I had the white noise issue this afternoon. After playing around 5 hours. To me it sounds like the virtual sound card lost sync with the amp. I've had this once before, but I had problems with stability on MC 19 and 2012 R2 at the time so I did nor report it. Since then I have sorted the stability issue (driver problems in 2012 R2). I've mostly been using USB to avoid the white noise trouble, waiting for Devialet to fix it.
MC 19 | Server 2012 R2 | AIR Ethernet | D250. Wifi card disabled, DHCP on ethernet
(10-Aug-2014, 23:25)ogs Wrote: [ -> ]To me it sounds like the virtual sound card lost sync with the amp.
@
ogs, what do you mean by that?
Also, do you recall pausing tracks during your listening session, especially, pausing for lengthy period of times?
(29-Jul-2014, 19:51)AlexS Wrote: [ -> ]@Georgethesixth, I thought Devialet was using unicasting for streaming. AIR does know the exact IP address of the Devialet (and shows it, when it's connected). So, why would AIR need to broadcast messages? Well, that's a topic for a different thread (perhaps, one titled "Know your Devialet" 
Based on what I can see by tracing network traffic between my Mac and Devialet 200, it looks as though AIR streams audio data via unicast UDP (host port 45456 -> Dev port 45455). Each UDP frame seems to carry a 962-byte payload when there's music playing at 16-bit/44.1 kHz. The payload size and/or frame rate will no doubt change for different sample sizes and/or rates, but I haven't tried anything other than 16/44.
The broadcast mechanism seems to run continuously whether the Devialet is in standby, powered up, or streaming. I assume this is just used for discovery. As @Georgthesixth found, the Devialet seems to broadcast to port 45454 several times per second from each interface (Ethernet and/or wi-fi).
(29-Jul-2014, 19:51)AlexS Wrote: [ -> ]I wonder what useful info you'd be able to gather with a network tap, even if you catch it at white noise act. How would you know if AIR sends "noise" packets to the Devialet, or if the noise originates in the Devialet itself (needless to say, I suspect the former). Perhaps, the packets would contain only 1's (no 0's) - that's what white noise is, is it not?
With a network tracing tool you could see the payload of the UDP packets which carry the audio data. Whether that would help or not depends on how the data is formatted. I'm guessing that it might be raw, linear PCM but I haven't any good evidence for that other than that the data rate is roughly right. In any case it would probably be quite difficult to distinguish music from noise by looking at the payload. (it could be easier if you were playing a test tone.) I'd expect white noise to be represented by a random-looking payload, not all-ones.
Speculating wildly...a burst of white noise might be caused if the AIR software sent data from a "random" region of memory on the host to the Devialet. Depending on how the AIR software is implemented this could occur due to memory management problems, which I think you've already suggested as a possible cause.
Not sure whether this helps or not…I hope so.
My tap has not arrived yet so I'm 'network blind'. The tap is needed because even when you use a SPAN port on a Cisco switch, you'll always miss some traffic.
What I find strange is that people are willing to pay serious money for an audio USB-cable, while Devialet is 'polluting' their network with a permanent stream of broadcasts. I had to reconfigure the switch to allow this number of broadcasts from one host. The switch interpreted the amplifier/DAC as the source of a denial of service attack.
It looks like Devialet wants to avoid the complexity of having to enter the amplifier/DAC IP-address but the way they solved this is hardly elegant.
To be clear : we bought the Devialet for the sound quality and not for the way they use TCP/IP.
(12-Aug-2014, 00:01)Georgethesixth Wrote: [ -> ]It looks like Devialet wants to avoid the complexity of having to enter the amplifier/DAC IP-address but the way they solved this is hardly elegant.
Indeed. There are any number of devices that solve essentially the same problem without recourse to continuous broadcasts (Sonos, just to pick an obvious example of one that "just works").
Apart from the network pollution, this can't be good for standby power consumption.
To be frank, I think Devialet have been a bit lazy and/or rushed with this.
