Dubbed Audio Style Guide - VO-Style Dubbing
Objective
This document contains a list of required practices when creating dubbed audio in VO-style (also known as UN-style) for Netflix content. Consult your Netflix representative for any specification not covered in this document. Please refer to the Lector dubbing guidelines for treatment on Lector style dubbing.
1. Translation Requirements
1.1 Main Dialogue
- All main dialogue in the source (original) language should be translated unless specifically noted.
- Due to timing limitations, some of the dialogue may be condensed/truncated as long as it retains all essential elements of the plot.
- Please refrain from dubbing redundant words such as character names and repetitions.
- Additionally, do not recreate laughs, hesitations, reaction noises, etc.
1.2 Character Names
- Do not translate proper names (e.g., Peter, Suzanne), unless directed by production or approved by Netflix. EXCEPTION: Famous historical/mythical characters and/or nicknames that might vary from language to language.
- Local pronunciation of character names can be used in the voice-over, as long as the names are consistently pronounced throughout the series.
1.3 On-screen Text Treatment
- On-screen text should only be included if plot pertinent and should be included in the Forced Narrative file (not dubbed).
- For treatment of main title and episode title, please reference the article Episode & Main Title Translations in Subs and Dubs.
- “A Netflix Original Documentary Series” and other Originals credits translations should be included in the forced narrative file, using the Netflix approved translations.
- Opening and End Credits should not be dubbed or included within forced narrative file, as these are usually not localized. Dub studio/voice credits should be conveyed by means of a Dub Card.
1.4 Foreign Languages
- For shows where foreign dialogue is sparse throughout, or has plot pertinent importance, a forced narrative file for the foreign dialogue will be created as opposed to covering with voice-over. (e.g.: “Dirty Money”: French Canadian being spoken throughout the episode of “The Maple Syrup Heist” which was translated with forced narratives.)
- If foreign dialogue accounts for roughly 40% or more of run time (e.g.: “Chef’s Table”), this dialogue would be better conveyed with VO instead of forced narratives, as too many forced narratives causes a disruptive experience.
- Do not replicate accents and/or grammar mistakes made by talking heads speaking a language other than their native language. The use of language accents should generally be avoided, because an inaccurately applied/imitated accent may be perceived as insensitive or offensive to certain cultural and/or ethnic groups.
- Seek advice from Netflix if it is unclear how to treat foreign dialogue or accents.
1.5 Songs
- Songs should be translated and included in the forced narrative file (not dubbed) only if plot pertinent and if the rights have been granted. E.g. "Wild Wild Country"
Please obtain song clearance from your Netflix representative.
1.6 Censorship
- Dialogue (including expletives) should be rendered as faithfully as possible, without using dialect or words that would otherwise introduce or soften a level of obscenity inconsistent with the original content. Tone, register and context should match the original.
- If expletives are bleeped in the original content, do not recreate the bleep in the voice-over. This would create a jarring experience due to the likely misalignment of the bleep sounds. Instead, translate the censored expletives in a milder way, and keep the bleep in the original mix audible to convey the creative intent.
1.7 Consistency
- Key Names & Phrases glossary (KNP) creation and entry within Netflix KNP tool is required during script translation and adaptation. Translated terms should be entered into the tool and “locked” by the start of recording. Existing, locked KNP translations must be followed. Details and requirements for the KNP workflow are outlined in our Dubbing Scope of Work.
1.8 Archival Material
- To retain the authenticity, integrity and historical intent of archival source content, translate and include in forced narrative file (do not dub). Seek advice from your Netflix representative if it is unclear whether or not a line/scene is archival, or if there is a high percentage of archival that may disrupt a viewer’s experience (with too much forced narratives to read), or other related queries. Remember that each title is unique and different treatment may be requested by the Netflix representative on specific titles.
- Anything historical and filmed outside of the film/series itself is typically considered archival material. Examples of commonly used archival material are: Recorded audio from confessions/phone calls, TV news footage, court footage, speeches from political figures or activists, previously released movie/TV content, etc.
2. Casting
- Talents can record more than one voice with the following limitations:
- Talents should not perform more than one voice in the same scene or sequence.
- Talents cast for a primary voice should not be cast for any secondary voices.
- Talents cast for secondary speakers should not voice more than one significant speaking role (20+ lines) per feature or series or perform in consecutive scenes.
- While exact voice matching is not required for VO, please look to cast talent with similar traits such as age range, vocal qualities, etc.
