Download Subtitle Edit. You will also need to download VLC Media Player.
Open the subtitle file with Subtitle Edit. If you save the video in the same folder and with the same name as the subtitle file, the software will load the video automatically.
You can also click on the upper right area of the interface to open a video file:
Download the necessary dictionaries by going to Spell Check > Get Dictionaries in the main menu:
Select the dictionaries you will use for the project:
For project-specific words, proper names, etc., it's very important to ensure consistency throughout the project.
For example, if a character's name is Zaboleff, make sure you add it to the name list so that the spell-checker flags any other spellings (for instance, if you type “Zeboleff”).
Click on Add name to list (Ctrl + Shift + L) and write the name in the pop-up window:
You can also right-click on a word in the subtitle edit box to add it to the Names list:
You can see the word lists by clicking on Settings > Word lists:
If you are provided with the original script, you can extract the dialogues with an online app: https://www.adobe.com/acrobat/online/pdf-to-word.html
If you have an srt file, you can add subtitles quickly by clicking on the voice line and dragging to the right, then right-clicking and selecting Add text here from the menu:
For Netflix the subtitles should be spotted to shot-change, that is, subtitles should not straddle a shot-change as far as possible.
When there is dialogue across a shot-change, there are some rules that should be followed:
No subtitle should START between 1 and 11 frames before the shot-change.
● If the subtitle starts up to 7 frames before the shot-change, move the in-time to the shot-change.
● If the subtitle starts between 8 and 11 frames before the shot-change, move the in-time to 12 frames before the shot-change.
No subtitle should END between 11 and 1 frames after the shot-change.
● If the subtitle ends up to 7 frames after the shot-change, move the out-time to the shot-change, respecting the 2-frame gap.
● If the subtitle ends between 8 and 11 frames after the shot-change, move the out-time to 12 frames after the shot-change.
In Windows, right-click on the video and select Properties, Details, and check the frame speed.
Open Subtitle Edit.
Go to Options>Settings, and in Profile select Netflix. Make sure you change the Time code mode to HH:MM:SS:FF so that you can move the timecodes by frame instead of by millisecond.
Click on Waveform/Spectrogram and Download Ffmpeg.
Close the Settings window and click on Video in the main menu. Select Generate/import scene changes.
Click on Generate scene changes with Ffmpeg, making sure you set the Sensitivity to 0.1 and the Time codes are in frames.
Wait until the shot-changes are generated.
Shot-changes will appear as a vertical white line (in very dark scenes, the program may not be able to detect shot-changes).
Make sure the subtitle whose time-codes you want to change is selected by double-clicking on it.
This is a selected subtitle straddling a shot-change:
To change the in-time, drag the time code to the shot-change until you see a dash line.
Now, using the steppers, move the in-time to the left to make sure it's not between 1 and 11 frames before the shot-change, the forbidden area for a subtitle to start.
These are the steppers to change the in-time:
Press the down stepper and the whole subtitle will move (not just the time code). Count the frames from the shot-change to the beginning of the audio and place the in-time either on the shot-change (respecting the 2-frame gap) or 12 frames or more before the shot-change (see screenshot at the beginning of this tutorial).
To change the out-time, drag the time code to the shot-change until you see a dash line.
Press the stepper to move the time code. Count the frames from the shot-change to the end of the audio and place the out-time either on the shot-change or 12 frames after the shot-change (see screenshot at the beginning of this tutorial).
These are the steppers to change the out-time:
Press the stepper several times to move the out-time (red dash line) 12 frames after the shot-change (white line).
Please follow Netflix Timed-Text Style Guidelines:
https://partnerhelp.netflixstudios.com/hc/en-us/sections/203480497-Timed-Text-Style-Guides
https://docs.google.com/spreadsheets/d/1oNJQBbfTCGBeziKjYftn-J5JU2KHFU7_CThZryQ_llg/edit?urp=gmail_link&gxids=7628#gid=0
You can generate automatic error reports on Subtitle Edit. Go to Settings > General and in Profile select Netflix:
At least the following errors should be fixed:
42 characters per line (don't break subtitles with fewer than 42 characters)
For example, this subtitle is only 41 characters in total, so it should be unbroken
Use the smart character U + 2026 […] instead of three dots.
Use an ellipsis when dialogue trails off or to indicate long pauses.
-I thought you were going to…
-What did you say?
I doubt… you'll like it.
Use an ellipsis to indicate that a subtitle starts mid-sentence:
…more time to do this.
Use an ellipsis when a sound cue or false narrative is inserted in a sentence split into two subtitles:
-[seagulls mewing]
-…none of them wanted to come.
