Abstract

The current study explores the special qualities of close-up dubbing in film. It looks at whether close-up dubbing represents a distinct practice in comparison to other shot types and examines how phonetic synchrony (lip-sync), isochrony, and kinetic synchrony influence translation strategies, particularly when such processes result in shifts in meaning. Using a multimodal corpus of twelve English-dubbed Korean films, dialogue shots were segmented and analysed using ELAN software. Although it is well-recognised that shot scale affects dubbing constraints, the findings empirically demonstrate that close-up dubbing leads to a significantly higher frequency of meaning modification. These results show the heightened technical demands of aligning dialogue with detailed facial expressions and lip movements. Furthermore, qualitative case studies show that translators frequently modify dialogue in close-up sequences to improve dramatic tension and emotional expression in addition to achieving lip synchronisation. Collectively, these findings indicate that close-up dubbing is a creative adaptation process that integrates linguistic precision with cinematic techniques.

Keywords: audiovisual translation, dubbing, close-up shot dubbing, dubbing into English, multimodal corpus

1. Introduction

Close-up shots have long been recognised as a cornerstone of cinematic expression, as established by early film theory (Andrew 1976; Balázs 1970). Bousé (2003: 123) asserts that films without close-ups are “a dull business” because the absence of intimate visual focus hinders audiences from forming deep emotional connections with characters. Modern film studies confirm that close-ups, by capturing detailed facial expressions, are vital for conveying a character’s emotional state and eliciting both intellectual and emotional responses (Carroll 1993; Tan 2005). Empirical research has demonstrated that close-ups enhance the speed and accuracy of recognising individuals and emotions (Cutting and Armstrong, 2016; Loftus and Harley 2005; Lampinen et al. 2014), suggesting that these shots offer audiences a more profound insight into the personalities and inner lives of characters.

In the field of audiovisual translation (AVT), particularly in dubbing studies, the significance of close-up shots is equally critical. Unlike medium or long shots, close-ups demand precise alignment between the dialogue and the actor’s facial movements. Dubbing in these shots must not only maintain linguistic accuracy but also ensure that the visual cues of the performance are preserved. Chaume (2004a, 2004b, 2008, 2012) stresses that dubbing professionals must synchronise key phonetic elements, such as open vowels and bilabial consonants, with the precise moments when an actor’s mouth moves in close-up shots. Despite the acknowledged importance of close-up dubbing, research in this area has been limited, and there has been little empirical investigation into how the specific characteristics of close-up shots influence translation strategies.

Building on established dubbing scholarship (Bosseux 2015; Chaume 2004a, 2004b, 2008; Spiteri Miggiani 2019), this study empirically investigates how synchronisation constraints driven by shot scale, particularly in close-up sequences, impact specific translation strategies. On this basis, the study formulates two research questions. First, how do synchronisation requirements affect the translation of dialogue in close-up shots, particularly in terms of meaning modifications? Second, what additional factors, beyond synchronisation, contribute to modifications in close-up dubbing, especially when the intended meaning is transformed? These questions aim to fill the gap in current research by exploring the unique challenges and creative adaptations required in close-up dubbing.

To address these questions, this study employs a comprehensive dataset drawn from the Multimodal Corpus of English-Dubbed Korean Films. The dataset includes twelve Korean film DVDs (Region Code 1) distributed in North America during the twenty-first century and is segmented by shot size using the ELAN tool, which enables multi-layered annotation of audiovisual content. Although synchronisation constraints in close-up dubbing are theoretically acknowledged, this study employs a quantitative methodology to empirically validate and precisely measure these effects across a broader dataset. By quantifying the incidence of different strategies (e.g., modifications versus reduction), the analysis enables statistical comparison between shot types and offers empirical rigor. Moreover, numerical analyses make it possible to identify patterns - such as the prominence of synchronisation-driven modifications in close-up shots - that remain largely speculative in qualitative research alone.

The analysis is structured in three stages: first, applying ten established dubbing translation strategies to categorise dubbing dialogue according to shot type; second, quantitatively assessing the impact of synchronisation on modifications in close-up shot dialogue; and third, investigating other factors that lead to modifications when synchronisation is not the primary influence. Ultimately, this study aims to determine whether close-up dubbing is a distinct practice and to identify the factors that mould translation strategies in close-up shots.

2. Literature Review

2.1 Previous Research in Dubbing

Subtitling and dubbing remain the two most prevalent modalities for translating audiovisual content (Chiaro 2009) and they continue to be the most extensively studied topics in AVT research (Wang and Jalaian Daghigh 2024). Previously, there was a clear national division in the preference for subtitling or dubbing (Ranzato and Zanotti 2019: 1). However, this once-clear distinction is now becoming increasingly blurred. Matamala (2018) argues that the conventional division between “dubbing countries” and “subtitling countries” has largely become a vestige of earlier practices. Rather than being constrained by established national norms, viewers now have more freedom to choose the translation mode that best suits their needs, largely due to recent technological advances.