(10-Aug-2014, 23:20)Jwg1749 Wrote: [ -> ]Thanks for the update. good to know they're looking into it.
Matt
Devialet have a Windows 8 laptop of ours which has the 'noise problem' and with which they can reliably reproduce the problem. We are waiting to hear from them about the results but of course they are on summer shutdown at present!
(12-Aug-2014, 22:31)thumb5 Wrote: [ -> ] (12-Aug-2014, 00:01)Georgethesixth Wrote: [ -> ]It looks like Devialet wants to avoid the complexity of having to enter the amplifier/DAC IP-address but the way they solved this is hardly elegant.
Indeed. There are any number of devices that solve essentially the same problem without recourse to continuous broadcasts (Sonos, just to pick an obvious example of one that "just works").
Apart from the network pollution, this can't be good for standby power consumption.
To be frank, I think Devialet have been a bit lazy and/or rushed with this.
I agree. I contacted Devialet about this a couple of months ago and also posted in the Support Desk section on here about it ( see thread "Unexpected activity on router"). There was a bug in the 'disable remote control app' function as well.
It's clearly not a major problem and not high priority for them to change/fix this but if enough people complain then maybe the priority will increase.
Regards,
Philip
(13-Aug-2014, 12:56)oxfordaudio Wrote: [ -> ]Devialet have a Windows 8 laptop of ours which has the 'noise problem' and with which they can reliably reproduce the problem. We are waiting to hear from them about the results but of course they are on summer shutdown at present!
Now this is indeed good news. And no more excuse for Devialet not to fix this once and for all.
@
oxfordaudio, what was the "white noise" pattern exhibited by this this Windows 8 laptop? Was this just lengthy listening sessions, which eventually resulted in the issue? Or something else?
(11-Aug-2014, 15:03)AlexS Wrote: [ -> ] (10-Aug-2014, 23:25)ogs Wrote: [ -> ]To me it sounds like the virtual sound card lost sync with the amp.
@ogs, what do you mean by that?
Also, do you recall pausing tracks during your listening session, especially, pausing for lengthy period of times?
Hi AlexS, sorry for the late response.
When I heard the white noise it reminded me of the noise I get when clock sync over an S/PDIF link to a DAC is lost or rather when the input circuit on the DAC can not lock to the sample rate of incoming data. AIR is designed to be 'asynchronous' i.e. the clock in the amp will control the virtual sound card running on the PC/Mac. If this sync is lost I believe noise is the result unless the mute circuit kicks in. I may of course be wrong here..
I do not recall pausing playback - not in the last hour or two before the noise anyway.
(12-Aug-2014, 00:01)Georgethesixth Wrote: [ -> ]It looks like Devialet wants to avoid the complexity of having to enter the amplifier/DAC IP-address but the way they solved this is hardly elegant.
To be clear : we bought the Devialet for the sound quality and not for the way they use TCP/IP.
I completely agree with this. Two of the main designers at Devialet has background from Nortel R&D and should have lots of network knowledge.
(12-Aug-2014, 22:31)thumb5 Wrote: [ -> ]Indeed. There are any number of devices that solve essentially the same problem without recourse to continuous broadcasts (Sonos, just to pick an obvious example of one that "just works").
Another example: slimproto, streaming from slimserver (LMS) to squeezeboxes. Rock solid
(13-Aug-2014, 20:47)AlexS Wrote: [ -> ] (13-Aug-2014, 12:56)oxfordaudio Wrote: [ -> ]Devialet have a Windows 8 laptop of ours which has the 'noise problem' and with which they can reliably reproduce the problem. We are waiting to hear from them about the results but of course they are on summer shutdown at present!
Now this is indeed good news. And no more excuse for Devialet not to fix this once and for all.
@oxfordaudio, what was the "white noise" pattern exhibited by this this Windows 8 laptop? Was this just lengthy listening sessions, which eventually resulted in the issue? Or something else?
yes it was - usually after 30-60 minutes playing consistently; often just loud white noise, sometimes distortion over the music.