These guidelines are for people familiar with SDH captioning and provide a short addition to our standard transcription guidelines to understand what’s expected when captioning SDH at Happy Scribe.

Any SDH project at Happy Scribe requires the same proofreading as with non-SDH subtitling, with the addition of:

• Dialogue specifications
• Atmospherics
• Speaker labelling
• Music notation

Technical requirements

Technical requirements are, unless otherwise indicated:

• 42 characters per second (CPL)
• 2 lines of subtitle, with preferably the top subtitle being shorter (but give preference to proper line breaks)
• 20 characters per second (CPS)
  • Stretch subtitle by 12 frames or 0.5 seconds past their original out-time to improve CPS.
  • Truncate subtitle by slightly simplying speech (only sparsely do this).
  • Allow CPS to reach 23 CPS as a last resort. No subtitle may reach a higher number.
• Subtitles should last at least 1 second and max. 7 seconds. Atmospherics should never last longer than 3 seconds, even if the sound they describe may be longer.
• All subtitles have a mandatory subtitle gap of at least 2 frames between each other (this is force-enabled).
• Ensure all subtitles are snapped to shot changes according to the requirements (this is force-enabled, but some automatic subtitles may need adjustments to snap).

<aside> ⚠️

If other presets are specified by the style guide, be sure to follow those presets instead.

Review the Transcription and Subtitles Guidelines for language-specific items and for an order of priorization in subtitle formatting.

</aside>

Dialogue

• Dialogue should captured as per normal subtitling standards, but unlike clean-read should include a selected amount of verbatim cues:

  • Filler words (offset by commas).

  1	But, like, have you considered it?
  2	So, yeah, it’s whatever.

  • False starts of 3+ words (separated with an ellipsis).

  1	There is something about…
  2	There is just something
  about that song!

• Atmospherics that interrupt dialogue should be preceded and succeeded by ellipses.

  | 1 | I just… [sighs heavily] | | --- | --- | | 2 | …need to get this off my chest. |

Atmospherics

• Sound descriptors are used for sounds made by the surrounding environment, by speakers or when instrumental music plays.

• They should be inserted if dialogue allows it and if they provide relevance (i.e. are not clearly seen and cannot be inferred from dialogue).

  In the below example, opt for A if reading speeds are high and if the sound is also seen on screen.

  A	Open up, it’s the Johnsons.
  B	-[Knocking]
  -Open up, it’s the Johnsons.

• Atmospherics should last a minimum of 1 second, just like other subtitles. Atmospherics should last a maximum of 3 seconds, even if the sound is heard for longer.

• It is not necessary to add an atmospheric simply because a sound is heard. Focus on atmospherics that introduce the mood or set the backdrop for the scene.

• Atmospherics are formatted in lower-case formatting, with the exception of proper nouns.

  1	[sombre music]
  2	[Muslim prayer]

• Any verbs should either be in the present simple (if sudden, punctuated, without a clear start/end, or if it benefits the CPS)

  1	[doorbell rings]
  2	[glass shatters]
  3	[tires screech]
  4	[lion roars]

  or should contain a present participle without the main verb (if prolonged):

  1	[people singing]
  2	[Spanish anthem playing]
  3	[birds chirping]
  4	[car engine revving]

• When there’s room, prefer to add descriptors that give a more refined representation of the sound.

  ❌	[door opens]
  ✅	[door slowly creaks open]

• However, when CPS or CPL is high, prefer to be concise to meet reading speeds.

  | ❌ | -[people shouting loudly in the distance] -[ominous music playing] | | --- | --- | | ✅ | -[people shouting] -[ominous music] |