While the demand for dubbing is steadily increasing, its representation in AVT research remains significantly lower compared to subtitling (Matamala et al. 2017: 12–14). In English-speaking regions, subtitling continues to be the most common form of AVT, with dubbing comparatively less prevalent (Perego and Pacinotti 2020: 43). Owing to historically unfavourable audience perception and a relative lack of scholarly interest, dubbing has not received much attention in Anglophone nations. Dubbing has often been dismissed as “inauthentic” (Ranzato and Zanotti 2019: 3) in English-speaking countries, where there is a vast supply of English-language content and a relatively small market for foreign-language products. When badly dubbed English versions of foreign films, particularly Hong Kong martial arts films, were shown in the United States in the 1960s and 1970s, American viewers began to react negatively to dubbing. The subpar quality of these dubs often led to mockery, with both audiences and critics deriding them as “otherworldly dubbing” (Bordwell 2000: 207).

With the growth of video-on-demand platforms such as Netflix, global demand for non-English films and dramas has significantly increased, resulting in a noticeable shift in Anglophone perception of dubbing. To encourage the consumption of non-English foreign-language content in the U.S. market, Netflix provides dubbing as the default viewing option. According to Netflix, for foreign-language series such as The Rain, Dark, and Money Heist, 85%, 78%, and 72% of U.S. viewers, respectively, chose to watch these titles with dubbing rather than subtitles (Goldsmith 2019). Consequently, the number of viewers engaging with non-English content via English dubbing is expected to continue rising globally. This growing demand for English dubbing, driven by the popularity of global streaming platforms, stresses the urgent need for further research addressing the challenges posed by the lack of well-established traditions, norms, and frameworks in English dubbing practices (Spiteri Miggiani 2021: 138).

2.2 Dubbing Studies from a Cinematographic Perspective and Close-Up Shots in Dubbing

Film studies have introduced an important new perspective to AVT research. The integration of film studies into AVT has proven instrumental in defining the sophisticated interaction between image, verbal language, and sound (Bartrina 2004). Accordingly, the value of  this interdisciplinary approach has been increasingly emphasised in AVT studies. Chaume (2004a) has been a key advocate for this perspective, proposing an integrated model that analyses dubbing from a cinematographic standpoint. His work represents a pivotal step in fostering sustained dialogue between translation and film studies. Similarly, Matamala (2010) argues for the need to examine the relationship between synchronisation and specific shot types, such as close-up shots, as well as the presence or absence of actors on-screen. Remael (2004) further emphasises that the semiotic complexities inherent in film texts must be accounted for in both AVT research and pedagogical practices.

The size of the on-screen character, as determined by the type of camera shot, plays a significant role in making the psychological connection between the audience and the character, while also influencing the actor’s performance (Tucker 2014: 10). In dubbing, shot size is particularly critical because it directly affects synchronisation, which is a fundamental technical component of the dubbing process. As the character’s mouth movements become more pronounced in closer shots, the relationship between shot size and translation strategies in dubbing becomes salient.

Camera shots are typically classified into eight categories based on their size: XCU (extreme close-up), BCU (big close-up), CU (close-up), MCU (medium close-up), MS (medium shot), MLS (medium long shot), LS (long shot), and ELS (extreme long shot) (Bowen 2017: 8–20). As shown in Figure 1, close-up and big close-up shots demand particularly meticulous attention to lip synchronisation. In these shots, the character’s facial expressions and emotions are highly visible, and even minor mismatches between dialogue and lip movements become readily apparent.

 

Figure 1. Eight types of shot sizes

Close-up shots - including close-up, big close-up, and extreme close-up - capture the character’s face from a short distance, making the details of the human face and facial expression visible. These shots occupy a privileged position in cinema as they objectify what Münsterberg (2004: 38) terms our “mental act of attention” and function as a powerful expressive resource within the language of film. Beyond their capacity for visual precision, close-ups generate what has been described as a “new dimension of emotional experience that lies wholly outside the time-space of empirical experience” (Carter 2010: xxxiii.), deepening viewers’ engagement with and understanding of fictional characters (Bálint et al. 2020).

The importance of dubbing close-up shots has been widely acknowledged in dubbing studies (Bosseux 2015; Chaume 2004a, 2004b, 2008; Spiteri Miggiani 2019). These studies emphasise the critical role of lip synchronisation in dialogue adaptation, particularly in close-up shots where any misalignment between speech and lip movements is becomes especially noticeable. The following section will explore this aspect of dubbing in detail.

