Date
1 - 9 of 9
Captioning question #Captioning
|
I received this question from a group setting up a virtual platform to offer courses:
I would love to ask your opinion about our online platform development.
We are about to implement captioning and the developers have given us two options:
‘Real-time’ captions, which show each word as it’s spoken, or
‘Final version’ captions, which compose a phrase or sentence after the person pauses. It’s more accurate apparently.
What would you recommend? Thank you, Sara Oser |
|
|
|
I prefer real time. I can hear, but miss a lot of words or phrases. Unless the captions are as close to the spoken word as possible they do me no good. The ASR has improved greatly during this past year, so accuracy hasn’t been a problem. Delay is always a problem.
toggle quoted message
Show quoted text
Thanks for asking. Carol
|
|
|
|
Mary Jarrett
I agree.
toggle quoted message
Show quoted text
Mary Jarrett MWjarrett@... (904) 635-6660 cell On Aug 5, 2022, at 4:07 PM, Carol in Boston <carolagate34@...> wrote:
|
|
|
|
Jack Clevenger, AZ
All, I prefer real time and allow for the mistakes that usually are minor. Having to wait for for That is my take."accurate" captioning really sets me back being at live meetings.
toggle quoted message
Show quoted text
|
|
|
|
Hi, Sara
I heard of captions that are not verbatim but display the essence of the conversation. I have not seen this in practice. It would be interesting to find a video that uses this method. Most people expect verbatim and may become confused seeing something different. -- Alan Katsura akatsura@... Moderator HLAA Tech Group |
|
|
|
I prefer final version. My experience is the delay is small enough to let me carry on a conversation. I’ve also seen captioning from Olelo where it is in realtime, but then it backs up and corrects.
|
|
|
|
After seeing Ed's post, I think I understand what "final-version" captioning is. It is verbatim but in small phrases. I can see where it may be more accurate and possibly easier to read since there is less likelihood of having to reread the captioner's corrections. Since normally there is a latency between the speaker/video and captions, accuracy is more important than speed. Does anyone use "final-version" captioning in their meetings? I will post a poll.
-- Alan Katsura akatsura@... Moderator HLAA Tech Group |
|
|
|
I just want to point out that preferences probably depend on the quality. ASR got a bad name because it was so often wrong. It has improved enormously. If anyone uses ASR with their Innocaption app you know how good it can be. CART is only as good as the provider, but it can never be as fast. The variation among CART providers varies just as the variation among ASR providers vary. But if you use a good ASR provider you know what you are getting. If you know your CART provider, you also so. But in most situations you ask for CART and have no way of knowing how good the provider will be.
toggle quoted message
Show quoted text
So ASR is less of a gamble and it has the plus of speed.
|
|
|
|
Arlene Romoff, NJ
Yes - I agree with this totally. That has been my experience as well. Arlene On Aug 6, 2022, at 11:45 AM, Carol in Boston <carolagate34@...> wrote:
|
|
|