Devialet Chat

I am curious about the manner in which Devialet has chosen to link the master/slave units.

1. Since they are linked via a digital connection, wouldn't the sound of the slave be dependent on the quality/flavour of the digital interconnect used?

2. Plus, Devialet has chosen to use RCAs instead of BNCs so there may be impedance mismatches?

3. For dual mono DACs, there is usually a clock sync connection between them. In this case, unless there is a clock sync function implemented by Devialet, it would appear that the slave has to extract the clock from the signal sent by the master, leading to potential timing issues?

4. With AIR, Devialet could have chosen allow data to be streamed to both master/slave units either wirelessly or wired. If wired, then if the same type of ethernet cable were used for both units, the "flavour" would be the same. This does not solve the clock sync issues though.

Hello nfnc,

The buss-cable are included when purchasing a second unit (slave unit) and it is a true 75 Ohms cable. Probably a different cable could change the sound but I would go for the one supplied since that cable is proved by Devialet.

I would guess that a time adjustment between the two amplifiers are necessary since the signal first arrive to the head unit and then has to be transfered to the slave unit. Most likely this must create a time delay if not the signal is delayed in the master unit. But the Devialet brain trust has its origin in the telecom business (french company Nortel) so I presume this is not an issue. Synchronization between digital telephone stations have been performed for 25 years by now :-).

Best regards

BjornWiman

(12-Jul-2014, 23:02)BjornWiman Wrote: [ -> ]But the Devialet brain trust has its origin in the telecom business (french company Nortel) so I presume this is not an issue. Synchronization between digital telephone stations have been performed for 25 years by now :-).

Best regards

BjornWiman

Totally off topic and pedantic, but Nortel is/was a Canadian company which bought the French company Matra telecom.
I think Nortel no longer exists.

Hi,

(12-Jul-2014, 20:10)nfnc Wrote: [ -> ]I am curious about the manner in which Devialet has chosen to link the master/slave units.
Me too, as I have bombarded Dev with questions concerning this as initially there was a solution of an AES/EBU to RCA link that Dev seem to replaced with the present Crystal Cable RCA-RCA. I never got an good answer from Dev with regards to why not use the AES/EBU to AES/EBU instead and why they had such odd solution from the beginning. I also had issues with the non BNC connecting possibility but then the answer from Dev was that I could use an BNC to RCA converter obviously missing my point with the question. In that same answer I had from Dev they say it will be 'impossible' to have true 75 ohm however they say it is not of importance indicating that Dev have some gizmo inside for impedance correction, however they did not confirm that either.

1. Since they are linked via a digital connection, wouldn't the sound of the slave be dependent on the quality/flavour of the digital interconnect used?
My point too! I can still hear differences between the units, and I tried swopping the channels with the result of that the slave still sounding a bit less than the master. I still relate this to 'burn in' as my 200 took some 500 hours to settle. But if it will not change I will for sure 'blame' the digi connection or the slave it self. I will for sure play around with cable options to rule out if cable gives a 'flavor' or it is the technology in the units that limits things.

2. Plus, Devialet has chosen to use RCAs instead of BNCs so there may be impedance mismatches?
See my answer above!

3. For dual mono DACs, there is usually a clock sync connection between them. In this case, unless there is a clock sync function implemented by Devialet, it would appear that the slave has to extract the clock from the signal sent by the master, leading to potential timing issues?
It would be very interesting to know more about this from Devialet! Typically Devialet answer very correct and technically good BUT when something is not under their control or they are aware about some less good issues they usually answer in an avoided way saying it doesn't matter for sound quality.
The reason for this answer I start to believe that the see their products as lifestyle products and not as HiFi/Hi End units which is not so smart that most of us owners seem to be hunters for optimal fidelity hence we want to optimize and understand.

4. With AIR, Devialet could have chosen allow data to be streamed to both master/slave units either wirelessly or wired. If wired, then if the same type of ethernet cable were used for both units, the "flavour" would be the same. This does not solve the clock sync issues though.

/Mike

(14-Jul-2014, 12:21)Mikeeo Wrote: [ -> ]

(12-Jul-2014, 20:10)nfnc Wrote: [ -> ]I am curious about the manner in which Devialet has chosen to link the master/slave units.

3. For dual mono DACs, there is usually a clock sync connection between them. In this case, unless there is a clock sync function implemented by Devialet, it would appear that the slave has to extract the clock from the signal sent by the master, leading to potential timing issues?
It would be very interesting to know more about this from Devialet! Typically Devialet answer very correct and technically good BUT when something is not under their control or they are aware about some less good issues they usually answer in an avoided way saying it doesn't matter for sound quality.
The reason for this answer I start to believe that the see their products as lifestyle products and not as HiFi/Hi End units which is not so smart that most of us owners seem to be hunters for optimal fidelity hence we want to optimize and understand.

Well digital data typically travels through cable at 2/3rds the speed of light so for a 1/2m cable that is 1/600 millionth of a second which is a mismatch I can live with, I guess.

(14-Jul-2014, 12:21)Mikeeo Wrote: [ -> ]Typically Devialet answer very correct and technically

Unfortunately this is not my experience. Even though I do have respect for the work they are doing, I feel like sometimes they try to avoid answering the question altogether or answer with half truths. It's a nice thing that they reply though.

(16-Jul-2014, 09:24)jjo Wrote: [ -> ]

(14-Jul-2014, 12:21)Mikeeo Wrote: [ -> ]Typically Devialet answer very correct and technically

Unfortunately this is not my experience. Even though I do have respect for the work they are doing, I feel like sometimes they try to avoid answering the question altogether or answer with half truths. It's a nice thing that they reply though.

jjo, see the rest of that sentence: BUT when something is not under their control or they are aware about some less good issues they usually answer in an avoided way saying it doesn't matter for sound quality

I do feel that Dev has changed or perhaps they are on their knees in tech questions they didn't expect due to that they are now 'life style products' manufacturer I do think that the trend has moved into the avoiding answering area.

/Mike

At the Oxford Audio Consultants' Devialet day today and Mathieu (their chief engineer) was there along with another guy from Devialet plus Pedro from Absolute Sounds.

Apparently the Slave is now to be called the Companion (to satisfy USA sensitivities).

"Significant Other" would be in the line if they wanted to be a little bit more sensitive

Well, posed the queries to DVLA last week but no response yet.

On the clock sync issue, while one assumes that DVLA have done their homework, it would be good to know what they have done. As some have said, I preferred their old, info-centric, website.

There are other manufacturers who are more upfront with what they have done to address the clock sync/time lag issue. For example, TotalDac employs a dedicated clock sync cable between their dual Mono DACs even while it connects the digital signal via AES/EBU. Avantgarde Acoustics, who link the master/salve units of their Zero 1 DSP active speakers wirelessly, indicate that they delay the master unit by up to 2 milliseconds to allow the slave unit to catch up with the master unit.

Mikeeo reports hearing differences in sound quality between the master and slave units in the 400 while IanG-UK is happy with his 800. Hmm...just go for the top of the line