Lush^2 - Share your configuration experiences

[PeterSt]

All,

 

On 8/18/2018 at 7:52 AM, rickca said:

OMG we're really going to need a separate topic for the Lush^2.

 

The past weeks a countless number of times various suggested to create a separate topic/thread for the JSSG 360 Lush. Now with the Lush^2 out, this seems to be the time.

The Lush^2 applies the JSSG 360 tweak in a formal way by means of special connectors. Here you see 10 of of 10,000 or whatever countless possibilities, the 3rd diagram on the second row only showing the original JSSG 360 application :

 

 

The virtual problem is that no-one will be able to try out all the combinations with some sense. Maybe after a while we may reach some consensus like (it is only a virtual example !) the middle shield always needs to be connected to both ends for the best exhibit of the cable. Maybe people with good knowledge/experience are able to tell what foremost never should be done like never leave the source connector disconnected from any shield (virtual example again !), which will eliminate 100s of possibilities.

All I can tell at this moment is that the difference between one or an other is more vast than what I have read from many of you, unless you tend not to exaggerate - haha.
 

On 8/18/2018 at 6:49 AM, tims said:

I tried the JSSG 360 on my Lush cable and was expecting good things from the reviews the 360 tweak has gotten here but was disappointed with a drop in SQ; flabby bass and digital glare - the pre-tweaked Lush sounded great in my system prior to the tweak.

 

And that (above) can happen just the same because of what the (special) shielding achieves. And yes, with the Lush^2 I myself have already experienced the "flabby bass" up to super umphy disco bass (both not right at all) just because I applied the wrong shielding combination / config. It is easy to have it wrong but it feels like it is more easy to have it right / for the better by random attempts. But what is the best ? And how does it relate to the PC in use ? Can we ever reach consensus ? Will it be possible to find out the technical mode and merit ?

Just discuss it is also fine; I hope the result and findings will bring us a step closer again to the best sound ever. Where possible I will try to have answers to questions as well, but do notice that I do not carry any baggage and math to show you why a certain config would work out better than the other.

 

As of  October 2018 the preferred configuration all Lush 2's are shipped with is 
A:B-W-Y-R, B:B-W-R (this nomenclature is explained in posts further down in this thread, like here and a bit more natively here). See photo below.

Preferred: the general consensus.

 

Then ... since March 2019 people tend to like my new finding, commonly known now as "PNF" (Peter's New Finding). This is

A:B-W & Y-R, B:B-W. You see this below.

 

 

Thank you all (see below),

Peter

 

Because the path to the JSSG 360 appliance deserves credit to @JohnSwenson, @lmitche and @austinpop but just as well to the all so many people who ordered the original Lush the past year and the really not few who applied the JSSG 360 tweak so we could get a sense of it really working - the Lush^2 is dedicated to all these people. 

 

[BigAlMc]

Hi Peter,

 

I'll be glad to share my experience and impressions when my Lush^2 arrive.

 

As you know I ordered two 70cm Lush^2. This is because I'm trying (hoping ) to beat one or both of the following:

 

Current Lush feeding my TX-USBUltra

&

Sablon Reserva feeding my DAC

 

Given the performance and rave reviews the Sablon gets I'm particularly intrigued in how the Lush^2 compares to that.

 

Cheers,

Alan

[PeterSt]

I'll try to answer a first few questions I found elsewhere :

 

4 hours ago, rickca said:

@PeterSt what connection configuration are you currently using with the Lush^2 that gave you such stunning results with your 14/28 configured as a 10/20?

 

Side note : "14/28" refers to a processor with 14 native cores, 28 in hyperthreaded mode with the further notice that in the BIOS this can be set to have 10/20 cores used only. It is just a means of implying a very different sound (quality), most probably impeded by a different (distribution) of radiation. This is just a wild guess at this moment. Do notice however, how this fits the different shielding approach because it would protect in different fashion.

All changes are applied to the 100% same PC (Phasure Mach III) - only the processor is physically changed, or is changed for its number of active cores in the BIOS.

 

That all said, I hope people see the importance of mentioned 14/28 configured as 10/20 do NOT give the same results for the Lush^2 best configuration (as far as I went with trying configs).

