Maximum bitrate question.

[Jeffster]

I noticed something that seemed strange to me after running the Prepare phase on a DVD today, the maximum bitrate for each segment of the movie was different, which is not something I've seen DVD-RB do before on the same Title. One segment even had a maximum bitrate of only 6949 Kbps!

Is this because of the changes in v0.96 mentioned in the changelog "Added measures to more accurately measure bitrates in the source stream and then distributes it accordingly in the backup", and therefore should be considered normal, or has DVD-RB gone crazy on me? :confused:


http://img229.imageshack.us/img229/5552/maxbitrate4mc.jpg (http://imageshack.us)

PAL DVD, with one 5.1 DD audio 448 Kbps
all other audio and subtitles tracks deselected
DVD-RB 0.96

[random asshat]

I can't say for certainon behalf of the author, but I'm pretty sure it's intended,
as I get similiar fluctuations in the max bitrate using v0.96.

It seems to be adept at pegging those "studio logos" in opening cells
and dropping off the max bitrate .. although this is probably decided
mathmaticaly rather than just my good luck.

still very nice ..

[Jeffster]

Thanks :)

I guess it's a new "feature" then with 0.96 and not an aberration.
It still seems odd to me though...

[jdobbs]

It gives better quality by not allowing the reencode to overstep the original streams peak bitrate. If it did, it would be stealing from other parts of the stream.

[Jeffster]

Thanks for the explanation... what are the chances of that ever happening though?


BTW - I used the new version of CCE Basic 2.70.01.07 for the first time yesterday, no compatability problems with DVD-RB and the output quality was excellent, as usual :)

[jdobbs]

The chances are pretty high. I ran several of tests. What happens often is that the original will have a higher liklihood of error on high demand scenes and the quants will be higher. As a result there will be error and possible blocking in the output. When DVD-RB reencodes, the error is already there (from the previous encode). Suppose, as an example, you have a macroblock error from the original... sharp edges, like those of a macroblock, are more difficult to encode and demand more bits. So you are adding bitrate for no reason other than to try and reproduce an error.