2.3 Synchronisation in Dubbing

Synchronisation, or synchrony, is a key aspect of dubbing, involving alignment between the translated dialogue and the visual elements on screen. Fodor (1976) identifies phonetic synchrony (matching lip movements and voice qualities), character synchrony (aligning dubbing actor’s voice with the original character), and content synchrony (ensuring translation consistency with on-screen action). Whitman-Linsen (1992) further categorises synchrony into two broad types: visual and audio synchrony. Visual synchronisation, which primarily affects dubbing dialogue authors, includes facial expressions, utterance duration, lip movements, and syllable articulation. Audio synchrony, by contrast, encompasses voice characteristics, paralinguistic traits, prosody, and cultural variances, all of which can influence dubbing actors’ performances.

Chaume (2004b) identifies four approaches to synchronisation in dubbing: professional, functionalist, polysystemic, and cinematographic. The cinematographic approach, in particular,  conceptualises synchronisation as an integral component of the film’s complex system of signs, emphasising three main types: phonetic synchrony, isochrony, and kinetic synchrony. This perspective stresses the need to integrate translated dialogue seamlessly with on-screen visuals to preserve both audiences’ sense of immersion and realism. Accordingly, this study adopts Chaume’s typology and focuses on phonetic synchrony, isochrony, and kinetic synchrony, given its analytical breadth and practical applicability.

2.3.1 Phonetic Synchrony

Achieving phonetic synchrony in film dubbing is particularly challenging as it requires translated dialogue to closely match the on-screen characters’ lip movements. Dubbing dialogue writers and actors frequently need to adjust or add words to harmonise speech with visible mouth movements. High-quality dubbing not only preserves the acting style in the target language but also guarantees that the dubbed dialogue aligns convincingly with the original lip motions.

In English, phonetic synchronisation primarily involves labialised vowels and consonants (Spiteri Miggiani 2019). However, the bilabial plosive /b/ and the labiodental fricative /v/ are absent in Korean. As a result, Korean speakers often perceive these sounds as the lenited /p/ and may struggle to distinguish them perceptually (Schmidt 1996). Given the limited role of bilabials and labiodentals in Korean phonology, it is unlikely that these sounds are extensively or systematically used in English dubbed dialogues. Therefore, an empirical approach is required to examine how bilabial consonants are actually deployed in dubbing practice.

In English, labialised consonants include /b/, /p/, /m/, /v/, and /f/. Among these, /v/ and /f/ do not exist in Korean. The Korean equivalents of /b/ and /p/ are the bilabial plosives ‘ㅂ’, ‘ㅃ’, and ‘ㅍ’, while /m/ corresponds to the bilabial nasal 'ㅁ'. In Korean, initial /m/ sounds are produced as bilabial nasals by obstructing airflow in the vocal tract while allowing some air through the nasal cavity. The consonant ‘ㅂ’ represents a voiced bilabial plosive, whereas ‘ㅃ’ and ‘ㅍ’ are voiceless bilabial plosives, produced by completely stopping airflow and then releasing it in a burst.

In Italian dubbing scripts, bilabial sounds and vowels are often explicitly bolded to draw the attention of voice actors (Spiteri Miggiani 2019). This practice reflects the observation that lip movements are more perceptually pronounced when bilabial sounds occur at the beginning of a sentence rather than in median positions. On this basis, this study focuses on bilabial consonants when they appear as initial sounds in words or sentences uttered by on-screen characters.

2.3.2 Isochrony

In dubbing, aligning translated dialogue with the duration of on-screen mouth movements, known as isochrony, is essential. Failure to maintain isochrony can result in noticeable mismatches, such as characters’ mouths moving without accompanying speech or remaining still while dialogue continues. Such discrepancies disrupt the viewer’s suspension of disbelief and undermine overall immersion. Romero-Fresco (2009) describes this phenomenon as the “suspension of linguistic disbelief,” where audiences may tolerate certain unnatural features of dubbed dialogue. Nevertheless, maintaining appropriate speech duration remains crucial for preserving this suspension and ensuring a coherent and seamless viewing experience.

2.3.3 Kinetic Synchrony

Kinetic synchrony, initially termed “character synchrony” by Fodor (1976) and later reconceptualised as “kinetic synchrony” by Whitman-Linsen (1992), refers to the alignment of  translated dialogue with actors’ body movements. This form of synchronisation is essential in dubbing, as facial expressions and gestures can vary significantly across cultures. For instance, Italians tend to use more frequent and expressive gestures during conversation, which may appear exaggerated or unnatural when dubbed into English (Schwarz, 2011). Differences in syntactic structure across languages can further complicate kinetic synchrony. In English, French, and Italian, verbs typically follow the subject, whereas in German, verbs often appear at the end of sentences. Whitman-Linsen (1992: 36) illustrates this with the English utterance “I've had enough!” where the key semantic element “enough” is at the end. In German, it translates to “Jetzt reicht’s mir aber!,” moving the focal element (reicht’s) to the middle of the sentence. Such structural shifts may misalign with the actor’s original gestures.

