Ethernet or wifi streaming and white noise on Mac

[IanG-UK]

I've had the opportunity to deal direct with the Devialet engineer who writes the coding for AIR.

My issue, for months, has been noise using AIR, more profound on hi-res (slightly quicker to happen on wifi than ethernet) but still an issue (which happens much more slowly) using 16/44 (again slightly quicker to happen on wifi than ethernet).

Today the engineer took control of my MacBook Pro whilst the noise was present, using ethernet and a buffer size of 5000ms (as he recommended). The noise, by the way, is continuous persistent disturbance, not drop outs or crackles. Apparently, the latter is usually associated with an inadequate wifi infrastructure whereas the former is a timing error associated with the pace at which the audio samples are extracted.

This is, apparently, similar to the issue with Windows, but cannot be eliminated using the same solution.

So they now know where the noise stems from, and need to find a way of solving it. I believe it is due to a one byte offset in the buffer manager - but one which is not solved by having a big buffer size.

[smokejumper]

Thanks for the update, it is nice to finally know what is going on. Beta 4 did not cure it for me, but it was better. I am a Windows user.

[Damon]

IanG,
That is an excellent update; getting the Devialet engineering staff involved while the problem is occurring is tremendously valuable. So many issues are difficult to diagnose by description, and electrical glitches are especially difficult. If Devialet can hear the problem, and 'see' your system at the same time, they will have better success solving it.

Thanks for that effort and the report.

[Womaz]

Great news Ian. The new beta did not work for me either, but when I move my laptop into the same room as my router , well so far I have had no white noise. It has been 2 weeks now with none. As soon as I move the laptop back into my work area, the white noise kicks back in within an hour or so.

Hopefully they now have more to go on and may be able to finally solve this.

[IanG-UK]

Great news Ian. The new beta did not work for me either, but when I move my laptop into the same room as my router , well so far I have had no white noise. It has been 2 weeks now with none. As soon as I move the laptop back into my work area, the white noise kicks back in within an hour or so.

Hopefully they now have more to go on and may be able to finally solve this.

Thanks. However, the fix for Windows, as already issued as a beta, will not work for Mac. The Windows fix is a workaround because the fault is one embedded in Windows: it is not an AIR fault per se. The fix involves using the best clocking device accessible on Windows - and it appears to improve things but not fix things entirely.

On Mac, AIR already uses the best clocking device in the Mac OS so they will have to devise another approach to solve the Mac issue.

Edit: Of course, this highlights a point often made by Guillaume. Computer audio might be inherently a botch because it is not dedicated; but more significant in my view is that Devialet will inevitably be subject to possible interferences as Microsoft or Apple make OS modifications with no regard as to whether Devialet are happy with them. So a white noise "fix" this year might need another in 2016 and another in 2017 and ...

[AlexS]

Great news Ian. The new beta did not work for me either, but when I move my laptop into the same room as my router , well so far I have had no white noise. It has been 2 weeks now with none. As soon as I move the laptop back into my work area, the white noise kicks back in within an hour or so.

Hopefully they now have more to go on and may be able to finally solve this.

@Womaz, this conversation probably does not belong to this thread, but rather the one specific to "WN on Windows", but I'll reply to your post anyway. I've just realized that you're not using an Ethernet connection, but WiFI. The fact that your problem is cured by the laptop's proximity to the router tells me that it's WiFi/signal strength dependent, and probably has nothing to do with AIR proper. I also suspect that the "white noise" you're getting is (somewhat) different from the variation most of us are afflicted with. Or else, there is a possiblity that with a poor WiFi connection, AIR has to work harder (there being more errors and retries to deliver the lost packets), and this makes it more vulnerable to the timing issues Ian had described to us. Could this (heavy traffic from AIR) also be the reason why the 24/192 format is more likely to expose the problem?
Anyhow, it's essential that you share your observations with Devialet: they may be useful to them.

[IanG-UK]

Good point Alex.

Maybe terminology has lapsed and we need to classify these noises as (a) continuous disturbance; (b) drop outs; or © crackles.Or something better.

As far as I can discern, crackles/drop outs are network related whereas continuous disturbances are clock related. But others here may be able to categorise/establish cause/identify solutions better than I can.

[Womaz]

Great news Ian. The new beta did not work for me either, but when I move my laptop into the same room as my router , well so far I have had no white noise. It has been 2 weeks now with none. As soon as I move the laptop back into my work area, the white noise kicks back in within an hour or so.

Hopefully they now have more to go on and may be able to finally solve this.

