Caption

German subtitles of a Dutch film originally recorded in English

Captions are the texts that appear at the bottom edge of an image, often superimposed on it, providing additional information about it or translating a narration or dialogue spoken in a different language. By moviegoers, because they allow the audition of the original actor's voice if the language of the production is not mastered, they are also used as an aid to improve knowledge of a language or as a substitute for the soundtrack for people with hearing impairments. In Latin America, subtitles are widespread in movie theaters, while dubbing is common on open-signal television stations. In Spain, however, they are restricted to a relatively specialized cinema circuit, with dubbed versions generally being shown in the rest of the cinemas and on television channels.

Types of subtitles

In 2003, coinciding with the publication of the first draft of the AENOR standard on subtitling, the researcher Jorge Díaz Cintas published Theory and practice of subtitling. English - Spanish, a work that constitutes a first approach to the categorization of subtitling typologies. Fifteen years later, based on Díaz Cintas' classification, Pablo Romero Fresco established two divisions when it came to categorizing subtitling. On the one hand, it includes subtitling according to its receiving audience, giving rise to two types of subtitling: subtitling —subtitling, just— and subtitling for the deaf —subtitling for deaf people.

The rise of the Internet and IDS platforms, together with the cultural change in Spanish society, has favored the appearance of new uses of subtitling. Trends such as the consumption of video through mobile phones in public spaces and with the sound turned off, access to other countries' own programming —in other languages— or a progressive consumption of series and films in their original version to the detriment of dubbing, have revitalized the use of subtitling and have revealed new uses, such as creative subtitling, karaoke or translation when audio is not available. The most recent table of subtitling typologies, which takes into account the characteristics of the new video-on-demand media, is that of Gema López Sánchez and Francisco Utray, published in 2020:

Television

Some television channels offer, through teletext (Closed Caption), subtitles adapted for the hearing impaired. When you select the teletext page with the subtitles, the TV shows them on top of the picture. Others even offer the option to add subtitles in another language for foreign people or in places where the video is not heard.

Film Subtitling

Cinematographic subtitling arose with the arrival of talkies (1927) with the aim of translating films that, unlike silent films, were no longer built on the basis of universal images, understandable to everyone. This new problem caused by the limitation of a script in only one language, led to the need to look for new solutions.

Technical aspects of subtitling

Subtitling presents a series of technical problems that make it difficult to execute: on the one hand, the space available on the screen is limited, which leads to the fragmentation of the text and, on the other hand, the lack of contrast between the background of the image and the color of the subtitles, which makes it very difficult to understand. It is important, therefore, that the subtitles are well synchronized with the speech of the actors, something that forces to cut and simplify the written text. On the other hand, the fact that the human eye has the average reading capacity of one second for a word that contains 5 to 8 letters. In this period of time, 24 images pass through the screen, which sometimes causes visual shocks that cannot be avoided. In addition, the fact of having to read the subtitles prevents the viewer from seeing the full screen correctly, since they must focus on the bottom part, depriving themselves of part of the image. Finally, the introduction of a written text on the screen can alter the aesthetics of the image, and supposes an extra effort for the viewer, since they must be aware of reading while paying attention to what happens in the rest of the shot.

Despite all this, subtitling is a very good option when translating movies into several different languages, as it represents the cheapest solution, much more so than dubbing.

For captions to be readable, it is recommended that each line remain on the screen for about three seconds. It is also of great importance to take into account that, regardless of the minimum time that the viewer needs to read, the subtitles must correspond to what is happening on the screen. For example, in the case of a dialogue, they should not stay much longer than it takes the actors to recite the text, as this could be confusing. Similarly, when an actor is speaking off-screen, subtitles should only appear when the actor enters the scene.

Sentences should be as clear and simple as possible, except for cases where the original text is confusing or ambiguous because the script requires it (for example, if one of the characters is delirious). They should also be fairly neutral. All exclamations will be dispensable, since it must be the actor who makes the viewer understand the tone in which each line is said. It is important that each line appear on the screen one at a time, although it is often impossible, because the spoken language is too long, and priority must be given to keeping the subtitle in sync with the original text. For this reason, it is common to see up to three lines at a time, although it is more difficult for the viewer to read without getting tired.

Media Containers

Other containers

There are other multimedia containers that can store subtitles in text format, which takes up much less space. These subtitles are written on the screen over the video image.

These containers are generally compatible with most home players. For this to be possible, you normally have to activate the subtitles (or press the subtitles button) on the remote control of the video player equipment, similar to how it is done on DVDs. Also, the subtitle file must be named the same as the one that contains the video.

