The history of subtitle formats reflects the evolution of video technology itself. From early frame-based approaches to XML experiments, each format attempted to solve the challenges of its era. Understanding these legacy formats not only helps maintain older content but also reveals why modern subtitle standards evolved as they did.
MicroDVD: Early Frame-Based Styling
While many consider MicroDVD (.sub) a basic frame-based format, it actually pioneered many styling features that modern formats take for granted. Using a straightforward {start-frame}{end-frame}Text syntax, MicroDVD combined frame-accurate timing with rich text formatting capabilities.
Consider this basic example:
{0}{24}First subtitle
This is a basic subtitle, with two subtitles each lasting exactly 24 frames. If the video this subtitle is applied to has a frame rate of 24 frames per second, both of these subtitles would last for exactly one second each.
Where MicroDVD innovated however was with its extensive styling system. Using bracketed escape codes, it provided detailed control over text appearance:
{0}{24}{c:$0000ff}{y:b,u}{f:DeJaVuSans}{s:12}Hello!
{25}{48}{P:10,10}World!
These lines demonstrate several features:
- Text color in $BBGGRR format (red)
- Combined bold and underline styling
- Custom font selection
- Font size control
- Positioning (10,10)
MicroDVD supported a comprehensive set of formatting options:
- Font styles: italic
{y:i}, bold{y:b}, underline{y:u}, strikethrough{y:s} - Font customization:
{f:FontName},{s:size} - Color:
{c:$BBGGRR} - Positioning:
{P:X,Y} - Multi-line text using the
|character - Character set selection:
{DEFAULT}{H:Greek}
SubViewer: Pre-SubRip Evolution
SubViewer (.sub) represents an interesting transition in subtitle format history, existing in two distinct versions. While its exact period of use isn't well documented, we know it predates the now-ubiquitous SubRip format.
SubViewer 1.0
The earlier version supported extensive metadata through an optional information section:
[INFORMATION]
[TITLE] Title of film.
[AUTHOR] Author of film.
[SOURCE] Arbitrary text
[FILEPATH] Arbitrary text
[DELAY] 0
[COMMENT] Arbitrary text
[END INFORMATION]
[SUBTITLE]
[COLF]&HFFFFFF,[SIZE]12,[FONT]Times New Roman
00:01:00.10,00:02:00.20
Oh, no.
SubViewer 2.0
The second version simplified the format, introducing a more compact structure with explicit line breaks:
00:01:00.10,00:02:00.20
This evolution toward simplicity foreshadowed the development of later formats like SubRip (SRT), which would become the industry standard.
SAMI: Korea's XML-Based Format
SAMI (.smi) represents one of the first XML-like subtitle formats. While its origins lie with Microsoft's Synchronized Accessible Media Interchange format, it found its primary adoption in Korea.
The format uses XML-style tags for timing and styling:
<SYNC Start=0>
<P>First subtitle appears immediately</P>
</SYNC>
<SYNC Start=5000>
<font color="#FFFFFF">
<P>Second subtitle after 5 seconds</P>
</font>
Notable features include:
- Timing in milliseconds
- HTML-style font styling
- Structured, XML-like syntax
- Support for multiple languages
This structured approach to subtitle formatting would later influence modern web-focused formats like WebVTT.
RealText: Streaming Text Pioneer
RealText (.rt) represented a significant shift in subtitle technology, treating text as a streamable media type alongside video and audio in RealPlayer G2. Unlike traditional subtitle formats, RealText approached captioning as a dynamic stream rather than a static file.
The format used HTML-like markup within a window container:
<window
type="generic"
duration="13:20.9"
height="60"
width="320"
bgcolor="#808080"
wordwrap="true"
>
<font color="black" face="arial" size="2">
<center>
<time begin="00:00.0"/><clear/>Opening text
<time begin="00:04.0"/><clear/>Next subtitle
<time begin="00:13.7"/><clear/><b>Bold text</b>
<time begin="00:16.7"/><clear/><font color="#008040">Colored text</font>
</center>
</font>
Notable features included:
- Text streaming synchronization
- Precise positioning control
- Dynamic text clearing
- Rich formatting options
- Integration with SMIL presentations
RealText's approach to treating subtitles as a stream rather than static text would later influence modern streaming subtitle technologies.
DVD Subtitles: The Image-Based Era
DVD subtitles, commonly known as VobSub, marked a significant departure from text-based formats. Developed in the late 1990s, this format made DVDs adaptable for international markets by supporting multiple subtitle tracks with precise rendering.