 

Thus easiest example, hopefully telling it all :

I had a best config for the 14/28 processor. Next I turned that into a 10/20 (BIOS) and the same Lush^2 config gave crazy better results (FWIW of course, and I hope I can be trusted in expressing something like this, for "some" real value).

Next step was installing a genuine 10/20 and the very same Lush^2 config did not work out at all. Now the two-folded importance :

 

1. This same genuine 10/20 processor with the original Lush(^1) was more than OK.

2. This same genuine 10/20 processor turned into a wildly better result ever heard, with now an other config of the Lush^2.

 

Ad 2 - For now this is just trial and error, unless we start to see the logic (and the purpose of this thread).

 

4 hours ago, rickca said:

Is it the same as the original Lush cable with the JSSG360 tweak as some users have already implemented?

 

Rick, to answer your question : yes, but this is only because I tried it as a "last resort" to enable a nice playback session. Anyway the answer is a full stop Yes, for this genuine 10/20 situation. But for the 14/28 and 14/28 tweaked to 10/20 it was a very different (there the very first diagram of those in the OP).

So the real answer should be : it depends on too many things to comprehend as of now.

 

To be precise : although the Lush^2 can be configured exactly as an original Lush and also as the JSSG 360 Lush you people so nicely created, it can not be the really same because of the material topology not being the same. For example, if you'd put a Lush and a Lush^2 next to each other, you won't see a difference except for the connectors. It is as think, looks the same, though feels very different (Lush^2 is more flexible) and the Lush^2 weighs more. So for shielding the JSSG 360 can be applied exactly the same, but not for dielectric and with that impedance and the other properties. Also, among you guys each other, you won't be having that the same (doesn't matter and only comparing among each other is now more tedious).

Fun is, this is all unimportant because with the different shielding setup, all the properties change (at least that is how it comes to me) and you can make any cable of it (so to speak). Read : suppose you make the outer shield (of the three) the active one, then the dielectric changes because of that. Say that now all is captured within the cable inside of the 3rd shield. This is different from all being captured within the innermost shield.

Again and please, this is how it is in my head. It could be the largest BS as of now. All we already know is that the SQ differences are huge AND that the Lush^1 is all about the cable's properties (and not really about shielding sec (as the only matter) at all). IOW, we may think we change the shielding (screening) but what may happen in reality is that all the properties change (impedance, capacitance, conductance) and not even in a linear fashion (like capacitance and inductance are normally related).

 

[PeterSt]

I don't really understand why one particular connection configuration may not be optimal for different PC's.

 

Me neither, but I hope we are going to find out. But a rough first attempt :

 

First we must understand how the Lush^1 was able to change the sound to how it does that (I think it is fair so say by now that almost everybody at least recognizes that and that only a very few don't like it for the better). Again we of course think about the measurement proposal, which as of now has turned into something more easy because we can always work with the same cable but change its configuration. And Yes, I again have set myself to do it (but don't hold your breath because it takes serious time to do it for real).

 

Anyway, for me it is relatively easy to see how the configuration required for different PC's "works", so to speak, because I have obviously in my mind how the Lush^1 operates. Still at this moment I wouldn't be able to really make something out of that because it really would be too much guessing. Therefore the only reasonable hint should be about different radiation from different PC's (or configurations in there as per my previous post) - that being filtered by the cable in different means per different config.

Reasonable, right ?

Well, no. No, because to me it does not make sense for real that filtering the frequency or noise of a USB cable would change sound (all assumed bit perfect, as usual). Unless of course we dive into the now old Lush^1 thread and proposed working of that - then we allow ourselves to reason from there. Anyway, this is not what I propose because it is too difficult and should be proven somehow all the way first. But now the how of that ...

So for now after all and FWIW : different radiation patterns require different means of shielding. Ready for reality improvement ...

[PeterSt]

 

6 hours ago, rickca said:

How can Lush^2 be more flexible than the original Lush with all these extra layers?  How does the thickness (diameter) of Lush^2 compare to the original Lush?

 

I guess I can refer to "suspension" of some kind. Internally the cable is kept together better and it does not need to "knack" (??) to bend. The better expression would be that it bends more smoothly and that it also bends back by itself nicely.