Examples of containers:

• AVI (.avi)
• MP4
• GMO (.ogm)
• Matroska (.mkv)

Subtitle Formats

• Format plain text:
  • SubRip (.srt)
  • MicroDVD (.sub)
  • Universal Subtitle Format (XML format with advanced features)
  • Substation Alpha (.ssa), and Advanced Substation Alpha (.ass)
• Other formats (.smi,.rt.txt,.aqt,.jss,.js)

Image format

• • Open Picture Subtitle
  • DVD (subtitles are saved in VOB files)
  • SVCD (unlike VCD, allows floating subtitles)
  • CVCD

For the sounds there is also a kind of subtitles to show on the computer or portable MP3 player. It is used to show the lyrics or some audio information. An example is the LRC (lyrics) format for displaying lyrics or written information in podcasts. They are also useful for karaoke.

Software and subtitles

There are programs adapted for comfortable writing of subtitles, video player programs that allow synchronizing a video file with a subtitle file, and video editing programs that allow including subtitle tracks from an external file or with tools suitable for the treatment of subtitles.

Write subtitle files

• Gaupol (GPL) supports the formats:

SubRip (.srt)

MicroDVD (.sub)

Substation Alpha (.ssa) - partly

Advanced Substation Alpha (.ass) - partly

MPsub (MPlayer subtitle)

SubViewer 2.0

MPL2

TMPlayer

• Gnome Subtitles (GPL) supports formats:

SubRip (.srt)

MicroDVD (.sub)

Substation Alpha (.ssa)

Advanced Substation Alpha (.ass)

MPsub (MPlayer subtitle)

MPlayer

MPlayer 2

SubViewer 1.0

SubViewer 2.0

Adobe Encore DVD

DKS Subtitle Format

Karaoke Lyrics LRC

Karaoke Lyrics VKT

Panimator

Phoenix Japanimation Society

Power DivX

• Subtitle Edit Filed on August 22, 2010 in Wayback Machine. (GPL) you can read, write and convert between more than 80 subtitle formats, such as:

SubRip (.srt)

MicroDVD (.sub)

Substation Alpha (.ssa)

Advanced Substation Alpha (.ass)

MPsub (MPlayer subtitle)

SubViewer 2.0

MPL2

Plain-Text

Adobe Encore DVD

Video player uses DirectShow or VLC media player

Autotranslation via Google translate

• Subtitle Editor (GPL) supports formats:

SubRip (.srt)

MicroDVD (.sub)

Substation Alpha (.ssa)

Advanced Substation Alpha (.ass)

MPsub (MPlayer subtitle)

SubViewer 2.0

MPL2

Plain-Text

Adobe Encore DVD

• KSubtile (GPL) supports formats:

SubRip (.srt)

• Tea supports formats:

SubRip (.srt)

Video playback

• Kaffeine (GPL) supports formats:

SubRip (.srt)

MicroDVD (.sub)

Substation Alpha (.ssa)

.smi

.asc

.txt

• VLC media player (GPL) supports formats:

SubRip (.srt)

MicroDVD (.sub)

Substation Alpha (.ssa)

SubViewer

VobSub (Image Track)

Sami

• MPlayer (GPL) supports formats:

SubRip (.srt)

MicroDVD (.sub)

Substation Alpha (.ssa)

MPsub (MPlayer subtitle)

SubViewer

VobSub (Image Track)

GMO

CC (closed caption)

Sami

VPlayer

RT

PJS (Phoenix Japanimation Society)

AQTitle

JACOsub

Video editing

• DVDauthor - Spumux (GPL) supports formats:

SubRip (.srt)

MicroDVD (.sub)

Substation Alpha (.ssa)

Advanced Substation Alpha (.ass)

.smi

.rt

.txt

.aqt

.jss

.js

For DVDauthor there is a desktop interface, like Q-DVDAuthor (SubRip - *.srt) or DVDStyler

• Subtitler of Cinelerra supports the formats:

SubRip (.srt)

• Avidemux (GPL) supports formats:

SubRip (.srt)

Advanced Substation Alpha (.ass)

VobSub (Image Track)

Subtitle Code Examples

Universal Subtitle Format code example taken from http://usf.corecodec.org/files/SVGsample.usf (broken link available at Internet Archive; see history, first and last version).

