Captions and subtitles

[Viggo]

I just like to reference this video about why you may consider subtitles and captions:
https://youtu.be/NFHDHcs4BvQ

For YouTube videos, it is a good idea to have separate SRT-files or similar text separate to the graphical video itself, because then the text can be translated to all native languages for the viewer of a YouTube video.

In the past couple of days I have read a bit about automatic generation of subtitles from voice and have tried to make subtitles to a couple of my youtube videos. I have also searched here for information. I have seen that Elusian made a tool with a link to here:

http://elusien.co.uk/shotcut/SRT/

I have not tried it yet.

As far as I can see, it helps to edit subtitles in the Shutcut framework.

And the Vrew should be downloaded free software to make subtitles:
https://vrew.voyagerx.com/en/

I have not tried this program, because I am a bit concerned about security. It seems to be developed in South Korea and perhaps it just used the Google machine for transcription. In the conditions you shall accept that the system sends audio to Google.

I have tried to make subscriptions to a few of my videos on my YouTube channel. YouTube do have a manual editor itself, at it do also make transcriptions for you on the basis of the voice. However for my 41 public videos YouTube only did that for one of them. And I guess it is because there is some technical terms, that it not easy to recognize, and I got a foreign accent. So I guess YouTube already considered the transcription result to be bad.

I have then tried to make Google docs make transcription via microphone input. I think the result was quite bad, and it would be better to do the job manually. Then I have tried the free version of otter.ei. It do it better that google, but you still got a lot of corrections to do.

There is a lot of other transcription software and services when you search on internet. So it would be nice to know if some others got experience with them.

What kind of experience you have got with auto-generated transcriptions and captions?

[Elusien]

I have tried auto-generation but with varying success. I generate the text for my tutorials by hand using a simple text editor (Notepad++), since I need a script for the vocal description anyway. Then I feed this into my SRT creator. The .srt file can be fed fully into Google Translate, or Deepl and a foreign language version can be created.

See below for more information:

Subtitle Displayer (subrip - .srt files)

Some time ago I released a Subrip Generator on my website to enable people to create a subtitle file relatively easily. This can be found here: SubRip SRT Subtitles Generator I have now created a SubRip Displayer, which is a fairly simple HTML file that displays in a webpage cues from a Subrip (.srt) file according to their timings. This can be found here: Resources for Shotcut You can us this HTML file as a “browser source” in OBS Studio, which then enables you to record the subtitles to a vi…

and here:

SubRip (.srt) Subtitle Generator

Why have subtitles for your videos? According to WHO about half a billion people have disability hearing loss (about 1 person in 16). Many people can’t or don’t turn on audio (e.g. 85% of facebook videos are watched on mute). Research has shown many people prefer to watch videos with subtitles switched on. Not everyone speaks your language (English 25%, Chinese 20%, Spanish 10%, Arabic 5%). Viewers are more engaged. Research has shown that on average non-subtitled videos were watched to 66% co…

See also the thread posted here:

Voiceover Practice

I find it incredibly hard to voice my videos, but I don’t want to procrastinate from this situation, I want to master in it. Are there any suggestions to how I can improve my skill, I have saw people like @jonray , EPIC VOICE GUY (in the forum) to speak very calm and well in their video, I wonder how are they so patient. I have borrowed a good mic from my friend to practice, as I will buy a new one when I will master voiceover. If I get better on it, I will start making videos again, having tho…

[Viggo]

Thank you for the input, Elusien.

I do not think, that you need to translate your SRT-file unless you know the language that well, so you can correct the errors of the machines translation. YouTube provide the user looking at a video to translate the subtitles to own language despite any effort by the creator to do a translation. I do not know about the other video providers.

For new videos I think I will follow your suggestion to make the written text first, because I think this will be a faster way. But sometimes I feel that my voice over can suffer in authenticity when I read a text, so I need some practice there.

But I then got some problem with the videos already made without subtitles. Some of them may need an update anyway, and could be replaced. However I am a bit worried, that YouTube may reduce some relevance status of newer videos produced, and you will have to start from scratch there.

[Elusien]

@Viggo

But sometimes I feel that my voice over can suffer in authenticity when I read a text, so I need some practice there.

I used to give presentations on computer-related subjects in front of large audiences. I would rehearse several times beforehand. Rarely did my speech follow exactly what I’d written as I liked to add small “asides” and interact with the audience. But if I were to create subtitles it would only take a small amount of editing of the original text to get the final version.

Don’t forget you can always edit the audio in Shotcut as well as the video. If a phrase/sentence doesn’t sound right you can always re-record that bit and substitute just that small audio clip.

Just make sure that creating your subtitles is the last step as you do not want to modify the video and then have to edit the .srt file by hand.