 

Diameter of both is the same; you wouldn't see the difference. Both feel completely different though because more is packed inside (I propose to think of a kind of emulsie stuff feeling). What's packed more is more air than anything else.

I think this can give you an idea :

 

 

Connectors are the same as those of the original Lush.

 

[rickca]

42 minutes ago, PeterSt said:

it does not need to "knack" (??) to bend

I think the English word you're looking for is kink.

[rickca]

56 minutes ago, PeterSt said:

The better expression would be that it bends more smoothly and that it also bends back by itself nicely.

I got a 1.5m Lush^1 primarily because of my concern about the unusual way the original Lush cable bends at the ends.  I figured I might need a little extra length to compensate for that.  Does this mean that I can use a Lush^2 that is a bit shorter than a Lush^1 to connect the same two components?  By ' a bit shorter' I certainly don't mean 0.5m but maybe 6 inches?

 

I'm not worried about the price difference ... it isn't very much.  I just prefer not to have cables longer than necessary.

 

Here's how you described this behavior of the Lush^1 in a post on your forum.  Is it still true of the Lush^2?

At the "A" side of the cable (this is the PC side) the cable is made so that it can bend up and down - not sideways. At the "B" side (DAC side) it only bends side ways.
Both combined implies that you need a bit of length to go from the one device to the other.
 

[RickyV]

Hi @PeterSt,

First congratulations on your new findings, very interesting. What processor are you using for the 10/20?

So in reading about Luch^2 i am still confused by my imagination, i am a diy’er, the tree shields go on top of the “ more or less lush^1”? So the two data lines and the 5V are not separated by shielding? In my imagination i would put the 5v between the red and yellow shield.

BTW the lush cable was my first high end cable i purchased.

[PeterSt]

11 minutes ago, rickca said:

I got a 1.5m Lush^1 primarily because of my concern about the unusual way the original Lush cable bends at the ends.

 

11 minutes ago, rickca said:

Does this mean that I can use a Lush^2 that is a bit shorter than a Lush^1 to connect the same two components?

 

Rick, no, not at all. Almst the contrary but it depends a bit.

Try to envision that the behavior of the Lush^2 is more like the ^1 did it only for the ends. This is harmless because at any length this can be twisted sufficiently so the ends of the cable turn out to be in a position you can use. Still the culprit is the same, though worse for practice :

 

11 minutes ago, rickca said:

At the "A" side of the cable (this is the PC side) the cable is made so that it can bend up and down - not sideways. At the "B" side (DAC side) it only bends side ways.

 

We couldn't hold to this any more. This now has become "at random". It is all related to how internal twisting ends up for the various layers and how the connection wires need to connect.

The net result is that you can better take a bit longer if you think it is critical. Certainly not shorter.

 

Thank you for this very useful question !

 

Edit : I added this text to the webshop :

Warning on the 40cm length :

Please take into account that the Lush^2 is flexible in one direction only - not in 4 directions (so to speak). And as we can not tell how this turns out (which connector, A and B, bends easily to which direction, you should consider that 40cm can be too short for your practical usage. Thus, envision that the Lush^2 can bend to two opposite directions only, you'd need to twist the middle part of the cable for which you have a trajectory of say 30cm (1 ft), in order to direct the connectors at the end to where they need to go.

[rickca]

The guys who think cables don't make a difference may need medical attention after reading this thread.

[PeterSt]

14 minutes ago, RickyV said:

What processor are you using for the 10/20?

 

Ricky, Xeon 4114.

 

15 minutes ago, RickyV said:

So the two data lines and the 5V are not separated by shielding?

 

Haha, they were and they are.

 

16 minutes ago, RickyV said:

In my imagination i would put the 5v between the red and yellow shield.

 

That could have been done, but it is in between yellow and white (middle and inner). This is because else it violates your original Lush^1 and I certainly did not want to create a new cable with all over different behavior. So that's why.

 

Very good thinking !

[rickca]

Perhaps with Lush^2 it's time to revisit configurations including an ISO REGEN.  Both @lmitche and @Superdad like source=>USPCB=>ISO REGEN=>Lush^1=>DAC.  Maybe with an optimal Lush^2 connection configuration we can go back to putting the ISO REGEN just before the DAC as the design intended.

