Pagine: < [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41] > | Ask me anything about subtitling Iniziatore argomento: Max Deryagin
| Max Deryagin Russia Local time: 21:31 Da Inglese a Russo AVVIO ARGOMENTO
Inna Ivanova wrote:
Dear Max,
I have a video file in Bulgarian. I want to find a program/website on which the subtitles can be automatically created in Bulgarian, which then ideally will be machine-translated in English.
Stepan told me of the program Subtitle Edit, but the audio to text of Whisper (ASR) doesn't work. I keep getting the error message - No text found. I tried all options for model - base, tiny, medium, large. Nothing worked. Vosk/Kaldi doesn't support Bulgarian.
I also went through a lot of other programs suggested by another Proz user, but they are all paid or do not have ASR, or do not support Bulgarian.
1. Could you let me know how to fix this error for Subtitle Edit?
2. Which FREE program/site can I use only to automatically create the subtitles in Bulgarian? Is there such a product?
3. If there isn't or you don't know of any such - which is the cheapest and most appropriate to use in my case?
Thank you very much in advance!
Hi Inna,
Since Bulgarian is a relatively "small" language in subtitling, there isn't much high-quality data to use for training ASR and MT engines to translate subs from and into Bulgarian, which means whatever current technology you can find out there won't be able to produce satisfactory results for you. So, good (and especially free) programs/sites for this kind of thing probably don't exist. I think the appropriate thing to do in your case is to subtitle traditionally, without any such automation. | | | Aegisub-merging lines when exporting to .SRT | Dec 2, 2022 |
Hello everyone,
I was wondering if anyone could help me with an issue I'm facing with Aegisub recently.
A client has asked me to create four separate lines with the exact same timecodes that he will then edit differently over the video. I create them with no issue in .ass format, but when I try to export the file as .srt, the four lines are merged into one, separated by line breaks. Is there a way to prevent this from happening and keep them separate when exported to sr... See more Hello everyone,
I was wondering if anyone could help me with an issue I'm facing with Aegisub recently.
A client has asked me to create four separate lines with the exact same timecodes that he will then edit differently over the video. I create them with no issue in .ass format, but when I try to export the file as .srt, the four lines are merged into one, separated by line breaks. Is there a way to prevent this from happening and keep them separate when exported to srt?
Thank you in advance! ▲ Collapse | | | Max Deryagin Russia Local time: 21:31 Da Inglese a Russo AVVIO ARGOMENTO
Maria Vardopoulou wrote:
Hello everyone,
I was wondering if anyone could help me with an issue I'm facing with Aegisub recently.
A client has asked me to create four separate lines with the exact same timecodes that he will then edit differently over the video. I create them with no issue in .ass format, but when I try to export the file as .srt, the four lines are merged into one, separated by line breaks. Is there a way to prevent this from happening and keep them separate when exported to srt?
Thank you in advance!
Hi Maria,
Yep, Aegisub auto-merges lines with identical timecodes upon .srt export, and so do many other subtitling tools. What you can do is open your .srt file in a text editor (e.g. Notepad) and manually rewrite the subs in the correct format. So you take something like this:
351
00:00:05,000 --> 00:00:10,000
Bread.
Bread.
Bread.
Bread.
And turn it into something like this:
351
00:00:05,000 --> 00:00:10,000
Bread.
352
00:00:05,000 --> 00:00:10,000
Bread.
353
00:00:05,000 --> 00:00:10,000
Bread.
354
00:00:05,000 --> 00:00:10,000
Bread.
And then save. That should do the trick. | | |
Brilliant! Thank you so much! | |
|
|
Sandra Silipo Regno Unito Local time: 16:31 Da Inglese a Italiano exporting a video with OOONA tools | Mar 15, 2023 |
Hi
I have just finished to subtitle a video on OOONA Tools and I would like to export it not just as an SRT file but also as an actual video that I can include in my portfolio. Ideally, a video that can be opened with Windows Media Player.
How do I do it?
Thanks,
Sandra | | | Max Deryagin Russia Local time: 21:31 Da Inglese a Russo AVVIO ARGOMENTO
Sandra Silipo wrote:
Hi
I have just finished to subtitle a video on OOONA Tools and I would like to export it not just as an SRT file but also as an actual video that I can include in my portfolio. Ideally, a video that can be opened with Windows Media Player.
How do I do it?