A similar challenge arises in Korean-English dubbing. Korean typically follows a subject-object-verb structure, placing verbs at the end of clauses, while English adheres to subject-verb-object order. This difference also appears in noun phrases containing relative clauses; English presents the head noun first, followed by the modifying clause, whereas Korean introduces the modifier before the main noun. These syntactic divergences require careful coordination between linguistic form and bodily movement in order to maintain kinetic synchronisation in Korean–English dubbing.

2.4 Dubbing Strategies

Matamala (2010) analyses three Hollywood films dubbed into Spanish and Catalan and proposes six translation strategies for film dubbing: reduction, repetition, amplification, modification, change in order, and deletion. These strategies address important aspects of the dubbing process and provide a solid analytical foundation for examining dubbed dialogue.

Building on Matamala’s framework (strategies 1-6 in Table 1), this study introduces four additional strategies to account for the specific challenges of Korean-to-English dubbing, expanding the framework to a total of ten strategies (strategies 7-10 in Table 1). The first, addition, refers to the incorporation of lines absent from the original Korean dialogue but present in the English version. The second, matching, captures rare instances in which syllable counts align between Korean and English dubbing, a phenomenon more commonly observed in Spanish-Catalan pairs. The third, transliteration, involves retaining Korean words or expressions in the English dubbing, frequently for exclamations or address terms. Finally, partial deletion describes cases where only a segment of a line is omitted, distinguishing it from complete deletion or reduction, which solely involves reducing syllable counts.

 

Table 1.  Ten translation strategies for dubbing: Six established and four newly introduced strategies

Unlike Spanish and Catalan, Korean and English exhibit substantial linguistic and structural differences, making direct comparisons difficult. To address these disparities, this study prioritises the analysis of shifts in meaning, word order, and repetition as primary strategies, while changes in syllable count are considered only when no other modifications occur. By adapting Matamala’s framework to this language pair, the study provides a more precise and context-sensitive approach to analysing Korean dialogue and its English-dubbed translations.

3. Methods

The dataset for this study comprises the Multimodal Corpus of English-Dubbed Korean Films, focusing on Korean film DVDs (Region Code 1) distributed in North America during the twelve-first century. The corpus was compiled by the author (see Yoon, 2025). The dataset includes twelve Korean film DVDs with English dubbing: Shiri (2002), Oldboy (2005), Memories of Murder (2005), Tae Guk Gi: The Brotherhood of War (2005), The Host (2007), Haeundae (2010), Man from Nowhere (2011), I Saw the Devil (2011), Masquerade (2016), The Admiral: Roaring Currents (2016), Veteran (2016), and Train to Busan (2017). These DVDs represent the complete set of Korean films commercially available with English dubbing at the time of dataset compilation (2020). One film - War of the Arrows (2011) - was excluded due to its primary use of Manchu rather than Korean.

These DVDs were converted into digital files using the encoding programme HandBrake (2020). To ensure the accuracy of both Korean and English conversations, Korean scripts were obtained from the Korean Film Archive and cross-referenced with the films to identify and correct any discrepancies prior to transcription.

In this study, the dataset was analysed using ELAN, a multimodal tool developed for annotating audio and video data. ELAN supports complex, multi-layered annotations, making it suitable for detailed analysis of audiovisual materials. The analysis involved segmenting each film into individual shots by selecting the “Two keystrokes per annotation” mode in ELAN during playback. Although film shots are typically adjacent units, this non-adjacent segmentation approach was required due to the limited number of shots containing spoken dialogue. By pressing the Enter key at the beginning and end of each shot, a segmented line was created to accommodate the insertion of annotations. On-screen dialogue was annotated with shot size information. To differentiate between on-screen and off-screen dialogue, two separate tiers were created: a shot tier, annotated with information about shot size, a sound tier, annotated with sound type information. For the purpose of this study, the analysis focused on the shot tier, which was annotated using eight distinct shot-size categories.

Further analysis was conducted on data extracted from a multimodal corpus. First, the previously introduced ten dubbing translation strategies were applied to categorise the dubbed dialogues, and a quantitative analysis was conducted by comparing the distribution of these strategies across different shot sizes. Second, the study examined how synchronisation constraints influenced meaning modifications in close-up dialogue. Third, for cases where meaning modifications in close-up dialogues were not influenced by synchronisation, additional contributing factors were investigated.

4. Findings and Discussion

4.1 Close-Up Shot Dubbing vs. Other Shot Dubbing

Of the 12,543 dialogue shots across the twelve films, 7,625 were on-screen (61%) and 4,918 were off-screen (39%), with an average of slightly over 1,000 dialogue shots per film. This analysis focuses exclusively on the 7,625 on-screen shots. To determine whether translation strategies differ between close-up and other shots, a statistical analysis was performed. Although films are generally segmented into close-up, medium, and long shots, this study categorises them into two groups: close-up dubbing and other shot dubbing. By grouping medium and long shots together as “other shot dubbing,” the study can more clearly assess whether the unique requirements of close-up shots lead to systematically different translation strategies. As illustrated in Table 2, there is a clear variation in the frequency and distribution of translation strategies between close-up dubbing and other shot dubbing; however, it is essential to verify these differences statistically.

