I'm in!
--A
Today I looked for Video+Subtitles software. Results:
1. SRT format looks good - let's stick to it;
2. Subtitle editors - I found gnome-subtitles which looks OK (and helps
to translate - two SRTs side-by-side, with preview). What it can't do,
though, is splitting the video automatically into time periods. Did your
software do that, Sergey, or how did you work it out?
3. Video and audio codecs: the lectures (at least L1 that I have) use an
AVI container with DIV3 video and Mp3 audio. They are not free! By
default, Linux distros do not ship them. The only alternative I found:
Theora (video) + Vorbis (audio). I am trying right now to transcode
Lecture-1a to these.
4. Container formats: AVI is good, but... It does not support SRT
subtitles, e.g. you can't package them into the same file. Ideally, we'd
like to package together with the video stream several languages in text
and audio streams, and let the user choose. Again, if we stick to free
containers, then it looks like we have Matroska (mkv), OGG, and OGM.
Which to choose? OGG is the most widespread, but subtitle support is bad
(unless you package it into Annodex, which looks complicated). OGM
works, but reportedly is a 'crude hack' that nobedy will remember in a
few years. I tried the Matroska and it worked with the AVI file
(put .AVI and .SRT inside) but failed to work with Theora/OGG.
5. Another (simpler?) way would be to use 'hard' titles, that is, render
the titles right into the video. I have not yet found software to do
that - most software seems to focus on 'soft' titles, and those do look
preferable.
--A
Thanks for the reply!
> > 2. Subtitle editors - I found gnome-subtitles which looks OK (and helps
> > to translate - two SRTs side-by-side, with preview). What it can't do,
> > though, is splitting the video automatically into time periods. Did your
> > software do that, Sergey, or how did you work it out?
>
> What do you mean by splitting the video into time periods? Why would
> you use it?
Oh, I mean the actual subtitle frames. You made 1012 of those - and
it's actually very good! I tried doing that once and it was so
time-consuming. Did you just manually decide where each frame
stops/starts? Or is there a software that can do that, maybe by sound
levels?
> > 3. Video and audio codecs: the lectures (at least L1 that I have) use an
> > AVI container with DIV3 video and Mp3 audio. They are not free! By
> > default, Linux distros do not ship them. The only alternative I found:
> > Theora (video) + Vorbis (audio). I am trying right now to transcode
> > Lecture-1a to these.
>
> Well, I am mostly MS Windows user and there're no such problems there.
> Video lectures I have (the lighter ones, DivX-encoded) can be played
> without any problem on WinXP.
No, there aren't any _technical_ problems on Linux either :) I
installed Mp3 and DIVX support - shhh - don't tell anyone. It is more
an issue of using FREE technology and being SAFE from litigation.
You know why Linux does not ship MP3 and DIVX? Microsoft does that,
and it was sued for $1.5 billion dollars for "patent infringements",
see, for example, here:
http://www.iht.com/articles/2007/02/23/business/web-0223microsoft.php
That is why people push OGG...
> > 4. Container formats: AVI is good, but... It does not support SRT
> > subtitles, e.g. you can't package them into the same file. Ideally, we'd
> > like to package together with the video stream several languages in text
> > and audio streams, and let the user choose. Again, if we stick to free
> > containers, then it looks like we have Matroska (mkv), OGG, and OGM.
> > Which to choose? OGG is the most widespread, but subtitle support is bad
> > (unless you package it into Annodex, which looks complicated). OGM
> > works, but reportedly is a 'crude hack' that nobedy will remember in a
> > few years. I tried the Matroska and it worked with the AVI file
> > (put .AVI and .SRT inside) but failed to work with Theora/OGG.
>
> All the video players I used loaded subtitles automatically when
> subtitles file name was the same as the video file name and they were
> located in the same directory.
> So no container was needed.
> Of course there can be players that can't load subtitles that way. Is
> it so on Linux?
Linux is OK - Totem loaded SRT just fine. I just want to clarify - I
absolutely agree that our main project aim will be to produce and
distribute the SRT files. Packaging them together with the video can
just be a fringe, well, some day in the future. In an ideal world we
would also have someone READ the text in Russian (etc.)
It is simply more convenient, especially when there are several
languages. Maybe at some point we'll also produce and distribute a DVD
with the translated lectures. Again, SOME day in the future :)
Cheers,
--A
>> What do you mean by splitting the video into time periods? Why would
>> you use it?
>
> Oh, I mean the actual subtitle frames. You made 1012 of those - and
> it's actually very good! I tried doing that once and it was so
> time-consuming. Did you just manually decide where each frame
> stops/starts? Or is there a software that can do that, maybe by sound
> levels?
It was really a manual and time-consuming process. You should also
constantly think about the best way to break speech into frames.
I don't know of any software that can help. However such software may
exists.
> No, there aren't any _technical_ problems on Linux either :) I
> installed Mp3 and DIVX support - shhh - don't tell anyone. It is more
> an issue of using FREE technology and being SAFE from litigation.
>
> You know why Linux does not ship MP3 and DIVX? Microsoft does that,
> and it was sued for $1.5 billion dollars for "patent infringements",
> see, for example, here:
>
> http://www.iht.com/articles/2007/02/23/business/web-0223microsoft.php
>
> That is why people push OGG...
But almost everyone uses mp3 anyway ;)
However I see your point here.
Do we infringe anybody's rights creating subtitles for video encoded
with proprietary codecs?
> Linux is OK - Totem loaded SRT just fine. I just want to clarify - I
> absolutely agree that our main project aim will be to produce and
> distribute the SRT files. Packaging them together with the video can
> just be a fringe, well, some day in the future. In an ideal world we
> would also have someone READ the text in Russian (etc.)
Hmm... Sounds really interesting. Who knows?
> It is simply more convenient, especially when there are several
> languages. Maybe at some point we'll also produce and distribute a DVD
> with the translated lectures. Again, SOME day in the future :)
There's a lot of work to do before that :)
But that would be really nice.
Regards,
Sergey Khenkin