Menu
▾
▴
#4 Calculations Are Way Off
The calculations seem to be way off. At first, I
thought it might be because three-pass encoding
optimizes in fast-motion parts of the movie. The DVD
I'm ripping is 1:46:43 long. My target file size is
650MB and a 96K audio bitrate. That calculates to 724K
video bitrate in quickrip. But I end up with an 811MB
AVI file.
Two- and three-pass encoding create approximately the
same sized AVI file.
Logged In: YES
user_id=83484
I'm using the latest CVS (as of 20030623).
Logged In: YES
user_id=83484
Maybe adding the "vratetol" option to the video codec
options would be a good idea. "vratetol" allows you to
specify a video bitrate tolerance, since the video bitrate
is variable.
I'm testing now with vratetol=1000 using one-pass encoding
and it seems to be keeping the filesize low. vratetol
supposedly defaults to 8000, where 1 = 1000 bits.
BTW, the movie I'm attempting to back up is "Equilibrium".
Logged In: YES
user_id=83484
vratetol=1000 seems to *really* tighten up the variable
bitrate of the encoding. However, compared to many other
DVD rippers, our bitrate calculations are conservative.
I've got a Windows bitrate calculator that says a bitrate of
746K is good for "Equilibrium". Whereas, Quickrip
calculates it as 724K. Perhaps with this vratetol option,
we could eek out a litle higher bitrate (instead of
multiplying the bitrate by 0.96). Or maybe we should make
vratetol a little more tolerant (say 2000?). Comments?
Logged In: YES
user_id=83484
Calculations only seem to be reasonably correct when doing
one-pass encoding. Three-pass encoding still produces an
800MB AVI regardless of the vratetol setting. Maybe the
optimization information produced by the first video pass is
causing the problem. Perhaps the optimization ignores the
vratetol setting? Or maybe three-pass encoding needs to
have a lower target video bitrate, since high-motion
portions of video cause the bitrate to increase well above
the target bitrate.
I haven't tested two-pass encoding to see if vratetol has
any affect on it. I'm about to test that now.
Logged In: YES
user_id=83484
Two-pass encoding also produces inflated AVI files. Two-
and three-pass create the same size file, it appears. In
this case, Equilibrium rips to a final size of 837063494,
when my requested size was 650MB. That's about 145MB too
big. Any thoughts why?
Logged In: YES
user_id=175451
Have the calculations been that far off for anything else?
Since I put in the 0.96 multiplier, I've never had QuickRip
be off by any more than 10mb.
Logged In: YES
user_id=83484
Well, I've only been working with Equilibrium. But I can
try ripping one of my other DVDs again for reference.
Trying to think what would have similar video patterns...
I'll try another DVD sometime during the week to see if I'm
having similar problems.
Logged In: YES
user_id=83484
Tried three-pass encoding with "Bourne Identity". It
appears to display the same issue. The movie is 1:58:10
long. Target filesize is 650MB. DivX bitrate is 645.
Resulting file is 821MB.
I'll check two-pass this evening...