 

Table 2. Frequency and distribution of dubbing strategies in close-up versus other shot types

The chi-square test results for close-up dubbing and other shot dubbing yielded a result of χ (10, N = 7,625) = 71.64819, p = 0.000. This finding indicates that the choice of translation strategies is not random but is systematically influenced by shot type.

Among the various strategies, modification exhibits the greatest discrepancy, representing 28% of translation choices in close-up dubbing versus 21% in other shots, a seven-percentage-point gap. This higher reliance on modification in close-up shots can be attributed to the rigorous synchronisation requirements associated with these shots, where detailed facial expressions and lip movements demand precise dubbing adjustments to maintain both linguistic accuracy and audiovisual coherence. These findings emphasise the importance of shot-specific dubbing strategies in achieving audiovisual coherence. The following sections examine the impact of synchronisation on close-up dubbing and identify additional elements that influence how dialogue in close-up shots is translated.

4.2 Close-Up Dubbing Influenced by Synchronisation

Achieving precise synchronisation in dubbing is crucial for ensuring that audiences perceive the dialogue as naturally uttering from the characters, rather than as a translation overlay. Synchronisation is particularly vital in close-up shots, where the visibility of lip movements is most pronounced. Any discrepancy between the character’s lip movements and the dubbed dialogue in such cases can break the viewer’s immersion. This section presents examples of how dubbed dialogue was altered from the original to achieve synchronisation, focusing on lip movements, speech duration, and body language.

Close-up shots, such as XCU, BCU, and CU, accounted for 71 (51%) of the 139 instances of dubbing alterations caused by phonetic synchronisation. The analysis, which builds on earlier studies, concentrates on bilabial consonants in dialogue delivered in close-up shots.

In the film Shiri, the character names “Park Min-seong” and “Park Yong-sang” were modified in the English dubbing to “Peterson” and the command “Pull the trigger!”, respectively, leaving out the original Korean names. These modifications involve the bilabial consonant ‘ㅂ’ at the beginning of each name, suggesting that the changes were made to achieve lip synchronisation. Additionally, these examples illustrate the substitution or omission of Korean names in favour of English alternatives during the dubbing process, possibly because Korean names might not be familiar to English-speaking audiences. Since these characters are peripheral and their names are mentioned only once, such modifications are feasible.

An XCU shot captures a character’s face with such proximity that it can induce discomfort in the audience; therefore, it is typically reserved for scenes requiring heightened dramatic tension. Within the corpus, XCU shots were employed only 23 times. In the film The Admiral: Roaring Currents, the Admiral Yi exclaims, “백병전이다,” which translates to “close combat,” signalling direct physical engagement with the enemy. While “close combat” is a direct translation, the English dubbed version modifies the line to “Prepare to attack!” to align the initial bilabial plosive sound ‘ㅂ’ with the labialised consonant /p/ to match the character’s lip movement. Given the prominence of lip movements in such a tight shot, the word “prepare” is chosen to synchronise with the original Korean bilabial plosive ‘ㅂ’ sound.

In the same film, the Japanese dispatch a suicide ship toward Admiral Yi’s flagship, which is out of ammunition due to prolonged combat, and Admiral Yi remains unaware of its approach. Captured by the enemy and aboard the suicide ship, Lim Jun-young secretly ascends to the deck and, from a distance, instructs his wife to alert the flagship to the threat. As depicted in Figure 2, Lim Jun-young, in a close-up shot, loudly exclaims, “임자!” (“Imja!”).

 

Figure 2. Phonetic synchronisation in the film The Admiral: Roaring Currents (2016) [01:35:29.752]

“Imja” is a term of endearment used to address one’s wife in Korean. In English, there is no direct equivalent; typically, a spouse’s name is used instead. However, since Lim Jun-young’s wife is unnamed throughout the film, creating a name for the English dubbing is challenging. Rather than transliterating “Imja,” which might be unfamiliar to English-speaking viewers, the English dubbing replaces “Imja” with “You must!” When pronouncing “Imja,” the lips close for the “Im” sound and open wide for the “Ja” sound due to the vowel /a/ (see Figure 2). In the English dubbing, elongating and emphasising the /u/ sound in “must” achieves lip synchronisation, albeit at the cost of a modification in meaning from the Korean line.

