last mile ethernet

[Kunter]

OK ! It's for me ! :-)

Some audiophilefr members and Jean-François Bacquet, the designer of Klinger Favre (french loudspeakers and amps brand) organized some listening sessions :

Totaldac Server + Devialet 800 + Klinger Favre Studio 30
vs
Totaldac Server + Totaldac Dual Dac + Orpheus Lab 3M monoblocs + Klinger Favre Studio 30

In fact, the Devialet 800 had been replaced by Devialet 250, as they didn't find a good link between master and companion…
After this sessions, Aerki (the Devialet 800 owner) did some tests between coaxial links, and found that the Crystal cable delivered with the companion was the worst of the three cables tested… and the SQ-Audio Pharos was the best one as it has a true 75 ohms connection… and costs 3 times less than the Crystal one ! And some other tests will follow !!!

About the Devialet / Totaldac comparison, here is the result :
- the Devialet 250 is a real good sounding solution, especially with the Totaldac Server
- but all the listeners agreed to say that the Totaldac Server + Dual Dac is better on all aspects (bass response, resolution, sound staging, dynamic…).

Even Aerki told that in order to improve his system, he should buy the Dual Dac… But it's 9900 Euros more !

What could be interesting is to have the feedback of Aerki when he'll find the good link between master / companion… I think that 800 could bring some improvement as right and left signal have their own DAC circuit...

Thank you for doing this Steph! Actually it would be great if Aerki could join us to discuss the digital link and of course the totaldac solutions!

Guillaume

Thank you Steph!

Do they try and compare

Totaldac Server -> Totaldac DAC -> Devialet Analog inputs
vs
Totaldac Server -> Devialet Digital inputs

Kunter

[nicoludio]

Thank you Steph!

Do they try and compare

Totaldac Server -> Totaldac DAC -> Devialet Analog inputs
vs
Totaldac Server -> Devialet Digital inputs

Kunter

This appears to be the general consensus (using Google translate):

"Between the System 1 and System 2, the match was short, it is rare to throw in the towel before the end of the first round By doing a little less concise qu'Aerki, and not not too frustrating subscribers to this thread, the gap between Systems 1 and 2 were: - shortened bass - bass missing Stamp - treble shortened - a lack of texture across the spectrum - a lack General precision - a flattening of sound levels - a lack of ventilation between instruments careful, no one said the system was bad !! System 2 was just a lot better, what we can not realize that putting them side by side (I know, I ramble). One of them is he the best in the world, of course not. ... That makes no sense. Speakers and speaker cables were the same. System 1 has also benefited from the small upgrade an excellent digital cable Accrolink used on the system 2."

System 1 was Totaldac Server + Devialet 800 + Klinger Favre Studio 30

System 2 was Totaldac Server + Totaldac Dual Dac + Orpheus Lab 3M monoblocs + Klinger Favre Studio 30.

---

Here are some comments from Aerki:

"Whatever the solution chosen intermediate amplification, a return of a rare level is obtained. Presence is dense, all in relief and sonic nuances. Intuitively, I would give the following weight to the responsibility of the various elements in the final result. Source 40, Amplification 20 Speakers 40 In my view, we compared the D250 Orpheus and the Unison blocks. 's D800 was out of the race, in fact the night before (With Nicoben, Cubase and PierreRay) it was clearly understood that the connecting cable / sync between the two amps played a very important role and was - to date (qq amp received days before) - not so much. Moreover, the same phenomenon has been demonstrated, and dramatically, changing the digital cable between D1 Server and amp (or D1 Server and dualdac). were two renditions of a similar nature between D250 and Orpheus, especially in the presence of Unison (tubes). In this configuration, two blocks Orpheus deliver a return slightly better than the D250. 's unison course plays in the warm register. Flatterers from the first moment. Mediums for me, spent the first moment I do not like the presentation of the music I would call overly "romantic". The colors in a photo taken in sepia tones, I prefer the density and contrast of Black & White, unaltered. Due to time constraints, there were still 450 km to go to find the island of France, we did not compare to the same config: two blocks 250 / Orpheus two blocks. Allow me to explain: using 2 D250, not by combining and synchronizing in D800 (source then arriving on one of the two amps and Signal sent to the other), but making them work in two independent mono blocks. Thus it overcomes the barrier of sync (and the famous cable digital connection that can all "ruin"). Through what I've heard, the results should be very close (very high end) and I certainly would hazard me not to pre-enter on the nuances that might intervene. So I'm going to change my config. Two blocks mono-powered D1server / dual dac. Each channel (speaker) will have its own line of conversion / amplification. Thierry can eventually we quickly re-explain the principle of DUAL dac and conversion through. One may also conclude from these tests and my approach that will then Devialet more employees as part of what they do, amplification. D1 Server After a few months there, the major contribution to my system is that Dual Dac. We enter here without any ambiguity in a new dimension of digital / analog conversion. Finally a few notes, so Unison, and I think for them shared in unison, notes borrowed from Albert Camus and conveyed Vincent Brient "Hats off, sir,""

What he appears to be saying is that he has plans to drive the Devialet 800 with TotalDac Server and two mono Dual DAC (one each for L/R channels). So, Server -> DAC -> D800 analog inputs seems to be better than Server -> D800 digital inputs.

Reading through the rest of the thread, another Devialet 800 owner felt that the conditions of the shoot-out was not optimal and that, amongst others, the AIR input should have been used. So, there will be another shoot-out between the Devialet 800 vs TotalDac Server/Dual DAC, scheduled for 28 August.

Apologies to Kunter for side-tracking...

[darthsteph]

In fact, if he had the money, Aerki will use the Devialets not as a 800 but as 2 x 250 in monobloc mode.

