To be fair with you, I never took the time to do some good testing about this. Doing 3-pass VBR would take more than 24h to process 3 movies, for me. Some people at doom9.org argue that doesn't make too much difference, some say it's 3-pass VBR is much better. I would like to see some tests and screenshots if possible.
But if you decide to test this, remember: the idea behind multipass VBR is make a better picture by optimizing bitrate distribution. So you have to encode long segments to compare (a recommend a full movie). Do the following:
1) a first pass a Q 30, for instance, (which will create the VAF anyway, that later will be used to multipass). Move the generated MPV somewhereelse or rename, to avoid overwriting in step 2.
2) Then do a 3-pass VBR with the bitrate calculated from the movie generated in step one. At the end, both must have about the same size.
3) Create DVD2AVI projects for both encoded versions and use the following script to compare:
a=mpeg2source("movie_encoded_with_q_factor.d2v")
b=mpeg2source("movie_encoded_with_3_pass_vbr.d2v")
stackhorizontal(a,b)
Open in VirtualDubMod the script, and zoom at least 2x. Scan throught the movie and let us know if you find any difference! I might try this we I have time also.
Have fun!