Isochrony was maintained nearly perfectly in all the analysed texts. Even in scenes where only the mouth moves without any dialogue, the very act of matching lip length serves as clear evidence of how highly this norm is valued in film dubbing. On-mute dubbing involves adding recorded dialogue that is not present in the original audio track (Spiteri Miggiani, 2019, p 131). It is often assumed that employing on-mute dubbing in close-up shots is difficult because the character’s face and mouth are clearly visible. Nevertheless, the present analysis reveals that on-mute dubbing is also utilised in close-up shots to implement kinetic synchronisation.

In the film Memories of Murder, the protagonist, Doo-man, pursues a suspect into a mining field. After the pursuit, the suspect hides among the miners, pretending to be one of them. Doo-man spots him when he notices the suspect’s underwear, which he believes belongs to the perpetrator. He then instructs all the workers to line up, and the camera pans to show each one. Standing in front of the suspect with his back turned, Doo-man suddenly turns to the camera and nods to his fellow detective, Yong-gu, indicating that he should handle the situation.

 

Figure 3. Phonetic synchronisation in the film Memories of Murder (2005) [01:07:29.317]

As illustrated in Figure 3, Doo-man merely nods to his colleague without uttering any line. However, in the English-dubbed version, the phrase “Go for it” is added on-mute to clearly instruct Yong-gu to arrest the suspect. This close-up shot is very brief, lasting less than a second, and Doo-man’ nodding motion distracts the audience’s gaze. Thus, the use of on-mute dubbing in this close-up is impactful due to the scene’s rapid pace and the character’s movement.

The following example demonstrates how dubbing dialogue is altered to correspond with a character’s physical gestures. Table 3 presents part of Park Mu-young’s dialogue from a confrontation scene in the latter part of the film Shiri, in which Park Mu-young, a North Korean special forces agent, faces off against South Korean National Intelligence Service agent Yoo Jung-won.

Table 3. Example of kinetic synchrony in Shiri (2002)

During this dialogue, when Park Mu-young utters the phrase “아주 잘 짜여진 연극 한 편” (“a meticulously staged play”), he raises his index and middle fingers to perform the gesture shown in the figure below.

 

Figure 4. Kinetic synchrony in Shiri (2002) [01:44:48.153]

In the Korean original, the line “잘 짜여진 연극 한 편을 보고 있는 셈” (“we are merely spectators to a meticulously staged play.”) is rendered in the English dubbing as “we’re this close to starving to death.” In the English version, the phrase “this close” is synchronised with Park Mu-young’s gesture in Figure 4. Since the parallel display of the thumb and index finger does not naturally evoke the idea of a “play,” employing the demonstrative adjective “this” effectively achieves synchronisation between the physical gesture and the dialogue. This example, therefore, illustrates how the creative adaptation of dialogue in dubbing serves to enhance the synchronisation between gesture and speech.

4.3 Close-Up Dubbing Influenced by Characters’ Facial Expressions

According to Bowen (2017: 222-223), it is advisable for characters to minimise or avoid movement during close-up shots. In these shots, where the character’s face dominates the screen, facial expressions become highly visible. A comparative analysis of original and dubbed dialogues in close-up shots reveals instances where the dubbed lines were modified in response to the actor’s facial expressions. This section presents three such cases.

The first example is from the film Shiri. In this scene, Park Mu-yeong, a major from the North Korean Special Forces, calls Yu Jung-won to inform him about a bomb planted in a building in Seoul. As shown in Table 4, Park Mu-yeong mentions “Shiri,” a native Korean fish.

Table 4. Modification of dialogue due to facial expression in the English dubbed version of Shiri (2002)

During the call, Park suggests gifting the “Shiri,” which has become an estranged family member like the nation, to Yu’s girlfriend’s aquarium. In this scene, Park’s voice is heard in voice-over, while Yu’s reaction is depicted in a close-up shot, as illustrated in Figure 5.

 

Figure 5. Yu’s close-up shot in the film Shiri (2002) [01:01:18:028]

In the Korean version, Yu responds to Park’s proposal by saying, “준다면 기꺼이 받지,” which conveys a willingness to accept the offer and can be literally translated as “I would gladly accept your offer.” The English subtitles translate this as “Why not?”, maintaining a nonchalant tone. However, in the English dubbed version, the line is rendered as “I don’t care. Do what you want,” introducing a somewhat hostile tone. Given that Park and Yu previously confronted each other and are portrayed as primary adversaries in the film, a hostile expression on Yu’s face is not in tune with the acceptance implied in the Korean dialogue. Therefore, the English dubbing appears to have modified Yu’s originally accommodating response to better reflect their adversarial relationship.

In I Saw the Devil, Kyung-chul is portrayed as an exceptionally heinous character, epitomising pure evil. In a final act of vengeance for the gruesome murder of his fiancée, Soo-hyun rigs a device designed to decapitate Kyung-chul upon the opening of a door and summons Kyung-chul’s family.

