how to extract closed captions from a wmv file

[scarecroe]

I have a wmv file authored by unknown means in 2017 that has embedded (not burned-in) closed captions (not subtitles).

I'm trying to extract the text from the closed captions to save as an srt or some other sidecar file.

I found CCExtractor, but it doesn't work with wmv files. Does anyone know of such a program?

[ProWo]

Use clever Ffmpeg-GUI.
Load your wmv, click main page, click extract closed captions, click extract.
Wait until has finished.
You will find the cc as srt file in your target folder.

[scarecroe]

This almost worked. I loaded up the file and went to the Main Page button. Under extract, there's a button for EIA 608/708 which is the 608/708 compliant cc data. However, after attempting to extract, it analyzes for a few minutes and then says that no closed captions were found. I tried to make it look at additional captions indexes, but increasing the data field didn't seem to work either.

[ProWo]

If you load the file, the cc are shown in the stream list?
If so, use the extract streams button.

[scarecroe]

When I load the file, it shows two streams:

Stream 0:0, Audio, wmapro, 2 Channels, 440 kb/s
Stream 0:1, Video, wmv3, 1920 x 1080, 1.78, 29.97 fps

I assumed the cc were embedded in one of those. They don't show up in VLC or SMPlayer; I can only get them to show up in Windows Media Player (wmplayer).

[El Heggunte]

I have a wmv file authored by unknown means in 2017 that has embedded (not burned-in) closed captions (not subtitles).

I'm trying to extract the text from the closed captions to save as an srt or some other sidecar file.

I found CCExtractor, but it doesn't work with wmv files. Does anyone know of such a program?

Give a try to GraphStudioNext. IF the playback graph detects a closed-caption stream, THEN this should be "extractable" (DUMPed) to a standalone text file.

[scarecroe]

It doesn't look like GraphStudioNext found a CC stream:


[Attachment 83469 - Click to enlarge]

It's definitely there though, toggleable in Windows Media Player (and not a sidecar file).

[ProWo]

When I load the file, it shows two streams:

Stream 0:0, Audio, wmapro, 2 Channels, 440 kb/s
Stream 0:1, Video, wmv3, 1920 x 1080, 1.78, 29.97 fps

I assumed the cc were embedded in one of those. They don't show up in VLC or SMPlayer; I can only get them to show up in Windows Media Player (wmplayer).

Try this:
Reload your file in clever FFmpeg-GUI, click main page, click multiplex.
Select the wmv3 stream only (on the left), the audio stream should remain unselected.
Set wmv as container and click multiplex.
Then return to the source screen, load the new created wmv file and repeat the EIA 608/708 extraction operation.

EDIT:
If this fails too, then mux the videostream only to mkv and try this mkv in CCExtractor.

[El Heggunte]

It doesn't look like GraphStudioNext found a CC stream:


[Attachment 83469 - Click to enlarge]

It's definitely there though, toggleable in Windows Media Player (and not a sidecar file).

That's weird
What's the name of the first filter (demuxer)?
It's the default Windows ASF demuxer, OR it is LAV Splitter source
LAV Splitter does not support the ASF "script commands", afaik.
But the "normal" ASF demuxer does.

[netmask56]

As a try out load it into mkvtoolnix. It should display the subs as separate tracks along with everything else. Make sure you have the very latest version of mkvtoolnix. Untick what you don't want and remux. Doesn't take long, a matter of minutes.

[El Heggunte]

As a try out load it into mkvtoolnix. It should display the subs as separate tracks along with everything else. Make sure you have the very latest version of mkvtoolnix.

AFAIR, mkvtoolnix doesn't support ASF/WMV files. And not because Mosu doesn't know how to handle them, but because he doesn't want to be sued by the Micro$oft lawyers.

[poisondeathray]

Did you try Windows Media File Editor , file=>export header file ?

[scarecroe]

Did you try Windows Media File Editor , file=>export header file ?

This worked! This exported to a txt file, but the format is unfamiliar to me:

<Script Type="TEXT"
Command="&lt;center&gt;&lt;div align=&quot;center&quot;&gt;&lt;font face=&quot;Arial&quot; size=&quot;3&quot; color=&quot;#ffffff&quot;&gt;&gt;&gt; All of the people were there&lt;br&gt;"
Time="1479810000" />

I'm assuming I can convert this into an srt format, but the converters I've used so far don't recognize it. I could probably manually do it, but I can't figure out the time signatures.

[jagabo]

Time="1479810000", I can't figure out the time signatures.

My guess would be microseconds. So 1479810000 microseconds would be 1479.81 seconds, or 24 minutes and 39.81 seconds. Does that match the dialog?