Subtitling in Translation: What You Need to Know in 2026

When people talk about subtitles, it can feel like they’re speaking a different language. The field has its own vocabulary, and that can make an already complex process feel opaque. But the complexity is earned. Subtitling is an intricate discipline, with challenges you do not face in plain-text translation.

Here’s the reality in 2026: subtitles have shifted from an optional accessibility feature to a baseline expectation. Most Americans now watch content with subtitles at least some of the time, and younger viewers are even more likely to keep them on by default. If you produce video content, understanding how subtitling works is no longer optional (AP News).

Key terms every content producer should know 

Queuing refers to the timing of when text appears and disappears on screen. A subtitler needs to keep each subtitle on screen long enough for viewers to read it, but not so long that it distracts, lingers awkwardly, or overlaps with the next subtitle. Good queuing is invisible. When it’s done well, you forget you’re reading.

Hardsub is an export option where subtitles are permanently embedded in the video. Think of it as writing words onto each frame with a permanent marker. Viewers cannot turn subtitles off, and editors cannot easily revise them later. That sounds restrictive, but hardsubs are useful when you need a file that will play anywhere without subtitle compatibility issues.

Softsub provides more flexibility. Viewers can toggle subtitles on or off, and a single file can include multiple language tracks.

Time codes are the timestamps that control when each subtitle appears and disappears. In an SRT file, they look like this: 00:00:08,242 --> 00:00:10,687. The timing can be precise down to thousandths of a second. When someone says they are “tightening up the time codes,” they usually mean they are fine-tuning subtitle timing to better match the dialogue.

Encoding tells software how to display characters. If you’re subtitling into languages that use non-Latin scripts, encoding is critical. When encoding is wrong, characters can display as gibberish. Japanese even has a word for this: mojibake (文字化け), meaning “transformed characters.”

File formats: Choosing the right container 

Professional subtitling relies on several standards, and each fits different distribution channels.

SRT (SubRip) is the most basic and widely supported format. It contains text plus start and stop times, and it works with most media players and subtitling tools. It behaves like a simple text file and can even be opened in Notepad. The drawback is that SRT offers limited styling beyond basic line breaks.

WebVTT was designed for web video and HTML5 playback. It supports styling and positioning that SRT cannot. Converting between SRT and VTT is typically straightforward, so the better choice often depends on what your platform or player expects.

TTML/DFXP is an XML-based format used by major streaming services. It supports rich styling, positioning, and multilingual tracks. Many platforms also require specific TTML profiles for delivery.

SCC and EBU STL are legacy formats used in broadcast television workflows. SCC is common in North America, while European broadcasts use EBU STL. Both still matter when you distribute through traditional TV channels.

Timing and reading speed: The science behind good subtitles 

Quality subtitles are not just translated text on a screen. They must match how viewers actually process information while watching video.

Reading-speed limits exist for a reason. Many guidelines recommend keeping subtitles around 15 to 20 characters per second for adult viewers. According to the Netflix Timed Text Style Guide, the platform sets a maximum of 20 characters per second for adult content and 17 for children’s programming.

Eye-tracking research shows what happens when subtitles move too fast. At 20 characters per second, viewers miss roughly 20% of words. At 28 characters per second, they skip closer to 25%. When subtitles flash by too quickly, viewers resort to surface reading and miss information. Worse, they spend so much effort reading that they cannot follow the visuals (AUSIT blog).

Professional subtitlers handle this by editing down text when necessary. Condensing or paraphrasing dialogue to meet reading-speed limits is better than forcing viewers to choose between reading and watching.

Other common guidelines matter, too: subtitles should stay on screen for at least 1.5 to 2 seconds, rarely exceed 6 to 7 seconds, and generally be limited to two lines maximum.

Subtitles vs. captions: Understanding the difference 

These terms get used interchangeably, but in professional contexts, they mean different things and serve different accessibility needs.

Subtitles assume the viewer can hear the audio but needs the dialogue in text form, often for translation or clarity. Subtitles typically exclude sound effects and speaker identifiers.

Closed captions (CC) are designed for viewers who cannot hear the audio. Captions include all meaningful audio information: spoken dialogue, music cues, sound effects (for example, “[door slams]”), and speaker identification when the speaker is not obvious.

Subtitles for the Deaf and Hard of Hearing (SDH) combine both approaches. SDH looks like subtitles but includes the extra information deaf or hard-of-hearing viewers need, such as “[ominous music playing]” or speaker labels for off-screen dialogue.

How AI has changed subtitling workflows 

The past few years have reshaped how subtitles get produced. Automatic speech recognition often generates a first-pass transcript, and humans then review, correct, and format it. Many organizations now use auto-generated captions as a starting point before human QA.

AI translation is also part of many multilingual subtitle workflows. As content distribution goes global, more producers localize into multiple languages in parallel instead of one at a time.

Live events have advanced even further. Real-time AI speech translation can allow presenters to speak in one language while the audience receives live translated audio and captions in their preferred language. This technology works well for webinars, town halls, and international conferences where pre-produced subtitles are not practical.

The best workflow is usually hybrid: AI for speed, plus human expertise for nuance, cultural accuracy, and final polish.

Why viewers actually use subtitles 

Viewer behavior is a big reason subtitling has become essential, and hearing loss is not the main driver.

The most common reason is simple: audio is hard to understand. Actors speak softly, background music competes with dialogue, and modern mixing often prioritizes atmosphere over clarity.

Other common reasons include understanding accents, watching in noisy environments, and following along while multitasking. Only a small percentage of viewers use subtitles primarily for language learning.

Subtitles also improve engagement. Videos with subtitles tend to have higher completion rates, and subtitles let viewers watch without sound in public spaces or catch details they might otherwise miss.

Short-form video has accelerated this shift. On platforms like TikTok, Instagram Reels, and YouTube Shorts, many users scroll with sound off. On-screen text and captions are often the difference between being skipped and being understood. For brands reaching international audiences, translated captions can expand reach without the cost of dubbing. Learn more about how to reach global audiences through social media.

The legal landscape in 2026 

Consumer demand would make subtitling important on its own. Legal requirements increasingly make it mandatory.

In the United States, the Department of Justice published a final rule in April 2024 requiring state and local government websites and mobile apps to meet WCAG 2.1 Level AA accessibility standards, which include providing captions for video content. Compliance deadlines of April 2026 for larger entities and April 2027 for smaller ones mean public sector video content now must have captions (ADA.gov rule page).

In Europe, the European Accessibility Act (Directive 2019/882) took effect on June 28, 2025. It requires streaming services, broadcasters, and digital platforms to support accessibility requirements, including captioning and subtitling obligations as implemented by member states (EUR-Lex).

Across regions, the direction is consistent: accessibility is not negotiable, and quality expectations are rising alongside legal obligations.

What this means for your content strategy 

Subtitling is no longer a post-production add-on. It is a core component of content strategy across industries, including media, education, corporate training, and government communications.

Build subtitling into your production process from the start. Choose file formats that match your distribution channels. Use AI tools to accelerate workflows, but keep human quality control in the loop. Provide same-language captions for accessibility and translated subtitles to expand global reach.

The audience that benefits extends far beyond people with hearing impairments. When you invest in quality subtitles, you are serving a much larger share of your viewers.

Need real-time captions and translation for your next meeting or event? Argo Translation’s AI Speech Translator generates live translated audio and captions in 60+ languages. Participants can access translations on any device, and you can upload custom glossaries for industry-specific terminology.