?xml version="1.0" encoding="UTF-8"?
! DOCTYPE USFSubtitles SYSTEM "USFV100.dtd" --
?xml-stylesheet type="text/xsl" href="USFV100.xsl"?
ΔUSFSubtitles version="1.0"
≥metadata
أعربية containing only SVG tax/title
▪
أعربية
≥email limitunmei@matroska.org
◊
 backward language code="eng"
≤2 / 2 / 2 1 2 / 2 2
Δ/metadata
Δsvgdefs
≤2
≤2="black-white_1" x1="0%" y1="0%" x2="0%" y2="33.3333%" spreadMethod="reflect" gradientUnits="objectBoundingBox"
%" stop-color="#000"/
"100%" stop-color="#FFF"/
Δ/linearGradient
≤2="black-white" x1="0%" y1="0%" x2="100%" y2="0%" spreadMethod="pad" gradientUnits="objectBoundingBox"
%" stop-color="#000"/
"100%" stop-color="#FFF"/
Δ/linearGradient
Δ/defs
Δ/svgdefs
≥subtitles
Δsubtitle start="00:00:02,000" stop="00:00:08.000"
vertical-margin="10%" alignment ="TopRight" scale-x="500" scale-y="500"
Δrect x="102" y="30" width="130" height="81" fill="none" strike="url(#black-white)" stroke-width="5" stroke-opacity="0.75"/
Δellipse cx="117" cy="445.5" rx="54" ry="41.5" fill="none" stroke="url(#black-white_1)" stroke-width="8"
≤448.78,103.659 373.171,137.805 443.902,196.341" stroke="none" stroke-width="0" fill="url(#black-white)"/105
Δ/svg
Δ/subtitle
Δ/subtitles

Sample MicroDVD subtitle code from the movie Citizen Kane

[INFORMATION]
[TITLE] Citizen Kane
[AUTHOR]CabRa
[SOURCE]Subtitles captured by SubRip 0.8b
[PRG]
[FILEPATH]
[DELAY]0
[CD TRACK]0
[COMMENT]
[END INFORMATION]
[SUBTITLE]
[COLF] fakeHFFFFFF,[STYLE]bd,[SIZE]18,[FONT]Arial
00:00:00.00,00:00:00.10


00:02:23.99,00:02:25.22
Rosebud.

00:03:12.23,00:03:15.03
"EN XANADU ORDENO CONSTRUIR[br]KUBLA KHAN..."

00:03:15.27,00:03:18.27
"A MAJESTUOSO DOMO[br]DE PLACER. "

00:03:22.79,00:03:24.51
At the legendary Xanadu...

00:03:24.75,00:03:28.14
Kubla Khan ordered to build[br] his dome of pleasure.

00:03:28.39,00:03:31.78
So legendary is today[br]eIXanadu de Florida,...

00:03:32.03,00:03:34.95
the world's largest private amusement park.

00:03:36.87,00:03:38.87
Here in the desert...

00:03:39.11,00:03:41.79
an artificial mountain was built.

....

01:53:10.47,01:53:15.03
Do you think it will be[br] fun to run a newspaper?

01:53:19.47,01:53:25.03
Subtitles by Andrés (Gf-X.com.ar)

Sample caption code in SSA

[Script Info]
; This is a Sub Station Alpha v4 script.
; For Sub Station Alpha info and downloads,
; go to http://www.eswat.demon.co.uk/
; or email kotus@eswat.demon.co.uk
;
; Note: This file was saved by Subresync.
;
ScriptType: v4.00
Collisions: Normal
PlayResX: 640
PlayResY: 480
Timer: 100,0000

[V4 Styles]
Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, TertiaryColour, BackColour, Bold, Italic, BorderStyle,
Outline, Shadow, Alignment, MarginL, MarginR, MarginV, AlphaLevel, Encoding
Style: Default,Simhei,24, fakeHffffff, strangerHffffff, strangerH232323232323,-1,0,1,1,1,1,1,2,20,15,0,1
Style: Default2,Simhei,20, fakeHffffff, strangerHffffff, fakeH666666666666,-1,0,1,1,1,1,2,20,20,15,0,1
Style: Default3,Simhei,30, fakeHffffff, fakeHffffff, fakeH33333333,-1,0,1,0,2,20,20,15,0,1
Style: Default4,Simhei,20, fakeHffffff, strangerHffffff, strangerH666666666666,-1,0,1,1,1,1,6,20,15,0.1

[Events]
Format: Marked, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: Marked=0.0:00:04.40.0:00:05.80,Default,0,0000,0000,000.0, ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ------ ---- ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Dialogue: Marked=0.0:00:06.00.0:00:07.40,Default,0000,0000,0000,手の,。
Dialogue: Marked=0.0:00:07.70.0:00:09.00,Default,0,0000,000.000.000.000.0,探.
Dialogue: Marked=0.0:00:09.40.0:00:12.40,Default,,0,0000,0000,000.0.000,...
Dialogue: Marked=0.0:00:12.60.0:00:14.00,Default,0000,000.000.000.000.000.
Dialogue: Marked=0,0:00:14.30,0:00:15.70,Default,0000,0000,0000, boundary -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ---- ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Dialogue: Marked=0.0:00:15.90.0:00:17.70,Default,0,0000,000.000.0, CCW,
Dialogue: Marked=0.0:01:56.20.0:01:58.90,Default, 0,0000,0000,0000,0000, participan,の,。
Dialogue: Marked=0.0:01:59.20,0:02:00.90,Default,,0000,0000,0000,0000,, would you believe?