Use two hyphens to indicate abrupt interruptions:
-Don't start with–
-Let me finish!
Use hyphens to indicate two speakers within a subtitle or two sound cues originating from different sources:
-Hello, my name is Andy.
-Nice to meet you.
-[thunderclap]
-[screaming]
Do not use italics when the speaker is within the same reality that we are seeing, for example, in a different room of the house or a different part of the building, behind a closed door, etc.
Use italics for:
1- Narration
2- Titles of books, plays, works of art, movies, albums, video games, etc. Use quotation marks for song titles.
3- Dialogue heard through electronic media like radio, TV, public announcement systems, etc.
4- A character's inner monologue that only we can hear:
5- When a character remembers a past scene:
Numbers above ten should be written numerically, except at the beginning of a subtitle:
You owe me 50 quid.
Fifty, I said.
Avoid using semicolons (;) and ampersands (&).
Exceptions: When part of a brand name.
Johnson & Johnson.
Always use commas before and after non-verbals (oh, uh-huh, etc.) and direct address (a person's name or any word used to address a person directly).
Examples:
I see, oh, all right.
Yes, sir.
It's wonderful, dear.
So, Martin, what do you think?
Come along, Dad.
Subtitles and SDH captions should be centered for Netflix deliverables.
Subtitles should be positioned at the bottom of the screen unless they cover important information, including initial credits. Avoid jumping subtitles.
In this example, the contents of the letter are plot-relevant so the subtitle should be placed at the top of the screen.
Line-breaks should follow Netflix specs. Syntactic units should never be split:
The line should be broken:
- after punctuation marks (ellipses, commas, periods, etc.)
- before conjunctions (e.g. and, but, for, if, nor, or, yet, so)
- before prepositions (e.g. in, near, to, with, after, besides, about, etc.)
The line break should not separate:
- a noun from its article
- a noun from an adjective
- a first name from a last name
- a verb from its subject
- a prepositional verb from its preposition
- a verb from an auxiliary, reflexive pronoun or a negation
- a short cue or name tag from dialogue
Here’s a list of prepositions:
https://simple.wikipedia.org/wiki/Preposition#:~:text=Prepositions%20often%20describe%20the%20position,be%20used%20to%20end%20sentences.
Phrasal verbs should not be split up:
https://simple.wikipedia.org/wiki/Phrasal_verb
Object pronouns should not be split from the verb:
https://en.wikipedia.org/wiki/Object_pronoun
Adjectives and their noun should not be split:
https://en.wikipedia.org/wiki/Adjective
Possessive pronouns should not be split from their nouns:
https://en.wikipedia.org/wiki/Possessive_determiner
Other syntactic units that should not be split: negative particle from its verb, participle from its verb, and adverb from the verb it modifies.
For unclear audio, slow down the video using the two green arrows at the bottom of the interface:
Select a speed of between 30% and 70% and play the video several times until you understand the dialogue:
Always verify proper nouns, song lyrics, etc. For example, if you hear something similar to “stavasky”, try using relevant keywords:
If you don't know how to describe sounds, search for images of an object to find out its name or do different searches using possible keywords:
The Reviewer should:
● Watch the whole film with subtitles
● Correct timing
● Correct spelling
● Check that translation is complete against audio –correct omissions, mistranslations, etc.
● Run Netflix Quality Check and resolve warnings
● Check the translator's credits
Please run Netflix quality check and resolve the errors before delivery on every step (after translation, after SDH, after review).
Tools > Netflix quality check
https://partnerhelp.netflixstudios.com/hc/en-us/articles/217350977-English-Timed-Text-Style-Guide
Only plot-pertinent sounds should be included, and only if the sound cannot be deduced from the image.
● Verify proper nouns, name tags, cues, songs, and foreign dialogue as necessary.
● Important sounds are captioned as sound cues to capture off-screen action, clarify onscreen action or complement dialogue and name tags are added if the speaker is difficult to identify.
● Lyrics and sing-song boxes must be formatted in italics and include quaver notes (♪) at the beginning and end of each subtitle. They are formatted without any other end punctuation, except for question marks, exclamation marks or ellipses.
♪ <i>Stay with me forever more
My living doll</i> ♪
IMPORTANT: Even if lines of song are interrupted by dialogue a closing quaver note is used only when the song actually ends. Example:
Box 1: One line:
♪ <i>You're a dolly</i>
Box 2:
I really like this song.
Box 3:
♪ <i>You’re so cute and lovely
Dearest little doll</i> ♪
● Interruption of songs should follow the standard use of an ellipsis. For example: ♪ <i>Happy birthday to…</i> ♪
· Include the source of the sound and its description.
· If the source is seen, only the description is necessary.