 

Figure 6. Kyung-chul’s big close-up shot in the film I Saw the Devil (2010) [02:19:02.944]

As illustrated in Figure 6, a close-up shot shows Kyung-chul holding a string tied to the door handle between his teeth. As his family attempts to turn the handle and open the door, Kyung-chul screams with all his might; however, the string in his mouth muffles his cries, preventing his family from hearing him.

 

Table 5. Modification of dialogue due to facial expression in the English dubbed version of I Saw the Devil (2010)

A comparison of the original Korean dialogue and the English dubbed version, as shown in Table 5, reveals that Kyung-chul’s character is portrayed as even more malevolent in the English dubbing. In the first line, the original Korean has Kyung-chul calling his son by name, ‘Sang-won,” whereas the English dub replaces the name with the derogatory term “shithead.” Additionally, in the close-up shot depicted in Figure 6, the original line “Don’t open the door” is altered in the English dub to “I’m not scared,” seemingly to further emphasise Kyung-chul’s demonic nature, even at the brink of death. As the look on Kyung-chul’s face appears sinister, the modified English dubbed line is more closely aligned with his expression than a direct repetition of “Don’t open the door.”

The final example pertains to the film Train to Busan, illustrating how an actor’s facial expressions influence dubbing translations. Seok-woo, embroiled in a divorce, is contesting custody of his daughter, Su-an. On Su-an’s birthday, in honour of her wish to visit her mother in Busan, they board a train, only to face a zombie outbreak. Initially estranged, the duo strengthens their bond through shared adversity. Table 6 presents a dialogue between Su-an and Seok-woo from the film’s climax.

 

 

Table 6. Modification of dialogue due to facial expression in the English dubbed version of Train to Busan (2017)

In an earlier scene, Su-an hesitates during a school performance, possibly due to her father’s absence, and stops singing. Later, she confides in her father, expressing that she was “really scared” earlier that he might not show up.

 

Figure 7. Su-an’s close-up shot in the film Train to Busan (2017) [01:23:05.326]

The close-up in Figure 7 reveals Su-an’s face expressing not only fear but also the anxiety and sadness stemming from the thought of never seeing her father again. While the English subtitles translate her line as “I was so scared,” the English dubbing opts for “I was so scared and sad,” capturing the emotions displayed in Su-an’s close-up. This modification in dubbing can also be attributed to the need to comply with isochrony.

These examples illustrate how dubbing adaptations can modify dialogue to better harmonise with character portrayals and emotional expressions. In close-up shots, where facial expressions are prominent, dubbing can be finely tuned to convey the character’s emotions more effectively, ensuring that the audience’s perception remains consistent across different language versions. Such modifications show the importance of considering visual cues and character development as integral component of the film localisation process.

4.4 Modifications to Enhance Dramatic Effect in Close-Up Dubbing

Since close-up and big close-up shots prominently display a character’s face, their use must be exercised with extreme caution and be highly selective. Bowen (2017) suggests that such shots are typically reserved for situations such as love scenes or moments when the protagonist displays anger or inadvertently reveals secret information. These circumstances naturally elevate dramatic tension. It is important to emphasise that the focus of the earlier section was on how dialogue meaning is adjusted to match facial expressions, whereas this section examines additional dialogue modifications that enhance tension and dramatic effects that are not present in the Korean original. In such situations, modifications to the dubbed dialogue are sometimes made to further amplify the tension. This section presents two examples where dialogue translations in close-up shots are altered to enhance dramatic effect.

The first example is taken from a close-up shot in the film Shiri. In this scene, Park Mu-young presses Lee Myung-hyun, who appears emotionally shaken, to refrain from betrayal and to commit suicide. It is revealed that Lee Myung-hyun is, in fact, a North Korean spy whose true name is “Lee Bang-hee.” By concealing her identity and engaging with Yoo Jung-won, an employee of South Korea’s National Intelligence Service, she is about to marry him. As illustrated in Table 7, Lee Myung-hyun confesses her prolonged identity crisis and decides to end her own life.

 

 

Table 7. Modification of dialogue to enhance dramatic effect in the English dubbed version of Shiri (2002)

Myung-hyun confesses that she experienced “a brief period of foolish confusion,” yet ultimately recognised the truth that she must act against the man she loves, as she is inherently “Lee Bang-hee.” She bemoans that changing her external appearance did not suffice to alter her inner self, and she concludes by admitting that she never escaped her identity as Lee Bang-hee before pulling the trigger. In the final line, reported in Table 8, she reveals her deep despair at having failed to transform her very essence from Lee Bang-hee to Lee Myung-hyun. However, in the English dubbed version, the segment describing targeting the man she loves is replaced by a confession of self-reproach: “I’ve played Hyun so long and I guess I started to actually believe that I was her.” Thus, while the Korean original shows Lee Myung-hyun failing to redefine herself, the English dubbing suggests that she has come to truly believe she is Lee Myung-hyun. Additionally, the final line is modified to “Maybe even part of me wanted to believe it was true,” emphasising that the dubbing renders her internalised identity differently.