SubRip Subtitle Code Example

1
00:00:00,394 -- 00:00:03,031
Previously on Sons of Anarchy...

2
00:00:03,510 -- 00:00:05,154
Put this on. He'll help.

3
00:00:05,274 -- tax 00:00:07,021
I'm fine without him.

4
00:00:08,099 -- 00:00:12,357
He sold the wrong person.
And now he's got my son.

5
00:00:12,477 -- 00:00:14,127
We know you have a fake passport.

6
00:00:14,522 -- 00:00:16,207
He's back in Belfast.

Sample karaoke or subtitle code in LRC for the song "Beauty And The Beast" by Celine Dion

[ti:Beauty And The Beast]
[Celine Dion]
[au:Howard Ashman]
[to:Celine Dion - All The Way... A Decade Of Song]
[by:Wooden Ghost]
[re:A2 Media Player V2.2 lrc format]
[see:V2.20]
[00:02.00]Beauty And The Beast
[00:04.50]
[00:04.80]Celine Dion & Peabo Bryson
[00:07.30]
[00:07.60]Music by Alan Menken
[00:10.10]
[00:10.40]Lyrics by Howard Ashman
[00:13.30]
[00:14.80]≤00:15.30 taxTale ≤00:15.70 taxas ≥00:16.20.06.00.70 taxas ≥00:17.20.
[00:20.90] pupil00:21.40 HCFCTrue ≤00:22.20.25 taxas ≤00:22.50 regulait ≤00:22.75 taxcan ≥00:23.10.
[00:27.20] habit00:27.70 relateBare vis00:28.10 voluntarily ≤00:28.50 taxe vis00:28.85 taxven ≤00:29.20 visfriend vis00:30.10.
[00:30.50] dose00:30.80 taxThen ≤00:31.25 taxsome dose00:31.75 taxbo vis00:32.05.05.05.
≤00:33.90 taxun cup00:34.70 taxex tariff00:35.00 tax excl.:35.30 taxed tax00:36.10.
[00:39.70] habit00:40.20 biblical Just ≤00:40.80 taxa ≤00:41.25 taxlit ten00:41.45 PHPtle ≤00:41.70 taxchan habit00:43.70 regulage vis00:44.00
[00:45.80] habit00:46.30.PHUSL:00:46.50.PHYMALL ≤00:00:47.35 orgexsay ≤00:47.70 taxthe ≤00:48.05.05 tax excl.00:48.80)
[00:49.15]≤00:49.55 PHPBoth ≤00:50.10.2005 taxa ≤00:50.35 taxlit≤00:50.70 PHPtle ≤00:51.15
[00:52.20] habit00:52.50 PHPNei habit00:52.90 regulather ≤00:53.30 regulaone ≤00:53.65 taxpre≤00:54.15.
....
....
....
[03:27.20] habit03:27.70 regulaOh... habit03:28.30 backwards.03:29.10.
[03:32.90]
[03:33.40]
[03:41.75]
[03:43.25] habit03:43.75 taxBeau vis03:44.25 regulaty θ03:45.10 and ≤03:46.15 taxthe vis03:47.75 tax ≤03:49.75 taxBea tax03:55.35 regulast≤3:55.65
[03:57.15]
[03:58.65]
[03:59.85]
[04:00.15]
[04:00.45]Internet service is provided by CDMI

Differences between subtitle formats

Most formats have their advantages and disadvantages. As a general rule, we can state that the more complex a format is, the fewer programs that process subtitles will be compatible with it:

Plain subtitles

• The most compatible formats of all are the so-called "Lands". They carry this name because they have no styles, no "tags", no effects. They are only the text of the dialogue with their respective times of initiation and completion of the dialogue. The format of subtitles "loud" par excellence is the.srt plain. This format is the most used of plain text formats being (in its version of "srt plain") the only one compatible with some of the softsubs DVD creation programs. In its evolution, until the.srt became more complex, and now if you want, you can add "tags" (of italic, bold and colors), and also if you want you can superpose lines of dialogue (that means that you can put two different dialogues that happen in the same time span), and there will be programs of the kind that process the subtitles, which will "understand" the tags on one side, and the other. Note that this is no longer "srt plain" and stops being compatible with simpler programs.