[RickyV]

@PeterSt are you at the x-fi audio show in Veldhoven this year would like to hear your new server and cable?

[PeterSt]

3 minutes ago, RickyV said:

@PeterSt are you at the x-fi audio show in Veldhoven this year would like to hear your new server and cable?

 

Since x-fi (or its environment) decided that no show reviews "need to" happen and we're only dependent on the visitors of the show itself, we decided it is not worth it and don't attend. So this year not either. Sorry !

I mean : the reviews (Internet) after the show ceased to exist.

 

Of course we could organize our own audio show with local CA members. At least we're then mutually guaranteed of the best sound. I mean, I will know it is fine and you will know bad sound (if so) will not be caused by the show room. Haha.

 

Fact is that it could be very interesting for the cable alone. Thus, showcase how tf it can happen that changing a bit of shielding vastly changes the sound. Nice gag for such a show. Could be done by anyone (but won't happen because next "they" like to have money for it and your Lush^2 doubles in price.

[PeterSt]

13 minutes ago, rickca said:

Maybe with an optimal Lush^2 connection configuration we can go back to putting the ISO REGEN just before the DAC as the design intended.

 

What is the status (quo ?) regarding this and what did the JSSG 360 tweak change ?

(I hardly followed - or remembered anything of this)

[rickca]

49 minutes ago, PeterSt said:

 

What is the status (quo ?) regarding this and what did the JSSG 360 tweak change ?

(I hardly followed - or remembered anything of this)

Here is a recent comment from Alex.

https://www.computeraudiophile.com/forums/topic/49599-a-proposal-the-objectivist-audio-review-magazine/?do=findComment&comment=862198

 

I'm not sure the JSSG 360 tweak of the original Lush cable changed the preferred ISO REGEN/Lush setup ... maybe @lmitche can comment.  However, the connection configuration flexibility of Lush^2 warrants a re-evaluation of the best way to combine them.

 

[austinpop]

6 hours ago, PeterSt said:

All,

 

 

The past weeks a countless number of times various suggested to create a separate topic/thread for the JSSG 360 Lush. Now with the Lush^2 out, this seems to be the time.

The Lush^2 applies the JSSG 360 tweak in a formal way by means of special connectors. Here you see 10 of of 10,000 or whatever countless possibilities, the 3rd diagram on the second row only showing the original JSSG 360 application :

 

 

The virtual problem is that no-one will be able to try out all the combinations with some sense. Maybe after a while we may reach some consensus like (it is only a virtual example !) the middle shield always needs to be connected to both ends for the best exhibit of the cable. Maybe people with good knowledge/experience are able to tell what foremost never should be done like never leave the source connector disconnected from any shield (virtual example again !), which will eliminate 100s of possibilities.

All I can tell at this moment is that the difference between one or an other is more vast than what I have read from many of you, unless you tend not to exaggerate - haha.
 

 

And that (above) can happen just the same because of what the (special) shielding achieves. And yes, with the Lush^2 I myself have already experienced the "flabby bass" up to super umphy disco bass (both not right at all) just because I applied the wrong shielding combination / config. It is easy to have it wrong but it feels like it is more easy to have it right / for the better by random attempts. But what is the best ? And how does it relate to the PC in use ? Can we ever reach consensus ? Will it be possible to find out the technical mode and merit ?

Just discuss it is also fine; I hope the result and findings will bring us a step closer again to the best sound ever. Where possible I will try to have answers to questions as well, but do notice that I do not carry any baggage and math to show you why a certain config would work out better than the other.

 

Thank you all (see below),

Peter

 

Because the path to the JSSG 360 appliance deserves credit to @JohnSwenson, @lmitche and @austinpop but just as well to the all so many people who ordered the original Lush the past year and the really not few who applied the JSSG 360 tweak so we could get a sense of it really working - the Lush^2 is dedicated to all these people. 

 

 

Very interesting development, Peter!

 

Thank you for the shoutout, but it's all really @JohnSwenson's and @lmitche's ideas - I just screwed up the courage to try a DIY project. It should be clear, but let me clarify anyway, that I have no financial or business ties with Peter - in fact I just heard of the Lush^2 last night!

