You are not logged in.

  • "Diana" is female
  • "Diana" started this thread

Posts: 673

Date of registration: Jul 16th 2011

Language Team: Italian

Focus Group: Translation Proofreader
Language Coordinator

Location: Italy

Thanks: 35344 / 100

  • Send private message

1

Monday, January 23rd 2012, 10:19am

how to synchronize a translation

Hi all! I already asked this question elsewhere to some of you ^^

We have to synchronize some videos whose timestamp was adjusted after the translations, but so far we did it copying and pasting string by string. Is there a tool to help us in doing this?
Any other idea will be really appreciated. :)

This post has been edited 1 times, last edit by "Diana" (Jan 24th 2012, 1:11am)


58 guests thanked already.

Posts: 3,012

Date of registration: May 24th 2011

Language Team: English

Focus Group: Timestamp Adjustment Team
Public Relations Group

Location: Malmö, Sweden

Thanks: 90516 / 2397

  • Send private message

2

Monday, January 23rd 2012, 5:52pm

Hi Diana!! :smiley:
I am not sure I understand what you mean by copying string by string? In dotsub's translation page for that video,don't you have the Italian translation there already but only that not corresponding fully to the English strings above? And only in some cases you have to cut some text from one string and pass it to the next one or the one before?
Signature from »nomada« Click to know how to: - Receive email notifications - Join the LTI biweekly Meetings

54 guests thanked already.
  • "Diana" is female
  • "Diana" started this thread

Posts: 673

Date of registration: Jul 16th 2011

Language Team: Italian

Focus Group: Translation Proofreader
Language Coordinator

Location: Italy

Thanks: 35344 / 100

  • Send private message

3

Tuesday, January 24th 2012, 12:41am

Hi! :wave:
yes, the translation is there but it does not correspond at all. Only the first 10 lines correspond. That means, for example, more than 1450 strings to copy/paste just in Social Pathology. And the clicks are likely to multiply, because also the length of each string changed. It's not the only case. We already went through MF, Future by Design and the Presentation by PJ + some of the shorter ones... and all of them were messy like this. We still have Social Pathology and Addendum to do! The fact is that translating, or reviewing, keeps the mind busy on something more intellectual that asks some creativity. While synchronizing is just boring and repetitive :wacko: ah, the text is not in the previous or next string, most of the strings are several minutes far from the corresponding text. And the new layout in dotsub doesn't help at all with this task. That's why I'm asking for help again!!!
:smiley:
Copying/pasting string by string of a translation made off line it's much easier than doing the synchro of a translation... if you see the difference. I did both and I'd be ready to waste time to copy/paste a translation made off line, but not so much for the syncrho. It is too many clicks and scrolls, and a pain if the video is longer than 20-30 minutes.

Hugs
:)

This post has been edited 1 times, last edit by "Diana" (Jan 24th 2012, 1:13am)


39 guests thanked already.
  • "Diana" is female
  • "Diana" started this thread

Posts: 673

Date of registration: Jul 16th 2011

Language Team: Italian

Focus Group: Translation Proofreader
Language Coordinator

Location: Italy

Thanks: 35344 / 100

  • Send private message

4

Tuesday, January 24th 2012, 12:55am

well, keeping on with my example, in social pathology we "lost" the 10% (I know it's still there!). But it means about 140 strings, this is not an average but gives you an idea how how far the text is from the corresponding strings: somewhere it's 140 strings away 8o because with the timestamp adjustment the text gets squeezed at the beginning and all the strings at the end remain empty. :|

42 guests thanked already.

Posts: 3,012

Date of registration: May 24th 2011

Language Team: English

Focus Group: Timestamp Adjustment Team
Public Relations Group

Location: Malmö, Sweden

Thanks: 90516 / 2397

  • Send private message

5

Tuesday, January 24th 2012, 3:58am

Hi Diana!!I tried to sync the Portuguese translation too to see how hard it could become. Yes, it is lousy. The text becomes more and more separate and it becomes hard to copy past it.

Here is a thread where Georgy presented us with a new tool a while ago: Subtitle Editors

Found the thread with the translations, I think that in a state prior to the timeshifting, not sure. I am a bit confused now :embarrassed:
[English] Peter Joseph: Social Pathology [Done]

Let me study this a while and I get back to you. Maybe Bruno has a better view of what can be done.

Hugs : )
Signature from »nomada« Click to know how to: - Receive email notifications - Join the LTI biweekly Meetings

42 guests thanked already.