Thanks,
Sandra
Hi Sandra,
If you want to do it from OOONA Tools directly, you will need to use their Burn & Encode tool (paid for separately). Alternatively, you can export your SRTs and then use Handbrake or any other similar software for burning-in. | | | BogMarley Austria Da Spagnolo a Tedesco + ... Timing and frames | Jun 22, 2023 |
Hey.
I got guidelines for a subtitling task and I am a bit confused with their timing instructions.
If there are keyframes, I am supposed to have an Out-time of 3 frames before a keyframe. If they start after a keyframe, the subtitles should start 2 frames after it or, if audio starts 7 frames after a keyframe, subtitles should be synchronised with the audio.
I'm working with Aegisub.
Is there a way to know how many frames I am from a keyframe? The au... See more Hey.
I got guidelines for a subtitling task and I am a bit confused with their timing instructions.
If there are keyframes, I am supposed to have an Out-time of 3 frames before a keyframe. If they start after a keyframe, the subtitles should start 2 frames after it or, if audio starts 7 frames after a keyframe, subtitles should be synchronised with the audio.
I'm working with Aegisub.
Is there a way to know how many frames I am from a keyframe? The audio wave only shows ms. I know, that in a 24 FPS video 1 frame = 42 ms. But do I really have to do the math all the time or is there a way to do it in a more efficient way?
Is there any way to automatise that process? I know that there is a Post-Processing-option in Aegisub, but it also only shows ms and there is no way to select frames.
How can I know the exact number of frames? Do I just have to guess?
Thank you so much for your help! ▲ Collapse | | | Max Deryagin Russia Local time: 21:31 Da Inglese a Russo AVVIO ARGOMENTO
BogMarley wrote:
Hey.
I got guidelines for a subtitling task and I am a bit confused with their timing instructions.
If there are keyframes, I am supposed to have an Out-time of 3 frames before a keyframe. If they start after a keyframe, the subtitles should start 2 frames after it or, if audio starts 7 frames after a keyframe, subtitles should be synchronised with the audio.
I'm working with Aegisub.
Is there a way to know how many frames I am from a keyframe? The audio wave only shows ms. I know, that in a 24 FPS video 1 frame = 42 ms. But do I really have to do the math all the time or is there a way to do it in a more efficient way?
Is there any way to automatise that process? I know that there is a Post-Processing-option in Aegisub, but it also only shows ms and there is no way to select frames.
How can I know the exact number of frames? Do I just have to guess?
Thank you so much for your help!
Hi Bog,
When the video window is active (i.e. when you click on it), you can jump one frame forward/backward by pressing the left/right keyboard arrows. You can change these "arrow" hotkeys to any key combination (and make them global) in the settings. If you enable "Play audio when stepping in video" (in Settings > Audio), you will be able to see on the waveform how many frames away you are from a keyframe.
But generally, Aegisub is not a good tool for working with frames. Subtitle Edit does a much better job, as far as free subtitling software goes. | |
|
|
Mr. Satan (X) Da Inglese a Indonesiano
Hello, Max.
Sorry for asking a random question. But do you have any plan to publish a book about subtitling or AVT? | | | Max Deryagin Russia Local time: 21:31 Da Inglese a Russo AVVIO ARGOMENTO
Mr. Satan wrote:
Hello, Max.
Sorry for asking a random question. But do you have any plan to publish a book about subtitling or AVT?
Haha, I was thinking about perhaps compiling all the "Subtitle Appearance Analysis" and/or "Toward Intelligent Subtitling" articles into books, but I just never have time for actually writing the articles to begin with. Not even sure if I'll ever finish my article on subtitle colours (it's sooo complicated!)
But hey, maybe one day... | | | BogMarley Austria Da Spagnolo a Tedesco + ...
Thank you so much for your help! | | | Yaotl Altan Messico Local time: 10:31 Membro (2006) Da Inglese a Spagnolo + ...
Max Deryagin wrote:
...
Haha, I was thinking about perhaps compiling all the "Subtitle Appearance Analysis" and/or "Toward Intelligent Subtitling" articles into books, but I just never have time for actually writing the articles to begin with. Not even sure if I'll ever finish my article on subtitle colours (it's sooo complicated!)
But hey, maybe one day...
You should because it's a very useful thread for those who want to know more about those subtitling details not so visible at a first glance when working in a certain software.