@Womaz, this conversation probably does not belong to this thread, but rather the one specific to "WN on Windows", but I'll reply to your post anyway. I've just realized that you're not using an Ethernet connection, but WiFI. The fact that your problem is cured by the laptop's proximity to the router tells me that it's WiFi/signal strength dependent, and probably has nothing to do with AIR proper. I also suspect that the "white noise" you're getting is (somewhat) different from the variation most of us are afflicted with. Or else, there is a possiblity that with a poor WiFi connection, AIR has to work harder (there being more errors and retries to deliver the lost packets), and this makes it more vulnerable to the timing issues Ian had described to us. Could this (heavy traffic from AIR) also be the reason why the 24/192 format is more likely to expose the problem?
Anyhow, it's essential that you share your observations with Devialet: they may be useful to them.

Hi I am using the Ethernet setting on the 200 but my lap top is wifi yes. To be honest I have not really found devialet to be that helpful in trying to sort out my issues. I now mainly use my CD transport to play music and I am hoping that Dialog solves the issue.

When the lap top can be in the same room then its not a problem. I am no computer expert but I have never had wifi issues in the past, but I do agree it seems to be that this is the issue. I have always avoided having my laptop as part of the music chain, other than for ripping CDs and I think I will go back to this in the future too as there are too many things that can go wrong. Software conflicts etc and all I want to do is listen to music.

Thanks for the input

[Hifi_swlon]

Great news Ian. The new beta did not work for me either, but when I move my laptop into the same room as my router , well so far I have had no white noise. It has been 2 weeks now with none. As soon as I move the laptop back into my work area, the white noise kicks back in within an hour or so.

Hopefully they now have more to go on and may be able to finally solve this.

Thanks. However, the fix for Windows, as already issued as a beta, will not work for Mac. The Windows fix is a workaround because the fault is one embedded in Windows: it is not an AIR fault per se. The fix involves using the best clocking device accessible on Windows - and it appears to improve things but not fix things entirely.

On Mac, AIR already uses the best clocking device in the Mac OS so they will have to devise another approach to solve the Mac issue.

Edit: Of course, this highlights a point often made by Guillaume. Computer audio might be inherently a botch because it is not dedicated; but more significant in my view is that Devialet will inevitably be subject to possible interferences as Microsoft or Apple make OS modifications with no regard as to whether Devialet are happy with them. So a white noise "fix" this year might need another in 2016 and another in 2017 and ...

I'm still catching up on all the crackling/popping/white noise issue, having only recently bought a D200.

I have had lots of popping/crackling noises (daily or every other day at least), and only once a kind of weird throbbing noise that got worse and worse. Of course the latter could have started out as popping and crackling, as I was out of the room for a long time and on the other occasions I just kill iTunes and/or Air as soon as I hear it and it goes away.

Is this white noise? I suppose it doesn't matter really, there's clearly an issue whatever it's called.

I'm not really that convinced that its an OS problem if that's whats being said? It just seems a little too convenient that Windows and OSX both have almost exactly the same bug that only seems to manifest itself in Devialet's proprietary streaming software? Or is it just me? And even a workaround for OS bugs should surely be fixable within a release cycle?

I'm using Air (Eth), on OSX 10.9.x, through structured cabling and a gigabit switch, on a network that I've never knowingly experienced any glitches other than since the Devialet's arrival…

I also experienced it via USB, but that was using Audirvana and I think that was probably the issue there (although I was still surprised -my Chord Hugo never had any issue). I stopped using USB as my goal is to move the music server to another room anyway. I might try wifi this weekend of I can be bothered.

[AlexS]

Good point Alex.

Maybe terminology has lapsed and we need to classify these noises as (a) continuous disturbance; (b) drop outs; or © crackles.Or something better.

As far as I can discern, crackles/drop outs are network related whereas continuous disturbances are clock related. But others here may be able to categorise/establish cause/identify solutions better than I can.

Ian, I think describing white noise as "continuous disturbance" is being too politcally correct.
White noise == very loud (although that depends on the current volume) steady (not throbbing ) hissing coming from both speakers (mainly the tweeters and mid range drivers). Occasionally, it starts with music barely audible against the relatively quiet hissing background, but soon switches to full blown loud hissing, with no music discernible whatsoever.

To those attempting to stream over WiFi (vs Ethernet) in any part of the music chain, this, in my opinion, is a hopeless undertaking, if you go beyond the CD format, unless you live in a secluded mansion. Even many network streamers (e.g. Moon MiND) specify in their documentation that you have to be fully wired, to stream above the 16/44.1 resolution. But then those streamers, accept the compressed files over DLNA, whereas AIR sends uncompressed packets to Devialet, further increasing the network traffic.
Incidentally, I was never able to use WiFi even for 16/44. I purchased an AirPort Extreme, and trouble shot the WiFi network, to make sure I was streaming on an empty channel. All to no avail. It would play fine one day, and give me hiccups the next one. Only later did I realize that my neighbor's wireless phone was responsible for those interruptions.
Going fully wired (to my dedicated AirPort Extreme router) solved all problems for me, except for the occasional white noise, which issue, it appears, is getting better and better with the recent iterations of AIR (beta-1 and beta-4).