The original .vob( 720/480, 6448kbs,FR 29/sec ABR 448kbs,Ass 48khz) at 97.9 mb gave these results for 720/480,6000,29.7,192,44. which should be comparable quality to the original. Then for 480/320

720/480 480/320

codec X X

xvid 3.82 6.3

mpeg4 4.00 6.5

wmv1 3.42 6.0

It is very interesting that decreasing the bit rate does not do nearly as much to reduce file size as you would think it would. In fact, if you go back and look at the data I put in the.rtf file, you'll see that 2000 kbs at 720/480 gave the SAME compression( X=3.7) as 6000 kbs averaged above. 2000 kbs at 480/320 gives the same also( X=6.3). In fact notice, there is no difference at all at 480/320 between 2,000kbs and 1,024 kbs!

The data in the .rtf file were gathered strictly to allow some empirical conclusions. If I had not been so tired and frazzled late at night( should have gone to bed,finished later), I would have included the data and conclusion I just put forth. My final conclusion: Changing the frame size has much more to do with the final file size than bit rate.

So you can maintain quality for the most part with xvid,mpeg4 and wmv1 at the native 6000 kbs and get a value of X=3.7 which might actually increase as the file size increases to 4.0 or more. The formulas, just as a reminder :

X = orig / comp.
orig = X (times) comp.
comp.= orig / X

This has been illuminating for me. I would not have predicted these results. It probably takes someone who is familiar with the actual codec algorithms and formulas to really understand things. Which goes to show how in life and science we intuitively assume things to be true that aren't because we really don't know what we're talking about! Evidence, evidence,evidence! What if the climate models are oversimplified( they have to be) or based on incorrect assumptions? Any errors in the inputs are grossly magnified, in addition, with the enormous complexity of the calculations. Sorry, off the subject,rambling again,lol..

Take care,
boomer