ogg.k.ogg.k at googlemail.com
2008-Nov-14 11:30 UTC
[ogg-dev] video chapters and subtitles in ogg containers
>> (odd, I did get this reply for Silvia, but not the original post) > > Hmm, it was properly CCed to the list.Yes, I found it in the spam bucket for some reason...> Chapters are a list of timepoints stored in the metadata. They are an > information for player software that is usually used to allow the user > to jump to certain significant points within a stream. This probablyI don't think anything currently in Ogg can do this. The closest I can think of is CMML's clip concept, but you'd still have to parse the entire stream to find them all. One could build an index of them to place them in a Skeleton message header though, but there is nothing to do that AFAIK.> Chapters like this are mainly used on video DVDs, but are also defined for > SVCD and OGM containers. My personal use is watching movies from SDcard > or harddisk. I'm the type of guy who often wants to watch only some > parts of a movie, and chapters greatly simplify finding a certain scene > (or skip one that is too exciting...).If you are transcoding a DVD, you might want to look at the diffs directory in libkate, there is a patch to Thoggen (Linux based DVD transcoder) that automatically converts DVD subtitles to Kate streams, though it also requires a set of patches to GStreamer (also found in the same directory) to be applied.
Hans J. Koch
2008-Nov-14 13:11 UTC
[ogg-dev] video chapters and subtitles in ogg containers
On Fri, Nov 14, 2008 at 12:30:25PM +0100, ogg.k.ogg.k at googlemail.com wrote:> > Chapters are a list of timepoints stored in the metadata. They are an > > information for player software that is usually used to allow the user > > to jump to certain significant points within a stream. This probably > > I don't think anything currently in Ogg can do this. The closest I can > think of is CMML's clip concept, but you'd still have to parse the entire > stream to find them all.That's nearly useless for DVD-like applications.> One could build an index of them to place them in a Skeleton message > header though, but there is nothing to do that AFAIK.I guess this is not neccessary. A simple chunk of metadata containing the list of times at the beginning of a file would suffice. OGM simply uses text stored in comment fields for that. I wouldn't recommend comments for a general approach, but whatever format we choose, it should probably not be too difficult to specify and implement it. Looking at some of the accessibility pages Silvia mentioned, I think that ogg specs are very much focused on web video. I'll follow that discussion and will remind ogg developers from time to time that there are still one or two people who watch videos from good old files on their disk ;-)> > > Chapters like this are mainly used on video DVDs, but are also defined for > > SVCD and OGM containers. My personal use is watching movies from SDcard > > or harddisk. I'm the type of guy who often wants to watch only some > > parts of a movie, and chapters greatly simplify finding a certain scene > > (or skip one that is too exciting...). > > If you are transcoding a DVD, you might want to look at the diffs directory > in libkate, there is a patch to Thoggen (Linux based DVD transcoder) that > automatically converts DVD subtitles to Kate streams, though it also requires > a set of patches to GStreamer (also found in the same directory) to be applied.Well, I gave up using graphical DVD rippers years ago since none of them does what I want. Meanwhile I got used to create my audio, video, subtitle, and chapter data using various commandline tools. The best solution for me would probably be oggzmerge understanding srt files and some textfile containing my chapter points. Thanks, Hans
Ivo Emanuel Gonçalves
2008-Nov-14 13:26 UTC
[ogg-dev] video chapters and subtitles in ogg containers
On 11/14/08, Hans J. Koch <hjk at linutronix.de> wrote:> The best solution for me would probably be oggzmerge understanding srt files > and some textfile containing my chapter points.Actually, you just have to use ffmpeg2theora + srt file. Though you won't have chapters for now. -Ivo
Conrad Parker
2008-Nov-14 19:31 UTC
[ogg-dev] video chapters and subtitles in ogg containers
2008/11/14 Hans J. Koch <hjk at linutronix.de>:> > I guess this is not neccessary. A simple chunk of metadata containing > the list of times at the beginning of a file would suffice. OGM simply > uses text stored in comment fields for that. I wouldn't recommend > comments for a general approach, but whatever format we choose, it > should probably not be too difficult to specify and implement it. > > Looking at some of the accessibility pages Silvia mentioned, I think > that ogg specs are very much focused on web video. I'll follow that > discussion and will remind ogg developers from time to time that there > are still one or two people who watch videos from good old files on > their disk ;-)that's a fair comment :-) we designed CMML for web video, and SIlvia's current accessibility work is targeting web video. For this discussion on generic Ogg mechanisms, perhaps we should take plain disk usage into account too ... I guess it doesn't help you to say that a web video player can get the table of contents -- a plain XML file of the CMML clips -- before playback by requesting the media resource with a Content-Type preference for text/cmml. Of course that's only the case if the server supports that, and if it's not a live stream (ie. if all the clips are known in advance). Conrad.