VobSub consists of two essential components:
The SUB File
Contains the actual subtitle data as MPEG-2 image streams. Unlike text-based formats, each subtitle is stored as a pre-rendered image, ensuring exact appearance across all playback devices.
The IDX File
Controls how the subtitle images are displayed:
# VobSub index file, v7 (do not modify this line!)
# Original frame size
size: 720x576
# Settings for display
alpha: 100%
scale: 100%, 100%
smooth: OFF
# Language index in use
langidx: 0
# English
id: en, index: 0
timestamp: 00:01:25:880, filepos: 000003800
This dual-file approach offered several advantages:
- Perfect rendering consistency
- Multiple language support
- Precise positioning
- Complex styling and effects
- Compatible across all DVD players
However, it also came with limitations:
- Large file sizes
- No text searchability
- Difficult to edit or modify
- Resolution-dependent quality
Legacy and Influence
VobSub's image-based approach laid the groundwork for future subtitle formats. Most notably, it influenced the development of PGS (Portable Graphics Stream), which is still used today in Blu-ray discs. PGS built upon VobSub's core concept of pre-rendered images while adding support for higher resolutions and improved compression, making it suitable for HD and UHD content.
USF: The Failed Universal Standard
In 2002, the CoreCodec project attempted something ambitious: creating a Universal Subtitle Format (USF) that would unite the best features of all existing formats. Using XML for flexibility and Unicode support, USF aimed to be the one format to replace them all.
Here's what a USF file looked like:
<USFSubtitles version="1.0">
<metadata>
<title>The Universal Subtitle Format sample</title>
<author>
<name>SubZap</name>
<email>blog-posts@subzap.ai</email>
</author>
<language code="eng">English</language>
<date>2025-02-21</date>
</metadata>
<styles>
<style name="Default">
<fontstyle face="Arial" size="24" color="#FFFFFF" />
<position alignment="BottomCenter" vertical-margin="20%" />
</style>
<style name="MusicLyrics">
<fontstyle back-color="#550000" color="#FFFF00" bold="yes" />
</style>
</styles>
<subtitles>
<subtitle start="00:00:06.000" stop="00:00:14.000">
<karaoke style="MusicLyrics">
<k t="700"/>La! La! La! <k t="1000"/>Karaokeeeeeeeee
</karaoke>
</subtitle>
<subtitle start="00:00:12.000" stop="00:00:14.000">
Hey, stop singing!
</subtitle>
<subtitle start="00:00:14.000" stop="00:00:16.000">
You'll wake up the kids!
</subtitle>
</subtitles>
The format was ambitious, supporting:
- Comprehensive metadata
- Hierarchical styling system
- Karaoke effects
- Image embedding
- Unicode text
Why It Failed
Despite its technical sophistication, USF faced criticism, particularly from the fansub community. Compared to ASS (which used simple comma-separated values), USF was:
- More verbose and complex
- Harder for software to parse
- Difficult to edit manually
- Lacking in editing tool support
By 2010, with its final 1.1 release, USF had failed to gain widespread adoption. Only a few players like VLC implemented basic support, and the format never achieved its goal of becoming the universal standard.
Historical Significance and Legacy
USF's attempt at standardization, while unsuccessful, highlighted important lessons about subtitle format adoption. Technical superiority alone doesn't guarantee success in the real world. What matters more is ease of use, community support, and compatibility with existing tools.
The format's failure was particularly telling: despite being chosen as the default subtitle format for Matroska containers and offering XML's flexibility, USF couldn't overcome its fundamental usability problems. Its verbose syntax made hand-editing impractical, and the lack of dedicated editing tools proved fatal. While ASS might have seemed primitive in comparison, its hybrid approach of INI-style configuration, CSV-like structure, and TeX-inspired formatting had evolved to match real-world subtitle creation workflows.
Perhaps most importantly, USF's story demonstrates how subtitle formats succeed or fail based on their alignment with community needs rather than technical sophistication. The format's demise influenced future standardization efforts to focus not just on technical capabilities, but on practical considerations and existing workflow compatibility. This lesson continues to shape subtitle format development today, reminding us that the best technical solution isn't always the most practical one.
This completes our journey through legacy subtitle formats. Each format, whether successful or not, contributed to our understanding of what makes subtitle formats work in the real world.
What's Next?
Now that you understand the history behind subtitle formats, you'll be better equipped to handle format conversion challenges. Some of our next articles will explore the technical aspects of converting between different subtitle formats while preserving as much formatting and functionality as possible.
