Bufferbloat...another audio quality impact?

[bear105]

I've been trying to track my repeated and random dropouts between my hosted scanner and the RR audio servers. I'm sure that sometimes, its a hit to my home network, the ISP or the Internet backbone. But this concept came to my attention (via Slashdot) and it sure looks like it could have an impact on our audio connections, as well. The simple idea being that our current Internet protocols, such as TCP/IP, were not designed to work with very large buffer's, common on today's routers, switches and broadband modems, and thus create excessive latency and jitter.

The idea seems to have some history but its really Jim Gettys that has put it together, with the help of other Internet Pioneers, such as Van Jacobson, David Clark and Vint Cerf. The good news is that the reputation of these individuals are getting attention with numerous companies, such as Comcast.

Jim Gettys Definition - Bufferbloat is existence of excessively large (bloated) buffers into systems, particularly network communication systems.

Bufferbloat FAQ page from Jim Gettys Blog.

I realize its best to Keep It Simple, but after trying numerous setups, running all kinds of network metrics, I am looking for something else to blame for my connectivity problems. Interesting reading none the less....

 

[GordonE]

I would suggest installing PingPlotter ( from PingPlotter , you can use it for a while without needing to purchase it ) and setting it up to ping the server that your audio is being sent to (audio3.radioreference.com) at regular intervals. After you see your feed drop, check to see if PingPlotter saw any pings not be returned around that time, if it did, look at the data it recorded to see where the problem was.

Here are some instructions I typed up in the past when helping another user:

* Enter "audio3.radioreference.com" in the "Address to Trace" field in the upper-left corner of the window.
* Change the "Trace Interval" value to 1 second.
* (optional) Click on the horizontal line that is displayed above the small graph in the bottom-right area of the window and drag it up to make the response time graph larger.
* (optional) Select Options from the Edit menu and then select Auto-Save. Put a check in the "Auto-save data" checkbox and then click on the Apply and OK buttons.
* Click on the Trace button in the bottom-left corner of the window.

The next time your feed goes offline, see if the graph in the bottom-right area of the window indicates any problems (the response time graph should look different than it did when you started the program).

 

[bear105]

Excellent tool and analysis technique. Will give it a shot...Thanks!