Quote:
Originally Posted by
Andysol 
You lost me.....
All I got out of your post is you need to lower your expectations. :-)
Really? Maybe I am expecting too much.
Here's the basic rundown. When you load a website, get an iMessage, email, etc the data comes from the source to your device. Your device will inspect and acknowledge the data and if anything is corrupted or lost it can request new packets that are sent and then rechecked. There are many levels of redundancy built into the networking model to ensure delivery.
With VoIP (and by extension, video conferencing) all that goes out the window. There is a session set up to link the two end points and then the data is sent back and forth without the redundancy, error checking, and retransmissions that you find with TCP/IP. This is because RTP is better for UDP/IP. This is because TCP likes reliability where as UDP likes speed. This is important to VoIP because any delay is jarring to the listener. Even if we weren't used to standard telco voice calls feeling instant it would be jarring.
Now for VoIP to work it's not simply using the transport protocol with the smaller, simpler header but also making sure that these very specific packets using these very specific protocols are sent first. We call this QoS (Quality of Service). That means when all these packets get routed through router after router across the globe that the ones being used for VoIP are sent with the highest priority. This type of effort is considerably more complex and costlier to setup and maintain.I know this because I've done this.
It's easy to think of a 64Kbps analog call that can be compressed to 8Kbps for a modem voice codec as tiny but when you need that protocol to have no noticeable delay (or other routing artifacts) it gets tricky. Now at this point I've only really mentioned VoIP but video conferencing is no different, expect for the fact that it has all the same pitfalls as VoIP except that video requires a lot more data than simply voice. Now there is the benefit that a missing or corrupt packet might not be noticeable to the naked eye as easily with voice but there is so much traffic for video compared to voice that the issue is significant, especially when you through in QoS.
Does that clear how a webpage is different from a voice call?