiguing.  Not sure if leaving the 4th layer floating is a wise thing to do.  I still prefer bullseye.

 

JSSG360^3.5      A: 1-2-3, 4       B: 1-2-3-4

Cools things down a bit, and takes away some strain, but still brighter than I remember with JSSG360^3 before adding the 4th layer.  Still prefer bullseye.

 

JSSG360^4      A: 1-2-3-4      B: 1-2-3-4

Better, but still seems brighter than I remember the JSSG360^3 (and yes, at this point I'm kicking myself for not leaving one Lush^2 cable with the standard 3 layers)  Makes me want to go back to Bullseye.

 

 

So based on above, I went back and reread some of the original JSSG360 posts and discussion.  In all the above configs, the USB ground line (Black on the Lush^2) was disconnected, meaning the shield layers were floating from the source and destination.  What would happen if I started bringing USB ground back to the party? 

 

JSSG360+360^3        A:   0-1, 2-3-4      B: 0-1, 2-3-4

This config connects USB ground from source to destination but leaves the 360^3 layers still floating.  Still too much brightness for me.

 

JSSG360x2s       A: 0-1-2, 3-4       B: 0-1-2, 3-4 (favorite)

So what happens if the inner JSSG360 pair is connected to USB ground on both sides, and the 2nd JSSG360 pair is floating?  This is a winner!  Great detail, less harshness than 360x2 with the more challenging tracks.  This is currently my new favorite.

 

JSSG360x2s1         A: 0-1-2, 3-4       B: 1-2, 3-4

JSSG360x2s1’         A: 1-2, 3-4        B: 0-1-2, 3-4

So what about floating USB ground on one side or the other? Still keeps the great air and detail of the 360x2s, but is a bit harsher on more challenging tracks.

 

 

JSSG360x2bullseyeS    A: 0-1-4, 2-3     B: 0-1-4, 2-3

So how about adding a USB ground connection to bullseye?  Would that bring it even with 360x2s?  Not quite.  Definitely cleaner and with less high end energy than bullseye, but I am finding myself wanting to go back to 360x2s

 

 

With the initial survey done, time for some extended listening with JSSG360x2s.   With that new baseline, I may be tempted remove the 4th shield layer on one of my Lush^2's, and directly compare JSSG360^3 to JSSG360x2s.

 

Empirical listening aside, I'm still trying to get my head wrapped around the field dynamics in play when you have this many shield layers. Given that I still don't have a hypothesis for why JSSG360^3 would be better than JSSG360, the complexity of having a 4th layer makes it even harder to understand why one config sounds different than another.   When are layers an antenna and when are they a Faraday cage?  In aJSSG360 config, aren't they both at the same time?  I need to ponder this some more, with something stronger than sparkling water.

 

 

[PeterSt]

okayyy ...

Man, what an exercise !

[ray-dude]

Complete madness!

 

Kidding aside, it still blows my mind how different shield configs on the Lush^2 can sound so different, and that has always nagged at me.  
 

Above experiments are about me trying to understand how different configs can sound so different, and test some hypotheses. At this point, I would not recommend anyone add a fourth layer, except to experiment. 
 

 

 

[elan120]

20 hours ago, ray-dude said:

JSSG360x2s       A: 0-1-2, 3-4       B: 0-1-2, 3-4 (favorite)

So what happens if the inner JSSG360 pair is connected to USB ground on both sides, and the 2nd JSSG360 pair is floating?  This is a winner!  Great detail, less harshness than 360x2 with the more challenging tracks.  This is currently my new favorite.

Thank you for suggesting this configuration.  It is working great in my system as well and is now my new favorite!

[TheAttorney]

On 10/12/2019 at 12:38 PM, FIndingit said:

Has the quest ended already?

 

See what happens when you speak too soon.

This is just the beginning!

 

At first, I was unsure about Ray's new notation style of replacing letters with numbers, but I'm warming to it now.

It will become more pertinent as we get to 6... 8... 10 shields.

 

Would that be too far fetched?

Not with the kind of crazy guys we have around here 🙃

 

[RickyV]

It's not long before we have 10cm thick USB cables that need support brackets. 

[ray-dude]

4 hours ago, TheAttorney said:

At first, I was unsure about Ray's new notation style of replacing letters with numbers, but I'm warming to it now.

It will become more pertinent as we get to 6... 8... 10 shields.

I actually prefer the other notation with the stock Lush^2, but it broke down when I added a second black wire for my 4th shield layer

 

Numbering the layers from inside to outside has helped me visualize what the different configurations physically are. For instance, for me having the USB ground layer physically closer to the data signal seems to be better than having it connected to the outside JSSG360 pair.
 

What does that mean?  Wish I had a hypothesis other than the ones that I proved wrong with these last series of listening tests

[RickyV]

I think I prefer the 0,1,2,3,4 notation, I keep forgetting which color is the inside one.

[fas42]

On 10/22/2019 at 12:46 AM, ray-dude said:

Complete madness!

 

Kidding aside, it still blows my mind how different shield configs on the Lush^2 can sound so different, and that has always nagged at me.  
 

Above experiments are about me trying to understand how different configs can sound so different, and test some hypotheses. At this point, I would not recommend anyone add a fourth layer, except to experiment. 
 

 

 

 

Not to me ... the receiving circuitry is too sensitive to the analogue qualities of the waveforms on the USB cable - everything you do to adjust, slightly, the nature of those waveforms will change the perceived sound, no matter how 'bizarre' whatever it is you do.

 

The proper, long term solution is to make the receiving side of the link far more robust - meaning that the DAC and other analogue areas don't react to this sort of fiddling.

[ray-dude]

1 hour ago, fas42 said:

The proper, long term solution is to make the receiving side of the link far more robust - meaning that the DAC and other analogue areas don't react to this sort of fiddling.

 

This I believe.  I have developed a deep distrust of the PHY receiver chipsets out there, because that is the only thing I can think of to blame.

[PeterSt]

On 10/22/2019 at 11:19 AM, TheAttorney said:

It will become more pertinent as we get to 6... 8... 10 shields.

 

Would that be too far fetched?

Not with the kind of crazy guys we have around here 🙃

 

To me it didn't feel wrong either. Btw, right now the colors go from light to strong, White, Yellow, Red. With Black as the connection to the outer world (the connector's chassis).

Bit of a problem : you'd first need to know the colors before you can apply the 1,2,3.

 

Lush^3  ?

 

Actually we have all the materials to make the additional braids en insulation layers which is related to the wider diameters of the extra layers. I could have one operational tomorrow. But then what ?

Ah, listen myself first. 🥶

 

[tims]

3 hours ago, PeterSt said:

 

To me it didn't feel wrong either. Btw, right now the colors go from light to strong, White, Yellow, Red. With Black as the connection to the outer world (the connector's chassis).

Bit of a problem : you'd first need to know the colors before you can apply the 1,2,3.

 

Lush^3  ?

 

Actually we have all the materials to make the additional braids en insulation layers which is related to the wider diameters of the extra layers. I could have one operational tomorrow. But then what ?

Ah, listen myself first. 🥶

 

Anyone recommend a braid to do these 'tweaks'?  The brand I have used in the past has been thick and unwieldy.

Thanks

[elan120]

5 minutes ago, tims said:

Anyone recommend a braid to do these 'tweaks'?  The brand I have used in the past has been thick and unwieldy.

Thanks

I have been quite happy with Flexo Shield.  It is very easy to work with, flexible, and the 4th layer I added on my Lush^2 still maintain its flexibility nicely.