How Do Streaming Services Like Netflix Decide When to Use Dubbing or Subtitling?

Dubbing and subtitling are both viable content localization solutions, but they differ in terms of cost, time, workflow, and the resources required. With the launch of new streaming platforms including Disney+ (2019) and HBO Max from Warner Media (2020), localization demand has risen as platforms seek new ways to succeed.

Dubbing involves substituting the original audio performance with a different language version that is fully acted out. Sometimes lip-sync dubbing is required and dictates that the new audio matches the on-screen actors’ lip movements as much as possible.

Subtitles are a text version of the speech that occurs in a video which appears on screen to allow viewers to engage with the content even if they do not understand the original spoken language in the video.

For YouTubers like MrBeast and streaming platforms such as Netflix, the answer is more languages and more global original content. Netflix disclosed that in 2021 they subtitled seven million and dubbed five million run-time minutes.

Although part of Netflix’s international success has come from introducing audiences to content from outside their home countries, “Creating a dub or a subtitle is a complex, multi-step process” (Netflix Technology Blog post) so it is essential to choose the right option.

To Dub or to Sub?

The price tag, timeline, and resources are not the only considerations for content producers and buyers of localization services, but also the needs and expectations of the target language audience.

Video dubbing can cost up to 15 times more than subtitling, thanks to studio costs, post-production editing costs, and the high volume of resources. Dubbing costs are also influenced by the content length and the number of languages.

Resources include voice actors (potentially high-cost depending on quality), translators, script adaptors, artistic directors, and sound engineers. Intense organization is required for this. 

The timeline generally takes longer than subtitling due to the multiple-stage process and the number of people.

Comparatively, subtitling is a simpler process and involves a smaller team. It can be done remotely as transcribers and subtitlers simply need access to the right software. Fewer resources means there are less organizational costs and the timeline is generally shorter.

The cost of subtitling depends on the complexity of the content and turnaround times, and while subtitling may be more cost-efficient than dubbing, using human translators to localize subtitles is expensive, and the price increases for low-resource languages. In addition, quality evaluation can be “as expensive as generating the translation itself”, according to Amazon Prime Video researchers.

Dubbing is generally considered the gold standard of content localization; high-quality dubbing can appear completely natural resulting in a better viewing experience. However, pacing, timing, and lip-syncing can lead to less accurate translations.

Great Expectations

Traditionally countries have been perceived to have a preferred method of language adaptation, meaning the audience is accustomed to viewing video content in a particular way. For example, Spain is usually considered a dubbing country, while Greece prefers subtitles.

David Orrego Carmona from the University of Warwick told SlatorPod that this traditional division of countries may no longer be applicable. Carmona explains that people generally watch dubbed content if they want to relax, whereas subtitles allow viewers to be more connected to a product, enable a deeper understanding of the content, or help with language learning. He continues that familiarity, habituation, and how individuals interact with content all affect why people prefer subtitles to dubbing and vice versa.

Rising Demand and the Effect of Automation

As demand for content increases so too does the need for media localization. Zoo Digital suggests localization is a comparatively low-cost “way to reach larger audiences”, estimating that media localization expenditure was less than 3% of content budgets (2022). 

Other demand drivers in subtitling and dubbing include the emergence of markets in central and eastern Europe, preference for shorter content among young people which has a higher proportion of dialogue, and the rise in non-English source content which is increasing demand for into-English-dubbing which, according to Sofía Sánchez-Mompeán, “is experiencing a Netflix-driven revival in non-dubbing countries”.

Various levels of automation are implemented into subtitling and dubbing workflows to help ease the strain, for example, machine translation. In 2018, Iflix told Slator they were already heavily investing in MT. Netflix has also disclosed they are employing speech-to-text (STT) technology to create English templates for subtitling and Disney has shown interest in synthesized voice technology reflecting the growing demand for dubbing across subscription video on demand SVOD platforms.

Amazon continues to pour resources into in-house research on subtitling and dubbing — both human and AI. Slator analyzed the current quality of AI dubbing, which, according to Simon Constable​, SVP of Global Language Services at Visual Data and President of the Entertainment Globalization Association (EGA), is “very far away” from being useful for film and television series.

In contrast, HOOQ, an entertainment platform made in Asia for Asia, launched SmartDub in 2019, an AI-based dubbing solution applying voice morphology to traditional text-to-speech processing.