TCP/IP cannot guarantee that data arrives "on time", so it doesn't matter whether one is using wi-fi or ethernet, it still has to be buffered and reclocked.
25 responses Add your response
Jitter only comes into play during the analog to digital conversion process which occurs all within the squeezebox using the internal DAC or between squeezebox and external DAC if external dac is used.
The only issue wit Wifi comes if teh bandwidth is not sufficient to provide data to the Squeezebox fast enough to convert in real time. The result is temporary delays in the playback until the Squeezebox receives enough data to proceed converting to sound in real time.
Wifi bandwidth is affected by noise/interference on specific channels, physical obstacles in the path of teh transmission between remote devices and router, and physical distance between router and remote devices.
Kijanki, yes, buffering helps reduce dropouts however I believe if throughput between server and Squeezebox client goes too low, the buffer still might not be filled fast enough to keep music flowing. That is what I experience on occasion with similar Roku Soundbridge product, and would expect similar with any network player device.
Yes, I have had occasional dropout issues with wireless G as well and found changing router frequency can help.
With the default frequency/channel configured in my router for wireless G, I found that my Rokus would loose their connection to the server temporarily whenever our wireless phones rang for an incoming call. The wireless phones were apparently interfering on that frequency. Changing router channel configuration solved that problem.
Hadn't considered microwaves. Next time I get a delay, I will have to check to see if the microwave upstairs is in use......
I don't think wireless dropouts will be a problem unless your signal strength is marginal. You probably already know whether this is the case if you use wireless in the house.
I certainly don't see any basis for a statement like, "For a hi resolution system, I would avoid wireless. For streaming background music and such, it is great."
Its true that higher resolution and lossless files have more data to transfer and may be more prone to delays due to throughput limitations on a wireless connection if present. For example, I seldom get a dropout playing mp3 files, but it can happen on occasion with CDs ripped to lossless .wav format.
Thanks for the clear and concise answer. The good thing about dropouts is there's no second guessing, like with jitter: you either get or don't get dropouts.
I'm taking note of bandwith as an issue to keep in mind, especially since I'd like to experiment with hi-rez files, that are large.
FWIW, my router is Wireless-N and the PC is 802.11b/g/n, and I'm moving the router so it needs to go through only 2 internal brick walls perpendicularly to get to the Touch (down from 3 walls at 45-degree angle). Does that seem ok/bad/borderline?
Seems like the Touch is a rather safe and cheap way to get into hi-rez without any significant downsides, even for a $15k system like mine (pre/amp/speakers). Do you agree? Of course a better DAC will be needed down the road.
I use a Touch in my system wirelessly and never have dropouts, but I gnerally have a 90-95% signal strength at my listening position. By comparison, I also have a prior generation squeezebox (SB3) that has been heavily modified to improve the digital output and I find the unmodified Touch to be as good or better. I wouldn't worry about jitter with the Touch.
"Does that seem ok/bad/borderline?"
Well, my internal walls are drywall, not brick, so hard to say but I suspect brick would be more of an obstruction.
You should be able to see the "number of bars" or % connection strength on the computer used as music server.
Does Squeezebox provide a similar readout of signal strength? If so, that would be the way to tell.
I would say with wireless G if you fall below 80-90% signal strength in either connection to router (from server or from Squeezebox) that rebuffering delays are more likely using lossless files ripped from CD. I use lossless .wav format.
Higher rez files utilizing more data per second of music will up the ante further. I do not have any experience there I can relate.
Newer computers offer wireless N which has significantly higher bandwidth. Sam true of routers. I would expect newer networked music devices to follow suit also at some point at which time the issues with higher res files will be relieved somewhat.
Audioengr reminded me that a Wifi network is shared and used by all active devices connected to it. it is only dedicated to streaming audio from server to player if nobody else is using it for tasks that consume bandwidth (web browsing, file downloads, etc.). So concurrent usage for things other than streaming audio can cause delays/dropouts.
For example, I have two Roku players on two systems that conenct via wireless G to the same music server computer. Often if I play both at the same time with lossless CD .wav files, the dropouts/delays will occur regularly and become an annoyance. Less of a problem as I mentioned earlier for lossy or lower resolution files, like MP3.
54Mbps is what wireless g can do optimally based on the spec I think, right?
Weak/sub optimal signals reduces that bandwidth.
Also I've found not all g routers are the same. Actual bandwidth at a particular range seems to vary from router to router. There may be many reasons for this
Processors and programming can vary.
Differences in the performance of the radio transmitter /receiver technology may be another factor. Antenna's used are one factor. I suspect there are others also.
802.11g runs at up to 54MbpsNot in the real world. 15 mbps would be a reasonable expectation for a fairly optimal setup, and I'm sure that less than 10 is not uncommon.
I don't have a feel yet for typical 802.11n performance, but given that the theoretical maximums that were advertised for a, b, and g were all essentially fictional, I'd be surprised if the same didn't apply there.
Al, I've never done the math or quantitatively measured bandwidths, but from what I have seen I suspect your real world estimates to be reasonable.
Wireless n probably also does not always perform to the max based on teh spec but I would expect perhaps still a significant improvement over g in practice.
I have a wireless n router but my server and players are all wireless G, so I have not really leveraged the current wirless n technology in practice to date.
Again, just to be clear for all, these are limitations that determine how fast digital data can move through the digital wireless "pipeline" as input to the Squeezebox. Jitter is not a phenomena that pertains to transmission of dogital data. Jitter comes into play during the DAC process. Data input to the DAC process is provided by the Squeezebox, regardless of whether the SB's internal built in or an external DAC is used.
I know this thread hasn't been active for a few months, but here's my two cents on the wifi issues I have experienced with the SB touch; First the touch will show you network and wifi status, including signal strength and a data integrity graph that can be set to specific KBPS rates so you can easily see if your network throughput is sufficient for the files you are playing. In my experience playing WMA lossless over wireless G, the only time I see buffering dropouts is when there are multiple users on the network doing high data rate downloads, even doing an internet speed test while playing music won't cause buffering if those two operations are the only network traffic. My issue with the SB touch has nothing to do with playback or sound quality, rather I have problems with the SB connecting to my computer after the computer has gone into sleep mode...if I try to connect (play music from the library) while the computer is asleep the network (SB server network, not the work group) fails, requiring a reset of the SB touch to reacquire the connection about 60% of the time. It also occasionally (10% maybe) fails to reacquire the network even if the computer is up and running before the attempt, but had been asleep since the last SB playback.
My solution is that I'm going to use an external USB hard drive to hold my music library, obviating the need to stream the data, it will eliminate the network issues (*never* have problems with Internet radio connectivity), and gives me another backup of the music to boot! It won't even be very costly as a decent external 1TB drive is under $100 these days.
This thread went offtrack to WiFi discussion, but does the OP suggestion of using the USB on the Touch as an output make sense when it as a perfectly good SPDIF connections, both optical and coaxial?
I have read that the Touch does not do well with large music collections through it's USB port because of the hardware just isn't as fast as a dedicated server.