Ralph Giles
2008-Nov-14 19:45 UTC
[ogg-dev] video chapters and subtitles in ogg containers
On Fri, Nov 14, 2008 at 5:11 AM, Hans J. Koch <hjk at linutronix.de> wrote:>> I don't think anything currently in Ogg can do this. The closest I can >> think of is CMML's clip concept, but you'd still have to parse the entire >> stream to find them all. > > That's nearly useless for DVD-like applications.I don't know. Wouldn't it work if the entire CMML clip track were packed at the beginning of the stream? That removes the seeks entirely. I actually think that should be allowed by the spec, although I'm not sure whether it is, currently. Is there a way to tell we have complete converage from a given collection of cmml packets? -r
Silvia Pfeiffer
2008-Nov-15 13:53 UTC
[ogg-dev] video chapters and subtitles in ogg containers
On Fri, Nov 14, 2008 at 10:30 PM, ogg.k.ogg.k at googlemail.com <ogg.k.ogg.k at googlemail.com> wrote:>>> (odd, I did get this reply for Silvia, but not the original post) >> >> Hmm, it was properly CCed to the list. > > Yes, I found it in the spam bucket for some reason... > >> Chapters are a list of timepoints stored in the metadata. They are an >> information for player software that is usually used to allow the user >> to jump to certain significant points within a stream. This probably > > I don't think anything currently in Ogg can do this. The closest I can > think of is CMML's clip concept, but you'd still have to parse the entire > stream to find them all. > One could build an index of them to place them in a Skeleton message > header though, but there is nothing to do that AFAIK. >Yes, CMML was definitely built with chapter markers on mind. What you do is get the full CMML file in parallel or ahead to the video file. It has all the clips defined. The clips have id tags, which give you text to identify the chapters. They also have title tags for more than just a one-word description. You should however be able to simulate that with a kate stream, too. When multiplexed into the file you will indeed gather the markers as you go along. Regards, Silvia.
Silvia Pfeiffer
2008-Nov-15 13:58 UTC
[ogg-dev] video chapters and subtitles in ogg containers
On Sat, Nov 15, 2008 at 12:11 AM, Hans J. Koch <hjk at linutronix.de> wrote:> On Fri, Nov 14, 2008 at 12:30:25PM +0100, ogg.k.ogg.k at googlemail.com wrote: >> > Chapters are a list of timepoints stored in the metadata. They are an >> > information for player software that is usually used to allow the user >> > to jump to certain significant points within a stream. This probably >> >> I don't think anything currently in Ogg can do this. The closest I can >> think of is CMML's clip concept, but you'd still have to parse the entire >> stream to find them all. > > That's nearly useless for DVD-like applications. > >> One could build an index of them to place them in a Skeleton message >> header though, but there is nothing to do that AFAIK. > > I guess this is not neccessary. A simple chunk of metadata containing > the list of times at the beginning of a file would suffice. OGM simply > uses text stored in comment fields for that. I wouldn't recommend > comments for a general approach, but whatever format we choose, it > should probably not be too difficult to specify and implement it. > > Looking at some of the accessibility pages Silvia mentioned, I think > that ogg specs are very much focused on web video. I'll follow that > discussion and will remind ogg developers from time to time that there > are still one or two people who watch videos from good old files on > their disk ;-)Not kate, no. Kate was definitely defined for offline work.>> > Chapters like this are mainly used on video DVDs, but are also defined for >> > SVCD and OGM containers. My personal use is watching movies from SDcard >> > or harddisk. I'm the type of guy who often wants to watch only some >> > parts of a movie, and chapters greatly simplify finding a certain scene >> > (or skip one that is too exciting...). >> >> If you are transcoding a DVD, you might want to look at the diffs directory >> in libkate, there is a patch to Thoggen (Linux based DVD transcoder) that >> automatically converts DVD subtitles to Kate streams, though it also requires >> a set of patches to GStreamer (also found in the same directory) to be applied. > > Well, I gave up using graphical DVD rippers years ago since none of them > does what I want. Meanwhile I got used to create my audio, video, > subtitle, and chapter data using various commandline tools. The best > solution for me would probably be oggzmerge understanding srt files and > some textfile containing my chapter points.We might be rather close to the specifications for that. Tool support is a different question, as usual. Regards, Silvia.