Log in

View Full Version : Answers to 2 nagging "Why can't i..." questions


mpucoder
14th May 2005, 17:45
I discovered this while designing MuxMan's disk planner. Originally I thought I could do Scenarist one better by grouping titles together within a titleset that used the same segments, and then assign track numbers to audio and subs within the group (allowing navigation between the titles in the group without any problem). But then I asked myself, what would happen if a title from one group jumped to a title in another. And the answer to that is the same as:

Q: Why do all the titles in a titleset have to have the same audio and subpicture streams?
A: Because if a jump is made from one title to another, depending on the player, it will either continue playing the same decoding stream, thereby possibly switching languages, or switch to the decoding stream as indicated in the new PGC, thereby requiring an entry at every slot, even if there is no stream.

Q: Why can't I jump from one titleset to another?
A: Same as above, sort of. By forcing the use of a dummy PGC in VMG the player can now choose audio and subpicture languages for the new titleset. The process of choosing audio/sub language according to user preference and available streams only occurs with a JumpTT command.

So it turns out my concept of a "group of titles" is exactly what a titleset is.

Trahald
14th May 2005, 19:18
Makes sense.. althought still would be nice if the ability were there.. obviously even with all the DVD rules, the jump to a dummy vmg or using jump (if grpm ==) commands in the first play make it possible and 99 vts sets gives you plenty of room to work with. All the dvd rules are an attempt to simplify (or make less dificult) the coding in the roms in standalones... and of course setting hardware minimums helps too. thanks for the info, mpucoder.

Hopefully nextgen is looser.. (or gives more vts sets)

mpucoder
14th May 2005, 19:26
Glimpse of the nextgen here (http://forum.doom9.org/showthread.php?s=&threadid=94437)

influenza
18th May 2005, 05:39
Why do all the titles in a titleset have to have the same audio and subpicture streams?
Personally I find this very annoying. There are more than enough examples of dvds (and thus authoring systems) that have other audio/sub layouts throughout pgcs. That could be more or less audio and even audio with different bitrates (on the same track number).

I was kind of hoping that muxman would be somewhat more volatile in this respect, as opposed to scenarist :)

mpucoder
18th May 2005, 06:10
It will be, but not quite the way I originally thought it could be. Originally I thought I could handle an extreme case where there are 2 titles in a 16:9 Only Letterbox VTS. Each has 16 subtitle languages, with seperate streams for wide and letterbox, so there are 32 streams for each title. The two do not have any languages in common, so there are 32 languages altogether. Besides the fact that this scenario doesn't allow for mapping the 16 unused languages to an unused stream number (they would have to default to one of the used languages), it would not work properly if there was any navigation between the two titles. And that is the purpose of the VTS compartmentalization. The DVD player can be assured that everything within a VTS is consistent.

What MuxMan 0.15 does do is ensure within each VTS that the "decoding stream" numbers in every PGC map back to the same stream numbers, and in the case of 16:9 subs will be valid for the same display modes. This does not mean they all have to be defined in each segment used in the VTS. Muxman generates a map usable for all the titles, and uses that map for all the titles without regard to the streams being present. In other words you will not have to use dummy audio or subs to do things like 2 titles, one with English audio, the other with French. MuxMan will assign a track for each, and both titles will have identical mapping. The main thing is that the streams have consistent usage within a VTS.

influenza
18th May 2005, 07:16
Muxman generates a map usable for all the titles, and uses that map for all the titles without regard to the streams being present. In other words you will not have to use dummy audio or subs to do things like 2 titles, one with English audio, the other with French. MuxMan will assign a track for each, and both titles will have identical mapping. The main thing is that the streams have consistent usage within a VTS.

That sounds really nice :).