Basic kbps/compression question

I am converting a screen capture video to wmv. There are two options, Video 9 and Video 9 Screen.

now, each resulting wmv video is the same frame size, same fps, and same kbps, but the resulting files are of a vastly different size; the "screen" wmv is like a third of the size.

My understanding was that the kbps, multiplied by the duration, gave you total file size, and that the visual quality really depends on how many fps and how much motion you tried to cram into each second.

but here I have same video, length, and kbps and the file sizes are different. I can only assume that the "screen" version actually dips far below the target kbps when it can. right? and that the kbps I selected is only a "maximum" threshold?

The complexity of the image would only affect file size if you have a VBR setup, right? Now, if the resulting video is unwatchable, I might decide to go back and re-encode at a higher rate (use a different motion factor in the Kush gauge calculations) or change the key frame frequency. Different codecs would manage compressing motion better than others, sure.

Here are two network impact graphs for the two different videos during streaming viewing.

Video 9


Video 9 Screen


That sure looks like CBR and VBR respectively.

But, Mediainfo shows the video 9 as
File size : 4.65 MiB
Duration : 1mn 16s
Overall bit rate mode : Constant
Overall bit rate : 511 Kbps
Maximum Overall bit rate : 514 Kbps


and the Video 9 screen as
File size : 2.08 MiB
Duration : 1mn 16s
Overall bit rate mode : Constant
Overall bit rate : 230 Kbps
Maximum Overall bit rate : 514 Kbps

not sure how you have a CBR with such a different between overall and maximum bit rate....

If the video portion was indeed encoded at 230 then that's a flaw in the software. I selected 500 kbps as the output quality. and if I right-click/properties the wmv file, the details show 500kbps as the data rate. Maybe windows media player is just going by the intial lump of data (reflected in that graph).

FYI this is the export functionality from the WebEx Recording editor, for editing and converting desktop recordings made with webEx Recorder. There's a kbps drop down selector, and identical settings were selected for each export.

And from what i've read CBR is preferred if you are concerned about predicatable network impact, not varying content quailty, motion, and spikes in kbps that affect buffering or QoS. I'm working in a govt facility so this is important.

thanks for your input everyone.