Forum Discussion
Short closed caption are not being recognized by Storyline 360
I use a separate workflow to generate clean .srt caption files for my instructional videos. Ideally, I then want to be able to quickly add each .srt to the corresponding videos in Storyline and be done. But I've noticed that when I play the course afterwards, some captions are being dropped. It looks like this happens when the caption is shorter.
Please see the following Storyline file to see an example. I used short_test.srt for slide 1 and short_test_2.srt for slide 2. The two .srt files contain the same captions, but the timestamps are shifted to see if it has any effect (it doesn't).
Notice that the first 2 captions ("To give" and "us a specific goal") are dropped, and only the third long one is captured.
Expected behavior: All captions, whether short or long, are recognized by Storyline upon closed caption import, so that there are no gaps in closed captions.
It's very time-consuming to comb through a large volume of instructional videos (courses are hours long) and we don't have the capacity to do this every time we create instructional content. This ultimately hurts our users relying on closed captions, who get confused / miss out on important context during the instructional videos due to the dropped captions.
Note: It looks like this doesn't accept .srt file uploads, so I'll paste them below here. You can copy-paste these into a text file and give it extension .srt to recreate.
short_test.srt:
1
00:00:00,346 --> 00:00:00,713
To give
2
00:00:00,713 --> 00:00:00,747
us a specific goal
3
00:00:00,747 --> 00:00:04,117
as to what to include in our baker report
we're creating from scratch,
short_test_2.srt:
1
00:00:00,000 --> 00:00:00,300
To give
2
00:00:00,300 --> 00:00:00,350
us a specific goal
3
00:00:00,350 --> 00:00:03,300
as to what to include in our baker report
we're creating from scratch,
6 Replies
rachelatkaiser I'm using Premiere Pro to generate the caption files off of the instructional video, so my generated .srt have real timings - Premiere Pro captures them according to the timing of the video's audio. But good catch on these two particular examples having unrealistically short durations - I took a further look and it looks like I had artificially shortened these two .srt timing markers to make the issue more obvious for investigation.
To see something with no alterations, please see the below .srt excerpt (which is taken from the actual instructional video file). I can't post the entire .srt file in this public forum but the excerpt below has the actual original timings. This shows that it takes roughly 1.4 seconds to read through "To give / us a specific goal", which makes more sense.
1_Sequence_excerpt.srt
15
00:00:46,346 --> 00:00:46,713
To give
16
00:00:46,713 --> 00:00:47,747
us a specific goal
17
00:00:47,747 --> 00:00:51,117
as to what to include in our baker report
we're creating from scratch,
To see what happens when imported, I've attached a Storyline file showing what happens when you import this .srt to a video. You'll see that it drops the initial "To give" caption. (To observe, click on the video, and in Video Tools Options click Manage Captions).
This is just one example, but I found other random drops throughout my instructional videos. The only common thread I could see between the dropped captions were that they were on the short side.
Gotcha. I see what you mean in this new file you shared. I think james598patton is right and this seems to be a Storyline-specific problem. I think it would be worth reporting it to them (even if it is a known issue, it's good to add to the pile). James's suggestion sounds good. I'm really sorry if you do have to adjust all these captions.
Thank you rachelatkaiser for the follow up, and I think that's a good idea. EricSantos , could you please add this to the discussion thread/ticket that already exists for this known issue? Also, could you share a link to the original thread/ticket so I can follow it for updates?
My understanding of Articulate's values is that you make users' time and productivity a priority, which is why you're putting a lot of effort into AI features. Given that, I'd like to make an argument to prioritize this fix: getting this captions issue fixed will have exponential effect on productivity for accessible courses and shouldn't take nearly as many engineering resources as a new AI feature. With this current issue, my workflow for getting closed captions added to a single course takes hours. When fixed, the workflow will take minutes.
DLearning How are you creating these caption files? The time intervals of the first two captions in both files seem crazy short to me. It personally takes me 1.5-2 seconds to say, "To give us a specific goal." It seems a little unrealistic to not have hit the one-second mark by the third caption in both files. I wonder if something may be a little off with the captions.