• Other formats "Lands":.sub,.sub type MicroDVD (the only one whose times depend on the framerate of the video, although they are very easy to convert to other formats that do not depend on the framerate), the smi/sami if it does not have color specifications or any of that, the same.txt, and also the.ssa/ass is considered "loud".

• All "Land" subtitle formats are very easy to convert one into the other, without fear of losing information during conversion. Therefore, if you have a program of those who process subtitles that "understands" any of these formats, but the format that is downloaded from the internet is different from that which "understands" the program, and is also "loud", simply becomes the subtitle of used and ready format. There's no danger in making conversions when subtitles are plain. The program most used to convert formats between itself is the Subtitle Workshop, which can open up an important variety of subtitle formats.

Subtitles with tags

- During the evolution of "programs that process subtitles", versions of the programs began to appear that could "understand" what in computer science are called "tags". A tag is something that looks like this:

This subtitle is locked in a tag called "i".

Or sometimes it looks like this:

This subtitle is locked in a tag called "i."{i0}

The feature of all "tags" is that they are divided in two: one opening and one closing, and everything that is enclosed in the middle is said to be "modified" for that tag. Therefore, the two texts that we show as examples are modified by the tag "i". The first example is in the.srt, the second example is in the Substation Alpha (ssa or ass), and what the "i" tag does, in both cases, is to put the text in italics. If the "closing tag" does not appear, then programs assume that the entire line is modified by the same tag:

This subtitle is locked in a tag called "i", as it has no closing tag, the rest of the line is modified by "i".

Another example:

This part of the subtitle is not modified by "i" taxo but the rest of the line is.

- Now, not all programs understand tagged subtitles, and some programs, which is more dangerous, "believe" who understand them but do not do it well. In principle, a program understands "well" a subtitle with tags, if this example:

≤3 This part with italic. This part without italics.

is displayed exactly as the text says, that is, the part enclosed between "i" is displayed in italics, and what is not enclosed in tags is not displayed in italics. Here we have our first problem with the Subtitle Workshop! because if this program finds a tagged.srt, it will modify it to look like this:

This part with italic.This part without italica.

Which is not what we wanted. We already found our first "incompatibility" between a program of the type that processes subtitles, and a subtitle.

Subtitles with effects

- A later step was taken by a particular program of the type that processes subtitles, which at that time was called VobSub (in current versions it is called VSFilter or Textsub). This program was the first that could process subtitles in ssa/ass format with complex tags, for example this text:

{move(80,180,300,180)}{fade(255,0,255,0,1000,3000,5000)}Jano

is modified by the "move" and the tag "fade" (When the tags are this complex, the VSFilter also calls them "commands", and others that are even more complex that we did not see in the example, are called "functions", here we are going to call "tag").

- As you may have guessed, in this case of ssa/ass with "effects", we have fewer programs that process subtitles correctly, which are only two:

- if you are going to see the subtitle in the form of "softsubs", the only one that processes them correctly is the VSFilter (ex VobSub),

- if you want to put the subtitled video on a DVD, there is no way to do a "vob" separated from the video showing the effects, the only way to show them on the DVD is to paste them into the video. The only filter that does that is TextSub (if you knew Avery Lee's Subtitler filter, it's an oldie filter that doesn't 'get' the new effects added by VobSub).

A particular problem with the VSFilter is that it is only for Windows systems, and porting your code to be cross-platform is too much work, given the poor documentation of your source code. For this reason, the MPlayer project created its own subtitle renderer of this type, called LibASS, giving the possibility to see the subtitles, regardless of the platform.

Styled Subtitles

- In addition to adding tags and making them more and more complicated, programs that read subtitles began to be able to process "styles". Why were styles necessary? In cases like this:

{fnComic Sans MSi1}Hey, wherever it goes there are only forests.
I want to go to Geffen and work.
Really? I'd like to continue hunting a little more.
{fnComic Sans MSi1}That would be good, but I heard it's a bit dangerous.
We'll be fine. Even if something happens, I'll protect them with this.

It's easier to create a "style" that is called, for example, "MyStyle", that says: "to all dialog lines that are modified by this style, show them according to the tag {fnComic Sans MSi1}", which is the same as saying "all lines of dialogue that are modified by this style, display them in Comic Sans MS font, and in italics".

The main advantage of styles is that when you want to make a change, for example make the font bigger, or change the color, or change the font because you found another one that you like better, you make the change only once, in the "MyStyle" instead of line by line.

You can find more information about the styles on the Substation Alpha page.