The final example is from a scene in Tae Guk Gi: The Brotherhood of War, in which a recruiting officer approaches the protagonist, Lee Jin-seok, who is fleeing as the Korean War breaks out (see Table 8).

Table 8. Modification of dialogue to enhance dramatic effect in the English dubbed version of Tae Guk Gi (2005)

While the officer is questioning able-bodied men between 18 and 30 years old for conscription, he encounters Jin-seok and asks his age. In the original dialogue, Jin-seok responds with “열여덟입니다” (“I’m eighteen”), but in the English dubbing of his response is modified. He replies, “Eighteen tomorrow,” thereby producing a heartbreaking effect that implies he is to be drafted only one day before his eighteenth birthday. Because the original does not mention his birthday at any point, the English dubbing appears to have been deliberately altered to add narrative interest and increase dramatic tension. Had the dubbing process focused solely on synchronisation, another option might have been chosen to match the six syllables in “열여덟입니다.” Considering that the line does not begin with a bilabial sound, strict lip synchronisation is less critical, allowing for various translations such as “I’m currently eighteen” or “As of now, I’m eighteen.” Thus, this case can be understood as a translation adaptation designed primarily to heighten dramatic tension.

The cases presented in this section demonstrate that modifications in close-up dubbing are used to boost dramatic effect and convey complex emotions more vividly. These cases collectively imply that dubbing modifications extend beyond simple lip synchronisation; they function as narrative tools that adapt emotional impact, thereby enriching the overall dramatic and psychological effect of the film.

5. Conclusions

With an emphasis on close-up dubbing, this study examined how translation strategies vary by shot type through an analysis of a multimodal corpus of Korean films with English dubbing. Out of the 12,543 dialogue shots identified, 7,625 on-screen shots were examined. The aim of the study was to ascertain whether close-ups, as opposed to medium and long shots, require distinct translation strategies. The result of the chi-square test indicates that the distribution of translation strategies differs considerably between close-up and other shot dubbing. Most notably, the modification strategy is used more frequently in close-up shots. These findings demonstrate how the dubbing process integrates  visual performance with translation to create a seamless cinematic experience.

The implications of these findings are multifaceted. First, the significant differences in translation strategies between close-up shots and other shot types demonstrate that close-up dubbing operates as a distinct practice. This observation suggests that dubbing is a creative adaptation process that must account for both visual and emotional cues. Because close-up shots focus intensely on facial expressions, dubbing must be both linguistically accurate and visually coherent in order to convey the actor’s emotions faithfully. Second, among the 576 modifications observed in close-up dubbing, only 71 cases were directly influenced by synchronisation. However, when focusing specifically on modifications driven by phonetic synchronisation, a total of 139 instances were identified, with close-up shots accounting for 71 (51%) of these cases. This suggests that synchronisation is not the primary factor influencing changes in close-up shots, even though close-up dubbing is more susceptible to synchronisation constraints than medium or long shots. Third, the increased use of adaptation in close-up dubbing demonstrates that, in order to maintain dramatic tension and achieve balance with the actor’s performance, translators frequently have to deviate from the original script. Because the actor’s facial expressions are presented in striking detail in close-up views, these visual cues might change the dialogue’s literal meaning and elicit sympathetic reactions from the viewers. Consequently, this often leads to dubbing decisions that intensify the dramatic tension. These results thus require a more comprehensive approach to dubbing, wherein visual information and cinematic techniques are tightly integrated into linguistic adaptations.

Despite its contributions, this study has limitations. The corpus comprises twelve films, which may not fully represent the diversity of dubbing practices across genres. Additionally, the analysis is restrained to the Korean-to-English language pair, which raises concerns about the extent to which findings may be generalized to other language combinations. Future research could expand the dataset to cover a wider range of films and explore other elements, including audience reception. Further insight into the decision-making procedures underlying dubbing modifications may also be obtained through in-depth interviews with dubbing professionals. Furthermore, because information about the specific dubbing studios and voice professionals involved in each film is not publicly available, this limitation is acknowledged. Given that the films span a wide range of production years (2002–2017), genres, and distribution contexts, it is reasonable to assume that multiple dubbing teams were involved. Nonetheless, the possibility that certain adaptation patterns may reflect individual or studio-specific practices rather than broad industry conventions cannot be entirely excluded. The conclusions of this study are therefore presented with this caveat, and future research to explore industry variation and professional influence more explicitly. Future research is encouraged to investigate industry variation and professional influence more explicitly. By addressing these limitations, subsequent studies can build on the present findings to develop a more thorough understanding of the complex interaction between shot type, visual expression, industry practice, and dubbing translation strategies.