3. Recording
3.1 Sync
- For this VO-style dubbing, allow original dialogue to be heard at normal volume for a beat--usually 1-3 seconds. Delaying the translation (leaving a few seconds with no translations at the beginning and, where possible, at the end) enhances authenticity. This may mean condensing translations to ensure the original audio is heard.
- However, if valuable information is lost because the translation is condensed too much, it is acceptable to have the VO sync with the original dialogue.
- Maintaining the same word order at the beginning and at the end can also enhance authenticity. (Applicable only for languages that allow for this technique.)
3.2 Director
- Please utilize a dubbing director for documentary and unscripted content. If you have reasoning to forgo a creative director, please make sure you reach out to your Netflix representative for approval.
3.3 Performance
Documentaries
- Use best judgment when determining tone and rhythm of speaker, as well as speed and length of time needed for the translation.
- Reminder, as mentioned in section 1.1: Efforts/reactions should not be recreated. Allow the original audio to shine through where possible.
Unscripted
- VO dubbing should not fully match the delivery of the original version. Authenticity should still be respected since the original audio can still be heard in the background. Instead, certain unscripted titles may merit more energetic performance due to the nature of the content (e.g. cooking shows, competitions). Please use best judgment when making these decisions in tone and rhythm of speaker. If in doubt, please reach out to your Netflix representative for guidance.
- Reminder, as mentioned in section 1.1: Efforts/reactions should not be recreated. Allow the original audio to shine through where possible.
3.4 Pronunciation
- See section 1.2 for treatment on character name pronunciation.
- Local pronunciation of places or foreign words is acceptable.
- Please ensure pronunciation of key terms is consistent across episodes and seasons.
4. Technical Aspects
For information on audio mix and delivery formats, please reference the Localization, Accessibility, and Dubbing Branded Delivery Specifications.
- For a 5.1 OV Printmaster, dip center channel only. Only for very loud sections or for films with very wide dynamic range, it’s acceptable to dip Left and Right as well, no more than 6db.
- For a 2.0 Printmaster, dip both channels
- OV may be manually dipped when necessary. Voiceover may not be raised above loudness spec to overcome very loud OV sections.
- Dip OV 6-12 db, per mixer discretion. Dubbed audio should be clear and intelligible with natural presence of OV dialog underneath. These are subjective choices relating to the dynamics of the OV and the perceived volume of the OV and Voiceover. Please use your best judgement for different types of content; some content is louder and more dynamic than others and the dipping, editing should accurately reflect what is best experience for that content.
- Side-chained compressors should not have attack time shorter than 2ms and longer than 15ms.
- OV level should hit Netflix LKFS spec for the territory.
- OV level should transition to/from dipped in no more than 5 secs, not rapidly but in a seamless way.
- Regarding EQ: VO should sound natural. Excessive boosting or dipping of frequencies should be avoided, except to address recording problems.
- Regarding Narration: Unless specifically instructed by your Netflix representative, please do not remove narrator stems/utilize a mix minus track. For example, in documentaries the narrator is often telling their story (as a first person narrator) and retaining their voice preserves the authenticity of the content. This is a very different function from an omniscient third person narrator that can be replaced by a local voice. If in doubt or if you have a recommendation for the best experience in your territory, please reach out to your Netflix representative for guidance.
Translations
Change Log:
2019-10-16
- Section 4 (Narration): Territory recommendation note added.
2019-07-19
- Section 4: Bullet point added on Narration.
2019-05-21
- Section 1.1: Last bullet point added to clarify reactions do not need to be recorded.
- Section 1.3: Linking out to article for Episode and Main Title Translations.
- Section 1.4: Three bullet points added with more details on foreign dialogue treatment.
- Section 1.5: Updated main title of show example.
- Section 1.6: Modified language.
- Section 1.7: KNP instructions now consistent with Dubbing SOW.
- Section 1.8: Clearer details on what is considered archival.
- Section 2: Casting section created.
- Section 3.1: Details added for expectations on sync. Second bullet point added.
- Section 3.2: Director section created.
- Section 3.3: Performance section created.
- Section 3.4: First bullet point added.
- Section 4: Fourth bullet point updated.
2018-08-13
- Section 1.4: First bullet point clarification on forced narrative treatment. Second bullet point modified on multilingual shows.
- Section 1.6: Second bullet point added for treatment of expletives.
PROVIDE FEEDBACK ON THIS STYLE GUIDE