I think you missed my point... with the older, slower service, there were NO quality issues. It was perfect, clear as a bell. It's only on the faster service that I'm having issues.

Purchased package is Max Turbo (24 Down, whatever up).

DMZplus is also configured to point at the laptop we're VPN'ing from... no change.

What I'm wondering... if I was to do the router behind a router idea, would I possibly be able to mark my VPN traffic with the same QoS markings AT&T uses for voice and see some improvement?

He understood your point, however you didn't post your actual internet bandwidth package (which now is stated as 24Mbps down/3Mbps up). BTW, this wouldn't make a difference anyway. VoIP rarely uses above .5Mbps even.

To put the device on DMZplus mode that's using the VPN, access the modem's GUI by typing in 192.168.1.254 into the address bar. Go to Settings -> Firewall -> Applications, Pinholes, and DMZ. Choose the device -> Allow all applications (DMZplus mode). Press save.

Router behind router would work if you put it in DMZplus mode with router behind router detection as well.

Well, my point was that it worked on the slower connection just fine, so when he said if I had the same speed on the new service, it would have problems... didn't get my point. :-)

As stated above, my work laptop (the device doing the VPN connection) is in DMZplus mode already.

Does anyone have any idea how AT&T marks voice traffic on the Uverse network? Perhaps I could do router behind router and pre-mark all of my VPN traffic as EF, IP Prec 5, or whatever they do.

You know, I've posted this story so many places, I guess in this one I missed a piece... my apologies...

On the old AT&T FastAccess DSL 1.5/384 connection, the service worked FLAWLESSLY. No drops, no jitter, no poor call quality. It was only after moving to the Uverse connection that the call quality went down.

Jitter isn't being measured in this case, but assumed (on my part) based on previous experience. I have access at work to a multi-billion dollar Cisco/Juniper network (I work for an ISP with service that spans the entire country). In previous roles withn the company, I've performed VoIP testing between all manner of devices... Cisco IAD's and ISR's, UC500's, Acme Packet SBC's, Sonus and Metaswitch devices, Hammer network probes, etc.

In this case, I'm unable to actually measure jitter due to a number of reasons, but it certainly approximates the symptoms I've seen in the past when it was a jitter-related issue.

For what it's worth, a Cisco soft phone seems to be performing much better than using either MS Communicator or LiveMeeting for making calls. Unfortunately, LiveMeeting is the one that I have to use the most, so it's much more noticeable.