વિષયના પાનાઓ: [1 2] > |
What went wrong here? દોર પોસ્ટ કરનાર: Edward Potter
|
Edward Potter યૂનાઇટેડ સ્ટેટસ્ Local time: 19:04 સભ્ય (2003) સ્પેનીશ થી અંગ્રેજી + ... |
Even though I am not fluent in Spanish and cannot see what the problem is, I can tell that interlingual live subtitling is an incredibly difficult job, and errors will find their way into the script whether you want it or not — as it often happens with subtitles in live news reports.
[Edited at 2015-01-21 13:54 GMT] | | |
Looks like a technical issue | Jan 21, 2015 |
The process failed, it looks like. I see parts and bits of otherwise pretty correct phrases. The "equipment" must be at fault, but I have no idea how subtitling works, so I could be completely wrong here. | | |
What went wrong here ? | Jan 21, 2015 |
I don't know Spanish .But to me,it looks like a Time coding issue . I have a suggestion . Upload this video in Amara site .( If you don't use Amara ,sign up ) and then redo time coding in Amara's software . Looks like the English time coding is also wrong . You may have to do subtitle synchronization for Spanish from start to finish . In one go finish the fresh time coding and then make fine adjustments later . It is easy to time code accurately in Amara software . | |
|
|
Language course? | Jan 21, 2015 |
Looks like the fast way to learn English-Spanish. | | |
Juan Jacob મેક્સિકો Local time: 17:04 ફ્રેન્ચ થી સ્પેનીશ + ...
...subtitling, no doubt.
Never heard about "live" subtitling made by humans.
Impossible, IMHO. | | |
Juan Jacob wrote:
...subtitling, no doubt.
Never heard about "live" subtitling made by humans.
Impossible, IMHO.
Have you never heard of interlingual live subtitling or just live subtitling? If it is the former, I can understand, as it is rarely used due to the difficulty and issues involved; if it is the latter, then you probably have never worked in any professional subtitling tool, as nearly 100% of them have functionality for live subtitling. Either way, you can read brief descriptions of these two types of subtitling on Wikipedia:
http://en.wikipedia.org/wiki/Subtitle_(captioning)#Live
http://en.wikipedia.org/wiki/Subtitle_(captioning)#Real-time | | |
|
|
Juan Jacob wrote:
...subtitling, no doubt.
Never heard about "live" subtitling made by humans.
Impossible, IMHO.
I have interpreted at conferences where someone typed up the given speech as it was being given so that, when it was finished, people could have a copy. At the same time, I translated it into English. I think there was a lag of 5 minutes while we waited for the printer to do its job. I won't say it was my finest translation, but it was not bad and served a purpose until the text could be polished. | | |
Voice subtitling | Jan 22, 2015 |
Is it possible to subtitle with speech recognition software? That would go faster. | | |
Robert Rietvelt wrote:
Is it possible to subtitle with speech recognition software? That would go faster.
Yes, of course. In WinCAPS Qu4ntum, for example, there is speech input functionality that uses integrated Dragon NaturallySpeaking for picking up your voice and converting it into text in subtitles. However, when it comes to speech input, there is one hurdle — homophones. Before starting speech-inputting text, you need to fine-tune your homophone substitution table (such substitutions are called "House Styles"), so that you don't have to struggle with words that are pronounced the same but have different meanings. | | |
I hope not, using YouTube's contrivance | Jan 22, 2015 |
Robert Rietvelt wrote:
Is it possible to subtitle with speech recognition software? That would go faster.
If you go to YouTube, you may turn on the voice recognition-based subtitling and see the results. If that is not hilarious enough, turn on the automatic translation into the language of your choice.
For some reason that I cannot identify, the most hysterical ones seem to be obtained from dermatology clinics institutional videos. Anyone willing to earn an Ig Nobel Prize could try doing scientific research on why. | |
|
|
José Henrique Lamensdorf wrote:
Robert Rietvelt wrote:
Is it possible to subtitle with speech recognition software? That would go faster.
If you go to YouTube, you may turn on the voice recognition-based subtitling and see the results. If that is not hilarious enough, turn on the automatic translation into the language of your choice.
For some reason that I cannot identify, the most hysterical ones seem to be obtained from dermatology clinics institutional videos. Anyone willing to earn an Ig Nobel Prize could try doing scientific research on why.
José, I think you are confusing speech recognition for voice recognition. Speech recognition software is used for converting your own speech to text, whereas voice recognition software is used for converting dialogue or narrative in a video to text. The first works very well, whereas the latter produces rather poor results, as you described.
[Edited at 2015-01-22 11:59 GMT] | | |
Correct, Max | Jan 22, 2015 |
Max Deryagin wrote:
José, I think you are confusing speech recognition for voice recognition. Speech recognition software is used for converting your own speech to text, whereas voice recognition software is used for converting dialogue or narrative in a video to text. The first works very well, whereas the latter produces rather poor results, as you described.
It's good that there is speech/voice recognition differentiation. Thanks for clarifying the terms.
Some smart-aleck wannabe video translators think of using Dragon NS to transcribe videos, then shoot text through Google Translate, and have it time-spotted by some automatic contrivance that merely distributes the playing time proportionately to the number of characters.
The result should be much worse than the video at hand here.
Some colleagues who used the late ViaVoice said that it took them months to train the software to recognize their speech, as well as to get used to a somewhat boring (for human standards) way of speaking that would optimize results. | | |
[Edited at 2015-01-22 17:45 GMT] | | |
વિષયના પાનાઓ: [1 2] > |