Thank you very much for this saga, Max | |
|
|
indicating foreign language | Jul 11, 2023 |
There's one scene I'm subtitling where two speakers are speaking Spanish but the viewer is not intended to understand. I know if it was just one speaker I would display [говорит по-испански]. If it's two speakers speaking Spanish, do I need to subtitle it as as a dialogue (displaying it one about the other) or would one subtitle suffice to make the viewer understand that it is both speakers?
Also, if it's there's a very long monologue, should [говорит по-и... See more There's one scene I'm subtitling where two speakers are speaking Spanish but the viewer is not intended to understand. I know if it was just one speaker I would display [говорит по-испански]. If it's two speakers speaking Spanish, do I need to subtitle it as as a dialogue (displaying it one about the other) or would one subtitle suffice to make the viewer understand that it is both speakers?
Also, if it's there's a very long monologue, should [говорит по-испански] be displayed for the duration of the monologue, or just for a second or so at the beginning of the monologue?
[Edited at 2023-07-11 21:00 GMT]
[Edited at 2023-07-11 21:01 GMT] ▲ Collapse | | | Max Deryagin Russia Local time: 21:31 Da Inglese a Russo AVVIO ARGOMENTO
whitelighter wrote:
There's one scene I'm subtitling where two speakers are speaking Spanish but the viewer is not intended to understand. I know if it was just one speaker I would display [говорит по-испански]. If it's two speakers speaking Spanish, do I need to subtitle it as as a dialogue (displaying it one about the other) or would one subtitle suffice to make the viewer understand that it is both speakers?
Also, if it's there's a very long monologue, should [говорит по-испански] be displayed for the duration of the monologue, or just for a second or so at the beginning of the monologue?
Hi whitelighter,
1. One subtitle suffices: [говорят по-испански]
2. Only for a few seconds. | | | Max Masutin Ucraina Local time: 18:31 Da Inglese a Russo + ...
I learned about Whisper in this thread and came upon this standalone executable of OpenAI's Whisper & Faster-Whisper for those who don't want to bother with Python; for Windows 7 x64 and above to be used in scripts or Subtitle Edit. Faster-Whisper is much faster than OpenAI's Whisper and requires less RAM/VRAM. It automatically uses GPU processing if CUDA is dete... See more I learned about Whisper in this thread and came upon this standalone executable of OpenAI's Whisper & Faster-Whisper for those who don't want to bother with Python; for Windows 7 x64 and above to be used in scripts or Subtitle Edit. Faster-Whisper is much faster than OpenAI's Whisper and requires less RAM/VRAM. It automatically uses GPU processing if CUDA is detected. A Tiny recognition model is enough for testing purposes, especially if not running on CUDA; for decent transcription, use Medium or larger. OpenAI needs FFmpeg in the %PATH% variable or in Whisper's folder. Some defaults are tweaked to transcribe movies and make it portable. As a standard, it looks for models in the same folder in _models\faster-whisper-tiny (or another model's title), otherwise you need to specify it in --model. The model is downloaded automatically to the Windows folder or similar. Preferably, get one here. In Subtitle Edit, the executable can be selected for CTranslate2 engine, just rename it to whisper-ctranslate2.exe. If not in Subtitle Edit, you might need a basic understanding of command-line scripts to use it. I refer you to my issue with it, in case you're proficient in matters like CUDA and the required drivers.
Code:
| @echo off
set File=
set Model=Tiny
set Device=CPU
:: Use a short audio file or track first to see the results and how it runs on CPU. If satisfied, specify CUDA. You might get errors.
set Whisper=Whisper-Faster_r134
%~dp0%Whisper% --language en --output_format srt --task transcribe --model %Model% --device %Device% "%~dp0%File%"
pause |
|
▲ Collapse | | | Pagine: < [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41] > | To report site rules violations or get help, contact a site moderator: You can also contact site staff by submitting a support request » Ask me anything about subtitling CafeTran Espresso | You've never met a CAT tool this clever!
Translate faster & easier, using a sophisticated CAT tool built by a translator / developer.
Accept jobs from clients who use Trados, MemoQ, Wordfast & major CAT tools.
Download and start using CafeTran Espresso -- for free
Buy now! » |
| Anycount & Translation Office 3000 | Translation Office 3000
Translation Office 3000 is an advanced accounting tool for freelance translators and small agencies. TO3000 easily and seamlessly integrates with the business life of professional freelance translators.
More info » |
|
| | | | X Sign in to your ProZ.com account... | | | | | |