It means :

Totaldac Server -> one Totaldac Dual Dac -> analog inputs of 2 x 250 monoblocs

In this case, the Devialets will be used only as amps...

[nicoludio]

Thanks for the clarification.

[Antoine]

Another two happy users of the Diamond!

http://devialetchat.com/Thread-What-is-t...04#pid4004
And my own experience: http://devialetchat.com/Thread-Antoine-s...42#pid3142

[ogs]

Indeed.

Bit OT but it would have been interesting to have had the option to use the TotalDac discrete R2R ladder DAC (or even BB PCM 1704) instead of the BB PCM1792 in the Devialet.

I have been thinking this too! If the DAC in the Devialets was replaced with an R2R design...! And Devialet should introduce a reclocker in their amps. Being as transparent as the Devialet amps are, jitter is a very bad thing.

---

In fact, the Devialet 800 had been replaced by Devialet 250, as they didn't find a good link between master and companion…

For this reason I have dropped thinking of adding another 250. The link between master and companion should not be SPDIF or AES, but something like the MSB Pro I2S.

[Confused]

I must admit to being totally baffled by the above posts. There are many good technical reasons why an 800 should sound better than a 250. Devialet themselves state that the 400 sounds better than a 250, a case of the dual mono configuration halving distortion. The 800 is another step up. I have also noticed fairly consistant options from a few people that the 800 offers less "grain", a higher resolution sound. (On top of the raw power and headroom, which is good news for speaker control) What the above suggests is that the crystal digital cable is so bad it negates all of this. The self same cable that Devialet give you in the box with a new 250 slave.

Now I am running a D800 with the crystal, am I insane?? Has anyone tried switching the crystal on an 800 with a genuinely audible improvement? I must admit that I don't think I could tell the difference between the Crystal and another digital cable, having said that, I must admit I've never tried a back to back test. I do have other digital SPDIF's I could try, but these are cables that are about a tenth of the retail price versus the Crystal, but I guess I should give even these "cheap" cables a go when I get some time?

[GuillaumeB]

I must admit to being totally baffled by the above posts. There are many good technical reasons why an 800 should sound better than a 250. Devialet themselves state that the 400 sounds better than a 250, a case of the dual mono configuration halving distortion. The 800 is another step up. I have also noticed fairly consistant options from a few people that the 800 offers less "grain", a higher resolution sound. (On top of the raw power and headroom, which is good news for speaker control) What the above suggests is that the crystal digital cable is so bad it negates all of this. The self same cable that Devialet give you in the box with a new 250 slave.

Now I am running a D800 with the crystal, am I insane?? Has anyone tried switching the crystal on an 800 with a genuinely audible improvement? I must admit that I don't think I could tell the difference between the Crystal and another digital cable, having said that, I must admit I've never tried a back to back test. I do have other digital SPDIF's I could try, but these are cables that are about a tenth of the retail price versus the Crystal, but I guess I should give even these "cheap" cables a go when I get some time?

Hey Confused I didn't get a Crystal cable with my 500, at the time they were shipping with unbadged Transparent Premium ones. As an experiment I borrowed a Transparent XL S/PDIF from my dealer... just to see. It was just supposed to be an experiment, I can assure you. The results utterly amazed me and made very little sense (how does this sync business work anyway?!). I also discussed this with Absolute Sounds and they agreed with me.

Now based on what I'm reading elsewhere it would appear that others are getting really good results by changing S/PDIF cables. Some are getting similar results with 300 euro cables (which I posted about in another thread).
I haven't heard the Crystal one so can't comment but I've heard of others saying that they aren't that impressed. There are also concerns being expressed about the technology used to link the two units. Could that make it more susceptible to jitter? Who knows. They did recently introduce a jitter "fix" though. Did it sound good before that fix? Well, yes!

As with all these things it depends on how far you want to take things. Some are quite happy with what they have and others feel the need to keep experimenting and pushing the envelope (you know all the bonkers stuff! ). Just because something sounds way better after a tweak doesn't mean the original setup was "broken".

Yes the Dual mono was a big step up from the single unit. But so was the TA XL cable I installed... as were some of the other tweaks. I wouldn't bother unless I got decent results each time.

Guillaume

[IanG-UK]

Cable influence, especially in the digital arena, remains a contentious issue even within the relative calmness of the DevialetChat forum.

What I continue to find unbelievable is the justification for charging astronomic prices at the "top end". And the multipliers which seem to apply.

For example, with many of us using 5m lengths for router to Devialet connection, the OP's Audioquest Diamond costs around £3,000 which is 55 times the price of Audioquest's base Pearl offering at £55, that in turn 10 times the price of stock CAT6 cable.

By comparison, looking at "real" hardware, the Devialet 800 at £21,000 is around 20 times the price of a base offering for pre/power/phono/dac/wifi from Arcam, being the FMJ A19 and the Airdac. I'm content that each of those probably retail at around ten times the component and manufacturing cost, but in the case of the Diamond I bet that multiplier is one hundred or more.

At the forthcoming Devialet Guildford event it would be interesting to get the Devialet guys to comment on ethernet properties from a technical viewpoint and get the Guildford guys to run some blind tests to see if differences are readily discerned.

[Antoine]

I'm quite sure the differences with the AQ Diamond ethernet are real, I don't know why but it does make quite an improvement compared to even the runner up Vodka.

Luckily the 0.75m version is long enough in my situation as I fully agree the prices are astronomical. It's the usual 80/20 (or 95/5 if you'd like) rule that applies here. To get the last few percent of attainable performance you have to pay an uneven, most of the times 'nonlinear' amount of money.