Posts: 3,012

Date of registration: May 24th 2011

Language Team: English

Focus Group: Timestamp Adjustment Team
Public Relations Group

Location: Malmö, Sweden

Thanks: 90516 / 2397

  • Send private message

6

Tuesday, January 24th 2012, 4:28am

Ok. Found something. ^^

I went to the forum post where Bruno saved all the translations before applying the English transcript timeshifted times to them. So they are with the original times still. I downloaded the Portuguese translation from that forum post and the English transcript from the Social Pathology dotsub working location . In Georgy's tool REFRAGMENT I uploaded the "Reference SRT" as the English SRt and the "Refragment SRT" as the Portuguese SRT, and clicked "Upload files and view".
The tool gives me lines where to insert the Portuguese text that is in the green box bellow, through copy-past, accordingly to the English reference. When I get to the end of the group of lines in that page, I click "refragment" and the tool gives me the next set of strings... until the document is reviewed. This is much simpler thaN using dotsub, after I tried both ways.

If you wish to pause and continue reviewing later, copy the project number, which you can insert in the front page of the tool where it says "Existing Project" and clicking on "View", instead of uploading the SRTs and starting all over again.

My project number is 85246102892203. I tried it and it takes me to the exact point where I left the works ;)

Hope this suits all the other proofreaders :thumbsup: Throw some feedback on us, please!!

And THANK YOU GEORGY!!! YOUR THE BEST!!! :love:

Editing: I don't copy-past but Cut and past, following Georgy's advice, the text from the green box to the lines/string boxes, because itmakes iteasier for us to not mistaken an also find easily the text that we need next.
Signature from »nomada« Click to know how to: - Receive email notifications - Join the LTI biweekly Meetings

2 registered users and 42 guests thanked already.

Users who thanked for this post:

brunodc, georgyvlad

  • "Diana" is female
  • "Diana" started this thread

Posts: 673

Date of registration: Jul 16th 2011

Language Team: Italian

Focus Group: Translation Proofreader
Language Coordinator

Location: Italy

Thanks: 35344 / 100

  • Send private message

7

Tuesday, January 24th 2012, 10:00am

Thank you sooo much, Nomada and Georgy!!! :bighug: we'll use the tool and then let you know how it goes ... :smiley:

(we also cut and paste, I'm just to used to say copy!!) Thanks again for doing some experimentations :thumbsup:


Hugs and kisses :kiss:

1 registered user and 43 guests thanked already.

Users who thanked for this post:

georgyvlad

  • "georgyvlad" is male

Posts: 52

Date of registration: Jul 24th 2011

Language Team: Bulgarian

Focus Group: English Transcriber
English Proofreader
Translator
Translation Proofreader
LTI Development Group

Location: Mountain View, CA, USA

Thanks: 1863 / 34

  • Send private message

8

Tuesday, January 24th 2012, 3:56pm

Great job, Nomada! Yes, the "frag" (refragment) tool was designed to do just that and it works like magic (or rather computer science) :-)

Diana, go ahead and just use the tool, just like Nomada described it. The tool is pretty intuitive and has plenty of directions, too.

The more people we educate in thinking about the "frag", "diff" and "tmshift" tools, the more we'll keep people focused on the fun work and let computers do the boring part :-)

Hugs,

Georgy.

2 registered users and 47 guests thanked already.

Users who thanked for this post:

nomada, Diana

  • "Diana" is female
  • "Diana" started this thread

Posts: 673

Date of registration: Jul 16th 2011

Language Team: Italian

Focus Group: Translation Proofreader
Language Coordinator

Location: Italy

Thanks: 35344 / 100

  • Send private message

9

Saturday, January 28th 2012, 4:10pm

Hi!! I just went through the process with a short video and it was much easier than in dotsub :)

What should I click next: Synchronize timestamp or Export refragmented ?

Georgy, please, could we paint the editor box in a lighter green? to have a little more contrast between the text and the background ^^
well, I'd also have a sort of idea to make the tool faster! I'll think of how to put it in words. :whacked:

Hugs :loveya:

1 registered user and 45 guests thanked already.

Users who thanked for this post:

georgyvlad

  • "georgyvlad" is male

Posts: 52

Date of registration: Jul 24th 2011

Language Team: Bulgarian

Focus Group: English Transcriber
English Proofreader
Translator
Translation Proofreader
LTI Development Group

Location: Mountain View, CA, USA

Thanks: 1863 / 34

  • Send private message

10

Saturday, January 28th 2012, 6:15pm

First click Synchronize, then Export - just like the instructions on the page.