 

3 hours ago, PeterSt said:

 

I think this can give you an idea :

 

 

Connectors are the same as those of the original Lush.

 

 

Peter, could you clarify the switches on the ends. How do you configure them, in relation to the diagram in your first post?

[lmitche]

33 minutes ago, rickca said:

Here is a recent comment from Alex.

https://www.computeraudiophile.com/forums/topic/49599-a-proposal-the-objectivist-audio-review-magazine/?do=findComment&comment=862198

 

I'm not sure the JSSG 360 tweak of the original Lush cable changed the preferred ISO REGEN/Lush setup ... maybe @lmitche can comment.  However, the connection configuration flexibility of Lush^2 warrants a re-evaluation of the best way to combine them.

 

Hi Rick,

 

No, I haven't reevaluated the iso regen position with the jssg360 lush innovation. My power cable for the iso regen is too short.

 

Larry

[PeterSt]

52 minutes ago, austinpop said:

Peter, could you clarify the switches on the ends. How do you configure them, in relation to the diagram in your first post?

 

Rajiv, everybody receives a paper in the package with (sufficient) examples on how to make the connections per desire. It really is easy with that paper in front of you.

I chose not to publish that paper because it may be suggestive for by me trialled connections, which is not the case at all.

 

 

I suppose you have seen this by now, but if not this is to wet your appetite. It is also the communications means for people putting forward their findings. It is as easy as :

I connected black to white and yellow to red for both A and B (and this comprises the genuine JSSG 360 config).

 

One thing is important only : really everything can be connected (or disconnected) to(/from) anything else and that per A and B connector. And this gives more possibilities than you'd imagine, because it really is "everything". There are no exceptions or limits.

 

I still think you deserve to be in the row of originators, because you hosted that thread and were/are a huge stimulator. So thank you for that.

Peter

[PeterSt]

Yesterday, for a first time (after 6 days or so) I changed my own config. Just because I knew I could change things it came to me that I wanted the highs a bit less pronounced (they were not harsh or anything, but could be overwhelming).

I changed from the normal JSSG 360 config (which is black to white and yellow to red at both ends) to a disconnected yellow and red at the B side. Thus :

 

- Normal (inner) shield is active from A connector to B connector;

- Middle and outer shield connect at the A connector;

- Middle and outer shield do not connect at the B connector.

 

The first thing that I noticed was a more liquid bass. I'd call it even musical. This also doesn't make it a placebo appliance because I wasn't expecting that.

Later into the session it became more and more clear that the behavior of the highs exhibited the same liquid. For my situation it will mean that less of the "electric butterflies" will fly with certain music, and although that is spooky in itself, at this moment I think I don't care about that really.

The liquid bass (say "singing") is what I kept noticing throughout (2 hours or so of time).

 

As always, this does not imply it will stick, because maybe I will start to see annoyances later. Point is, that I have these kind of contributions in my mind for this thread, so later we can use it as a reference or plain guidance in general.

To be useful, this of course this needs confirmation by others, with at this moment no idea whether anything will work out the same for people.

 

Peter

[rickca]

1 hour ago, PeterSt said:

Normal (inner) shield is active from A connector to B connector;

- Middle and outer shield connect at the A connector;

- Middle and outer shield do not connect at the B connector.

What do you think happens to the signal when a shield is connected at A but not connected at B?  Would connecting a shield to B but not A be equivalent?  Was this configuration just a shot in the dark, or did you have some expectation about how it might affect the sound (an educated guess)?

[PeterSt]

Just now, rickca said:

What do you think happens to the signal when a shield is connected at A but not connected at B?  Would connecting a shield to B but not A be equivalent?

 

Rick, no, that will not be the same.

The sheer reason to provide such options spring from the time (2010/2011) that I had to decide whether the shield of USB needs to be connected to the DAC's chassis, some other grounding point, or not at all. I never look back at this any more (we connect it to "some" grounding point) but at least back then, there were no consistent answers to be found. Many solutions were provided, but all with a different reason and I actually could not decide.

