1 vs. 2 pass VBR

[Bobby]

Could someone give us some facts on 1 vs 2 pass VBR in terms of quality if the intended result in a DVD?

My current workflow is to generate an MPEG2 from PRE7, and then bring that into DVDAS for final menus and DVD generation.

I also noticed that that my hour (actually about 1:05) of video is coming out at only about 3.9 GB from PRE7 if I output to DVD image on my hard drive, and if I generate an MPEG it is about 3.4GB using the supplied widescreen DVD preset for MPEG generation. So obviously I have some parameters set too low?

[Paul LS]

I guess it should be 1 pass VBR vs 2 pass VBR vs CBR. So, as far as I understand one pass VBR can not see "ahead" and can only analyze a single frame and set an average bitrate on the frame alone while predicting future frames, this limits the bitrate variation allowed. Whereas in two pass the video is scanned on the first pass to determine from the content of the video frames (fast movement compared to a stationary scene for example) what the optimum bit rate is. And then it is encoded on the second pass. This allows the size of the file to be optimised compared to a single pass. However, there is an upper limit to the bitrate set by the DVD standard... around 9.2Mb/s dependant on the audio format, as total bitrate must not exceed 9.8Mb/s. So this can not be exceeded. If you have space on the DVD you could use CBR set to 9.2Mb/s (8Mb/s in in PE) and get the highest quality independant of frame content. However if you have limited space then VBR is better as it can asign a higher bitrate to fast action whereas with CBR you are stuck at the lower bit rate determined by the available space.

[Bobby]

Thanks Paul. So, to paraphrase: if it fits, using CBR would be the best alternative. I just set up a run with CBR at 9 (the highest the slider would go) and let's see what it generates.