Your words are my command - lighter green it will be :-)

I am all ears about the idea...

Hugs,

Georgy.

1 registered user and 40 guests thanked already.

Users who thanked for this post:

Diana

  • "georgyvlad" is male

Posts: 52

Date of registration: Jul 24th 2011

Language Team: Bulgarian

Focus Group: English Transcriber
English Proofreader
Translator
Translation Proofreader
LTI Development Group

Location: Mountain View, CA, USA

Thanks: 1863 / 34

  • Send private message

11

Saturday, January 28th 2012, 6:32pm

Diana,

It is lighter green now. I am even thinking I might do the Synchronize step automatically after the final refragment is done, so I don't confuse the user...

Hugs,

Georgy.

49 guests thanked already.
  • "Diana" is female
  • "Diana" started this thread

Posts: 673

Date of registration: Jul 16th 2011

Language Team: Italian

Focus Group: Translation Proofreader
Language Coordinator

Location: Italy

Thanks: 35344 / 100

  • Send private message

12

Saturday, January 28th 2012, 8:20pm

Hi Georgy,

I just created a new project to check it! and yeah, it's a lighter shade!! Our eyes say "Thank you" :)

And it is fine the two step thing, it's not confusing at all! I asked just to make sure ;)

Hugs

46 guests thanked already.
  • "Diana" is female
  • "Diana" started this thread

Posts: 673

Date of registration: Jul 16th 2011

Language Team: Italian

Focus Group: Translation Proofreader
Language Coordinator

Location: Italy

Thanks: 35344 / 100

  • Send private message

13

Saturday, January 28th 2012, 9:35pm

I imagined a page with two columns, in a column the reference and in the other an editor where (instead of cutting/pasting the text into cells) we "break" the text into rows ... by pointing the pointer where we want to change string and pressing a key (enter or tab). Rows will be in line with the reference strings and the tool would recognize those rows as strings. Hope it's clear and not too stupid, I know nothing about programming.

Then because of the continuity in reading, it should be more relaxing to have the reference in the column on the left and the empty part on the right, at least for those reading from left to right. What do you think? ("you" includes frequent users!)

:kiss:

1 registered user and 40 guests thanked already.

Users who thanked for this post:

nomada

Posts: 3,012

Date of registration: May 24th 2011

Language Team: English

Focus Group: Timestamp Adjustment Team
Public Relations Group

Location: Malmö, Sweden

Thanks: 90516 / 2397

  • Send private message

14

Monday, January 30th 2012, 10:34am

Well, as we are in the topic of improving the tool, I wonder the following:
The tool presents the transcription strings and correspondent translated text in the green box bellow, that have differences on the timings between the 2 (transcription and translation) of 300 milliseconds. Is this correct Georgy? If so, I am wondering about all those long chunks of text that appear in the green box; they might correspond not only to strings that were merged or split or having words transferred from one string to the next; because the chunks of text are very long, I wonder if most of those strings haven't been through a shift in the initial or starting time with no change of the content? If so, they wouldn't need to be language revised and so less strings to copy-past during the application of the Frag tool. Once again, if this verifies, maybe we could include, in the tool data input, a comparison of the 2 English SRTs (before and after timeshifted) and the tool would tell which segments of text were altered and which were just shifted in time? Then the tool would exclude these last ones from being evaluated by the user. hehe, I hope you get my point. Le me know if it makes any sense?

hugs
Signature from »nomada« Click to know how to: - Receive email notifications - Join the LTI biweekly Meetings

39 guests thanked already.

Posts: 3,012

Date of registration: May 24th 2011

Language Team: English

Focus Group: Timestamp Adjustment Team
Public Relations Group

Location: Malmö, Sweden

Thanks: 90516 / 2397

  • Send private message

15

Monday, February 13th 2012, 4:10am

Remembered one more thing Georgy. I wonder if the tool shows only the strings that have time differences or the ones that have only text differences. If it would show this ones too, we would be sure that wedontneed to revise the translation again in dotsub in order to catch all the changes. We wouldjust need to watch the whole video and call it done : ) Does this makes sense?

hugs
Signature from »nomada« Click to know how to: - Receive email notifications - Join the LTI biweekly Meetings

42 guests thanked already.
© Linguistic Team International 2018
Context In Motion