Two line Subtitles from Text files ...

PeterWright wrote on 6/25/2008, 4:10 AM
I have done a fair bit of sub-titling, mainly using Regions from Vegas.

I have a current project where I want to use someone else to transcribe the audio content and produce a text document in Notepad which I hope to then import into DVDA and space out the subtitles accordingly.

The problem is, from the Manual it seems that each line of text results in a separate subtitle, whereas I want to have two line subtitles.

Is this possible?

I'd appreciate any help.

Comments

TheHappyFriar wrote on 6/25/2008, 6:02 AM
you can do multi-like subtitles direct in DVDA. I did that & exported that to a .sub file.
here's the results:

0001  00:00:00:00  00:00:59:05  blah blah blah blah	hiyall

khjkhj



0002  00:00:59:05  00:01:58:11  hi how are you



0003  00:01:58:11  00:02:57:17  i am fine



0004  00:02:57:17  00:03:56:23  i love cookies



0005  00:03:56:23  00:04:55:29  :D :| :(


The first two lines are two separate lines of the same subtitle in DVDA. So if you format it like that in the text file (and rename it as a .sub as it doesn't understand that as a .txt file) you should be set.

great opportunity for someone to make a 3rd party vegas subtitle generation program for easy editing. :)
PeterWright wrote on 6/25/2008, 7:39 AM
Yes, thanks HF - I've exported a file like that from DVDA, having previously created the subs as Regions in Vegas.

My problem is that the person doing the transcribing doesn't have DVDA - they'll be listening to the audio and creating a .txt file in Notepad - how to create a line break and two line subtitles?.
TheHappyFriar wrote on 6/25/2008, 9:11 AM
just lay it out like I pasted.
0001  00:00:00:00  00:00:59:05  blah blah blah blah	hiyall

khjkhj


0001 = page # of subtitle.
00:00:00:00 = start time of subtitle
00:00:59:05 = end time of subtitle
blah blah blah blah hiyall = 1st line of text
khjkhj = 2nd line of text

0002 would start the 2nd page of subtitles
PeterWright wrote on 6/26/2008, 6:52 AM
Yes, but the typist was only going to transcribe the words whilst listening to the audio, not worry about timecode ...

BUT ANYWAY

This morning I found that my typist friend can't do it for at least 10 days, so I'm doing it myself (almost two and a half hours of Presentations!)

AND

I've made a MAJOR discovery - for me at least!

It's much better to do subtitles in DVDArchitect.

WHY?

Previously I was creating Regions in Vegas - the Regions name was the words spoken in that segment and this was exported as a .sub file to import into DVDA

BUT

At the time you typed the Region names after listening to the audio, you didn't know what it would look like in the DVD - all you see is the Region name in small text above the Vegas timeline - then in DVDA you still have to choose fonts, size, colours, backgrounds, transparencies etc.

SO

If you render the Video project as MPG2 and ac3, then import into DVDArchitect, you can open a timeline, add a Subtitle track, listen to the audio and type straight into Subtitle Events, and

SEE WHAT IT WILL FINISH UP LOOKING LIKE AS YOU TYPE.

Similar to what first drew me to Vegas, the first software I found which could give me live preview via software only, so I could see the results of what I was doing as I did it.


Going for a lie down.

TheHappyFriar wrote on 6/26/2008, 7:04 AM
Yes, but the typist was only going to transcribe the words whilst listening to the audio, not worry about timecode ...

that's why I made the comment someone could make quick $$ making a DVDA sub-generator program. it takes a 2nd person to re-edit the transcription & put in the sequence/time info. Would be easier to have the app figure out how long it's taking you to type & then add this info right in to the file so if you're transcribing at a 1:1 ratio (every time you type it's talk, no rewinding, etc) it automatically does it for you. :D
MPM wrote on 6/26/2008, 8:12 AM
Totally FWIW...

DVDA can be a bit sluggish playing from the timeline to check sub timing - it can start slow, & sometimes not want to catch up, even after stopping/starting a few times. Clearing undo history seems to help, but if I were timing every event, rather than syncing a text file, I think I'd check out using DGIndex & VFAPI, substituting the fake avi file from VFAPI for the imported mpg2, & wav rather than ac3. I haven't tried it in DVDA, but am basing that on the improvements I've noticed in Vegas with VFAPI vs the orig. mpg2 on the timeline.

As far as a DVDA sub generator goes, right now there are several transcription programs, plus several converters. All one has to do is wind up with DVD Studio Pro format sub text files & import. While something DVDA specific would be nice of course, DVDA hasn't been that popular in the mainstream IMHO for the coders writing this stuff to even think about. I'd suggest that if enough people wanted & showed that they wanted DVDA support, if only thru a dedicated thread, they might be persuaded to add whatever's necessary to existing code so it worked better with DVDA.

As far as the separate line problem etc... I would guess that as the typist was listening & typing, using a transcription program would produce times that, even if inaccurate, would form a much better base to build on having time codes at least. And if nothing else would be easier to start stop than using notepad & a viewer. If the text files wind up with single line subs only, Subtitle Workshop (plus probably other sub software) will easily combine or split lines.

Anyway, maybe (hopefully) something there will give a few ideas?
TheHappyFriar wrote on 6/26/2008, 8:54 AM
this is something that could be made by jetdv or the vasst team pretty easily i'm sure.

the time a wav plays should be = to the actual length of the fine. my wife's transcription software seems accurate. infact, instead of the screens like DVDA does, if TV style subtitles could be done that would be better. The kind where it scrolls, doesn't just appear/dissapear. odds are that would need to be a separate track of video though. :/
MPM wrote on 6/26/2008, 9:58 AM
Oh, I agree...

If I was a coder I'd likely take a look at the open source alternatives first, to make sure I could add something, and to keep from re-inventing the wheel. Then I'd work on making my project compatible with the widest number of apps to increase my potential user base. OTOH if all I could accomplish was a work-a-round so DVDA users didn't have to use something already available, not sure I'd go any further.

At any rate, Google & you'll get a bunch of hits on transcribing or creating subs or captioning. This doc talks about how to do it using my favorite, Subtitle Workshop:
http://www.calvin.edu/admin/av/Transcribing%20with%20Subtitles%20Workshop%20QuickStart%20Guide.doc

And Videohelp has a few in their tools section (the first one if you have text without timings):
http://www.videohelp.com/tools/DivXLand_Subtitler
http://www.videohelp.com/tools/Jubler
http://www.videohelp.com/tools/SubMagic

Getting the text down is critical, getting the timings is next, and getting them into your DVDs is almost trivial. The different text formats are mostly just a style of notation for times and line breaks. Here's a (hopelessly) corny example of the MAC DVD Studio Pro format DVDA likes [reflection of my mood :-) ]

01:31:58:07 01:32:02:04 ♫ There's a place in the sun
01:32:02:04 01:32:06:04 ♫ And before my life is done
01:32:06:04 01:32:11:30 ♫ Gotta find me<P>a place in the sun

In the common srt format:
1646
01:31:58,218 --> 01:32:02,121
♫ There's a place in the sun

1647
01:32:02,122 --> 01:32:06,125
♫ And before my life is done

1648
01:32:06,126 --> 01:32:11,998
♫ Gotta find me
a place in the sun

Or you can convert to scc for addition to the mpg2 itself as CC (using Encore or DVDLab Pro), but it's all the same text & timings - just written differently.

Edit: Oops... Sorry, while it looked OK entering the characters, *after* posting the music notes turned into ♫ - apologies.