How do I convert vobs/mpeg 2 to avi?

[Sgt_Strider]

Also does this decrease quality or what? Since I'm only doing a format conversion...

[Neo Neko]

Well this is not per say the propper category for this question. But I will try to answer it as best I can.

MPEG1 and MPEG2 video have no way to be put in an AVI container as we all know. And of corse it will be better not to have MPEG4 in AVI someday. So this means not only a container conversion but a format conversion. And lossy format conversion incurs loss. Like making a copy of a copy of a VHS. If it is done right the loss is not to bad. But there is some loss.

In terms of MPEG1 and MPEG2 audio it is possible under very rare circumstances to get them in AVI and play them back. But it is generally more trouble than it is worth.

As to conversion/transcoding check the guides here at doom9. It is one of the hottest topics.

[jggimi]

It's more than a format conversion. Most mpeg-2 streams are greater than 4 GB; .avi filesize is restricted to 4GB -- 2 GB in older software. There are some who report audio sync errors above 2GB, too.

So there must be a compensatory reduction in bitrate as well. And you're already using and familiar with Gordian Knot, which does this.

If you're looking for ways to convert MPEG-2 rather than .vob streams, continue to use DVD2AVI, just open your .mpg or .m2v files from the Open (F3) menu.

[ChristianHJW]

Dont try this with AVI. Although the OpenDML AVI standard does allow files > 2 GB there is almost no proper support for this in the AVI handling tools ( even Vdub cant handle it correctly, as Avery was pointing out ), and also MPEG2 video in AVI is a big no-no ( the only existing MPEG2 VfW codec will only output I frames ... lol ).

It could be done with OGM i believe, if Tobias added this functionality to the OGM DShow muxer and parser ( no idea how difficult it was though ).

The matroska container dev crew is planning to make MPEG2 video streams possible to be stored in matroska files, and we wont rely on DirectShow here for sure, but will implement our own way, as usual with a x-platform compatibility in mind. The problem here is that the MPEG container is basing on a number of 'levels', each of those containing certain type of information. Some of it can be represented in the matroska container by default, so we can drop those levels and dont need to store them twice ( saves overhead ). Some of it is vital, so we will want to simply copy these and treat it as data. For other information levels it may be necessary to introduce new matroska elements ( i love EBML ) to support them properly, or we decide to copy them 1:1 as they are in the MPEG container. Biggest problem to respect here is that we dont have UCI for the time being and we dont want to hardcode MPEG2 decoding support into the matroska parser ( for licensing reasons ), like we will do with MPC and Vorbis as an interims solution. As a result of this we may have to be able to call existing DShow filter for decoding, means that the output from our matroska parser/splitter has to be compatible with the MPEG2 DShow filter's input. Any suggestions or comments from the MPEG experts here ( Nic, trbarry, -h ) are more than welcome, as always.