[shouting]
· For a sustained sound, use the present participle form of the verb.
[dog barking]
· For an abrupt sound, use the third person verb form.
[sighs]
· If a character makes a sound not visually apparent, caption as a sound cue.
[sighing]
· If it is not clear who is making the sound, identify the speaker as part of the sound cue:
[John coughing]
· Sound cues can be used to capture moments of silence that affect the plot. For example, if a character turns off the radio and it is important for the viewer to note this, use the following sound cue:
[music stops]
· When a character is mouthing dialogue, and the dialogue can be visually deciphered but not heard, the following cue can be used:
[mouthing]
· When a character is saying something, but the audience cannot hear nor visually decipher what is being said, the following cue can be used:
[inaudible]
· Add only pertinent sound cues that are relevant to the plot. Please do not add unnecessary cues like [crickets chirping] during a night scene and [birds chirping] during a day scene.
Though some sounds can be “seen” by a viewer, such as an explosion, many on-screen sounds cannot be observed. If unsure whether to a cue, watch the video without sound. If the lack of sound creates ambiguity within the scene, add a cue.
· For incidental background chatter where the audience can see the people moving their mouths, you can use the following sound cues:
[people speaking indistinctly]
[murmuring]
Do not use [chattering].
· If the indistinct chatter communicates a specific emotion or action, use an appropriately descriptive sound cue. For example:
[cheering]
[jeering]
If it’s obvious from context and visual information what people are doing, no sound cue is needed. For example, if people are seen applauding and giving a standing ovation, it is not necessary to add the sound cue [cheering].
· If the sounds are from two different sources, they cannot be captured in the same cue. Separate the cues using speaker tags.
· If two sounds are made by the same person/source in quick succession, they can be captured in the same box and formatted as follows:
· Place the sound cue where it occurs in the dialogue in the following cases:
1. A speaker makes a sound that needs to be cued in the middle of dialogue
They said the singer was admitted to the hospital
for [clears throat] “nervous exhaustion.”
2. A speaker makes a sound that needs to be cued at the start or end of dialogue.
[coughs] What was that?
Think I caught a bug. [sneezes]
Note that the punctuation should go before the sound cue, not after.
· Audibly censored words:
Get the [bleep] out of here!
· If there is no audible beep, the sentence should be written as follows:
Get the…out of here!
(Please note the lack of spaces before and after the ellipsis)
1. Onscreen music:
● Identify or describe all on-screen music (that is, music that characters can interact with).
● Indicate the source of the music being played (for example, radio, band, speaker, etc).
2. Background Songs
● When cueing background songs and adding lyrics, do not include lyrics if they are of an overly fragmented nature i.e. snippets that appear between foreground dialogue, as this will distract the viewer from the main action of the scene.
3. Score
● There is no need to include cues for score, unless the client considerations request it, or it falls under one of the following two instances:
The first instance where you would need to identify or describe score is if it provides a clue to the plot of the movie, such as the Jaws theme that plays whenever the shark is approaching. In this case, this is the only way the viewer realizes that the character onscreen is in danger.
The second instance where you would need to describe score is when it is contrary to the mood of the scene (for example, a dramatic scene with happy music).
● Use the format: [XXXX instrumental music] where XXXX is an appropriate adjective, (heroic, sentimental, etc.)
● If a song cannot be identified or the piece is generic, musical sound cues should be used.
● If the music is heard in the reality of the scene, identify the source as necessary.
[“Coming Back to Life” playing on radio]
● When including song titles, make sure the title has been verified from a credible source. If this is not possible, use a generic cue.
● The artist's name is not always included in a music cue. We must determine if it is relevant to include or not. It is usually included when it is unexpected or when it is part of a joke or to emphasize the context of the music choice.
● When a character's mouth moves but they are not speaking actual dialogue and it is not obvious from the action what is happening, ensure that a sound cue or non-verbal expression is inserted so that the viewer does not think dialogue has been excluded.
● You can use either a sound cue [sushing] or a non-verbal expression such as “Shh” to represent the non-verbal dialogue.
● Avoid representing superfluous non-verbal expressions in dialogue that do not add to viewer understanding.
● If the non-verbal superfluous dialogue is indicative of something important, like the character is scared or lying, it must be captured. Sometimes, instead of including the actual non-verbal expression, try using a cue to better represent this superfluous dialogue.
If it is necessary to indicate how dialogue is said when using a name tag, separate the name tag from the sound cue.
- [John whispering] Where are you?
If you did the translation, please add a subtitle in the end:
Translated by [Name]
in your respected language. Please find Translator's credit here. E.g.
you can use pen name at the end of file.