The wizard essentially floods your upstream connection with data to illustrate the difference in call quality when PBM is off and on. In our baseline test call (without any interfering data or PBM involvement) our voice was played back without any problems.
In the second call, with interfering data being sent but with PBM inactive, the playback was unintelligible, indicating that our outgoing Skype voice data had in fact been squelched by the competing traffic. Finally in the last call, with the data still flowing but PBM turned on, our playback quality was markedly improved and almost as clear as it had been in the first call.
Built-in demos are all well and good, but to see how well PBM prevented nonessential network traffic from impeding the important stuff in the real world, we used Skype to make calls that included not just voice but video as well. While the Skype call was in progress, we uploaded a multi-megabyte file via YouSendIt, a Web site that provides an alternative to sending large files as e-mail attachments.
Although we expected this scenario to degrade our Skype call to some degree, we were surprised that the call quality as reported by the other party remained unchanged-- which is to say, good-- whether or not PBM was enabled (you can toggle it on and off via the tray icon). Even when we increased the number of files being uploaded simultaneously to two, then three and verified the increased amount of outbound data via Traffic Manager-- Skype worked fine with or without PBM.
We then tried the Skype calls again, but this time while sending a large file to someone via e-mail and simultaneously uploading another file to an FTP site. Again, the presence or absence of PBM didnt cause any detectable difference in Skype call quality. (Because Traffic Monitor is unavailable when PBM is disabled, we couldnt use it to compare the upstream data rates when PBM was on to when it was off.)
The fact that we were hard pressed to see a PBM effect in action outside the demo suggests that the demo was an extreme scenario that you probably wont encounter unless you have an especially slow or taxed connection.
We dont doubt that PBM is actually performing traffic management, but evidently it wasnt necessary given the 4,250/360k connection we enjoyed in spite of our attempts to overload it. With something like lower-end DSL service (typically 768k up/128k down) or a cellular modem connection, which is slower still, it would have likely been more useful.
Its also worth noting that because most people use shared Internet connections and PBM runs on individual PCs, the software cant manage your systems use of bandwidth relative to others. In other words, PBM can make the best of what you have, but how much you have depends on how many others are on the network and how much bandwidth theyre using. Propel says a future version of PBM will introduce the ability for multiple computers to work together to optimize bandwidth.
For some people, Propel Personal Bandwidth Manager may be worth the $29.99 price tag (as of this writing its being offered at a discounted introductory price of $19.99) just for the detailed connection usage data it provides. Most will want to start with the fully-functional 30-day trial version of PBM from Propelpbm.com, so you dont have to part with any cash until you determine its usefulness given your own connection speed and usage patterns.
Pros: Built-in Traffic Monitor shows you exactly what applications are putting a load on your Internet connection.
Cons: May miscalculate available Internet bandwidth; benefits of bandwidth management not always evident on a 360K upstream cable modem connection; currently not Vista-compatible.
Joe Moran is a former editor and analyst with Ziff-Davis Publishing. He's currently principal of Neighborhood Techs, a technology service firm in Naples, Fla. He holds several industry certifications, including Microsoft Certified Systems Engineer (MCSE) and Cisco Certified Network Associate (CCNA).