a question about DVDrb

[mooms]

hello

First i would thank all the guys here for this great forum and also jdobbs for DVD rebuilder.

I love this program, it's really the best alternative between the transcoders with poor quality and the high ends programs too much bloated.

i using it since 3 months and a question keep bothering me so here it is:


Why splitting the movie in so much parts when encoding ?


doing that is not the best way to achieve maximum quality in VBR encoding am i right ?


i suppose this is for the sake of maximum flexibility (applying some filters only in a part of the movie etc...) but making a option to encode the movie in one time with only one AVS file could be a good option if i am right.
(at least in the movie only case)


i mean; correct me if i'm wrong but the principe of multipass encoding is to analyze the movie first so the encoder can give more bytes where it's needed.

So now if you're only encoding 2 or 5 minutes of film at the same time like DVD-RB does; the encoder can only analyse that part when it run the analysis pass right ?

Let says you have high bitrate needs in a scene a the beginning but 20 minutes later the movie have a lot of quiet scenes; the bytes -possibly- gained for theses quiet scenes cannot be used for the scenes of the beginning because DVDRB does not encode the whole movie in one stage.

I think this is obvious but maybe i was wrong somewhere in my theory.



I hope it will be clear for all; sorry for my english it's not my native language.

[blutach]

Have a look at the ECL file created as part of the preparation stage. It contains various bitrates for each segment after a quick analysis of the segment (rather than an overall 1shot does it all approach).

Regards

[mooms]

sorry but i don't understand your answer.
what's the point about the ECL file containing various bitrates for each segment ?


Regards

[jdobbs]

doing that is not the best way to achieve maximum quality in VBR encoding am i right ? That would be incorrect.

There are many very good reasons/advantages to doing encodes by segment. For one thing -- they may not all have identical coding characteristics. Another is for editing convenience.

There have also been many discussions on this subject -- and a search will give you more insight. But here's the bottom line: DVD-RB does a "first pass" scan of the original MPEG data when it does the PREPARE phase. The data it pulls from that scan determines the minimum, maximum, and average bitrates applied during ENCODE. It works just like a single long encode because that "first pass" data is used to decide the allocation across segments. In fact, if you were to compare the resulting VBR bitrate profile between a single long encode and a DVD-RB encode you find it is almost identical.

[mooms]

i have searched for this but i haven't found the proper answer.


In fact, if you were to compare the resulting VBR bitrate profile between a single long encode and a DVD-RB encode you find it is almost identical.

So i was wrong.
Thanks for your answer.