 

What I recall of it, no options existed to leave the PC-side (or A connector) disconnected. This is logic, because a normal USB cable just would have the shield connected (both ends btw) so the options was just not there (and still is not, normally).

So what that was about was for the DAC designer the decision how to configure that, and obviously I have not been the only one to decide that (all DAC designers with USB input need to decide this, unless there's suddenly an enforcing law about this ).

 

Because it has been in my mind back at the time, for me it was the logical first next option to try. Well, almost, because this is only about the middle and outer shield, but you can bet that the next config I am going to attempt is leaving out the inner shield as well, and with that thus all on the B side.

 

On a side note, I already tried to have both ends fully disconnected, but although with the other processor, that sounded like a drag and didn't last for 10 minutes (muffled highs and nothing of interest).

I suppose (but should measure it) that not using the shield at all is bad to USB noise (the 8K multiples). This would be logic because it is an inherent USB frequency (tamed by schielding ?), plus I recall the sound from my first USB attempt, with poor results. What was wrong with it was the high USB noise (I could measure). Btw, most people would fall in the pitfall of such a setup (or any with higher USB noise) because it excels in the bass. But no highs and thus no interest (dead bird music).

 

I suppose I can come up with more "logical" trials and eliminations. For me they must spring from listening because many types of sound will remember me of some kind of set up ever back and how I solved that (issue assumed) or let it stay. And, of course, I must first have more experience on what causes which.

 

Imagine - assumed that the one sided shield connection like this indeed does a few things - what will happen with one shield connected to B only (I would take the middle for that) and the other connected to A only (the outer). Flow assumed (ending up in some kind of radiation), we'd now have two flows in the opposite direction. Will they cancel perfectly ? Net effect ?

What if the 3rd shield is normally connected end to end ? And if that is for the better, shouldn't it be better the middle one ? or would that eliminate the canceling effect and shouldn't it be better the inner one after all ?

Already the thoughts can go anywhere. 

[lmitche]

One of the practices of Naim Audio is to create a star grounding scheme between various components despite use of multiple outboard power supplies. In order the achieve this scheme, proprietary interconnect cables are used with multiconductor din connectors. These cables are also directional with screening connected to ground at only one end of the cable. Therefore when assembling a system black bands on the cable are used to mark the end that should be connected to the signal source.

 

Knowing this, I was surprised when John Swenson suggested that both ends of the screen meet at the endpoints with no connection to ground. Nevertheless John's scheme certainly works well here.

[PeterSt]

57 minutes ago, lmitche said:

both ends of the screen meet at the endpoints with no connection to ground.

 

It could be that someone misinterpreted. Ground is not equal to the screen / shield. Of course for some cables it is (like coax) but not here.

I can't quote readily, but of what I recall is that the both additional screens meet each other, and not the original inner screen. 

... which, thinking of it, would be a bit weird if that were to be about a wire only, because John talks about wires, not "360" screens.

So you must be talking about an other text.

 

It is more complicated, because depending on the in-device(s) design, it is very easy to let end up the shield at a same point as the ground does. So only when that can remain fully separated, it is useful to nog connect the shield to the ground of the cable to begin with (now the designer of the device can decide).

And FYI : where the Lush(^2) does not connect the shield to the ground (wire) in any event or config, the difference shows in our Phasure NOS1a/G3 DAC; the shielding *has* to follow a special route because otherwise the isolation (same as in the ISO Regen) does not work or only works with errors. So that fragile the end to end connection of a shield is.

For that matter it was almost surprising to see that all of the Lush^2 configs we randomly tried, have no USB errors anywhere. Thus even with all shields disconnected. This should at least tell that the way the shielding has been set up (with thus three screens) already should work for the (way ?) better, because I can't imagine USB to work error-free with a normally one-screen shielded cable but that screen not connected at both ends. I must say, I never tried it either. Yes, thus with the Lush^2 now.

[fas42]

Just a quick note to say that I went through a "mad" period of trying to shield everything, on a totally conventional CDP -> amp -> speakers setup, and experimented with every combination of connecting the shields to anything else - and got the same sort of results as Peter did, . A sort of insanity overcomes one, - thank goodness, I've calmed down a bit now  ... there will an optimum way of doing it, but it may take quite a bit of investigation to work that out.