@
matthuisman subs now work, thank you very much for that
There is one problem however, which I hate to bring up since you went to the trouble for subs in the first place, and it appears to be in the pycaption subsystem, so I doubt you can do much about it
I took Southpark S23 E02 (band in china) as a sample.
the subs displayed onscreen for opening dialog are
[ Humming ]
Stan! Family meeting!
Nobody cares about that!
Come on!
now looking at the srt file converted by the addon
25
00:00:37,000 --> 00:00:42,000
[ Humming ]
26
00:00:42,000 --> 00:00:45,000
Stan! Family meeting!
27
00:00:42,000 --> 00:00:45,000
Get downstairs!
28
00:00:45,000 --> 00:00:45,000
I'm writing a song, Dad.
29
00:00:45,000 --> 00:00:46,000
Nobody cares about that!
30
00:00:46,000 --> 00:00:49,000
Come on!
so it appears lines 27 and 28 were skipped and looking at the timestamps it's easy to see why. the timings are the same/overlapping
so I took a look at the source sub file on hbo (I mean they could just made life easier by using a more common sub format, sigh)
<p region='pop1' style='basic' begin='00:00:37:01' end='00:00:42:06' tts:origin='35% 84.66%' tts:extent='27.5% 5.33%'>[ Humming ]</p>
<p region='pop1' style='basic' begin='00:00:42:08' end='00:00:45:00' tts:origin='37.5% 79.33%' tts:extent='52.5% 5.33%'>Stan! Family meeting!</p>
<p region='pop2' style='basic' begin='00:00:42:08' end='00:00:45:00' tts:origin='45% 84.66%' tts:extent='37.5% 5.33%'>Get downstairs!</p>
<p region='pop1' style='basic' begin='00:00:45:02' end='00:00:45:17' tts:origin='10% 84.66%' tts:extent='60% 5.33%'>I'm writing a song, Dad.</p>
<p region='pop1' style='basic' begin='00:00:45:19' end='00:00:46:19' tts:origin='30% 84.66%' tts:extent='60% 5.33%'>Nobody cares about that!</p>
<p region='pop1' style='basic' begin='00:00:46:21' end='00:00:49:19' tts:origin='55% 79.33%' tts:extent='20% 5.33%'>Come on!</p>
and the timings are the same here.
so I went on a bit of a google trip, and came across
https://github.com/yanncoupin/stl2srt
this appears to be another python conversion utility.
this utility appears to merge those rows with identical timestamps and also using the milliseconds timings that pycaption appears to drop out
so when I throw the XML subs file into that, I get
14
00:00:37,041 --> 00:00:42,250
[ Humming ]
15
00:00:42,333 --> 00:00:45,000
Stan! Family meeting!
Get downstairs!
16
00:00:45,083 --> 00:00:45,709
I'm writing a song, Dad.
17
00:00:45,792 --> 00:00:46,792
Nobody cares about that!
18
00:00:46,875 --> 00:00:49,792
Come on!
I've got big news!
which then appears works perfectly...
so I know you mentioned writing your own custom converter, but perhaps this will help avoid that need?
