Files
SubMiner/docs-site/subtitle-annotations.md
2026-03-10 19:47:16 -07:00

6.6 KiB
Raw Blame History

Subtitle Annotations

SubMiner annotates subtitle tokens in real time as they appear in the overlay. Four annotation layers work together to surface useful context while you watch: N+1 highlighting, character-name highlighting, frequency highlighting, and JLPT tagging.

All four are opt-in and configured under subtitleStyle and ankiConnect.nPlusOne in your config. They apply independently — you can enable any combination.

N+1 Word Highlighting

N+1 highlighting identifies sentences where you know every word except one, making them ideal mining targets. When enabled, SubMiner builds a local cache of your known vocabulary from Anki and highlights tokens accordingly.

How it works:

  1. SubMiner queries your Anki decks for existing Expression / Word field values.
  2. The results are cached locally (known-words-cache.json) and refreshed on a configurable interval.
  3. When a subtitle line appears, each token is checked against the cache.
  4. If exactly one unknown word remains in the sentence, it is highlighted with nPlusOneColor (default: #c6a0f6).
  5. Already-known tokens can optionally display in knownWordColor (default: #a6da95).

Key settings:

Option Default Description
ankiConnect.nPlusOne.highlightEnabled false Enable N+1 highlighting
ankiConnect.nPlusOne.refreshMinutes 60 Minutes between Anki cache refreshes
ankiConnect.nPlusOne.decks [] Decks to query (falls back to ankiConnect.deck)
ankiConnect.nPlusOne.matchMode "headword" "headword" (dictionary form) or "surface" (raw text)
ankiConnect.nPlusOne.minSentenceWords 3 Minimum tokens in a sentence for N+1 to trigger
subtitleStyle.nPlusOneColor #c6a0f6 Color for the single unknown target word
subtitleStyle.knownWordColor #a6da95 Color for already-known tokens

::: tip Set refreshMinutes to 1440 (24 hours) for daily sync if your Anki collection is large. :::

Character-Name Highlighting

Character-name matches are built from the active merged SubMiner character dictionary, which auto-syncs character data from AniList for your recently-watched titles. Matching names are highlighted in subtitles and become clickable for full character profiles — portraits, roles, voice actors, and biographical detail.

How it works:

  1. Subtitles are tokenized, then candidate name tokens are matched against the character dictionary via Yomitan's scanning pipeline.
  2. Matching tokens receive a dedicated style distinct from N+1 and frequency layers.
  3. This layer can be independently toggled with subtitleStyle.nameMatchEnabled.

Key settings:

Option Default Description
subtitleStyle.nameMatchEnabled true Enable character-name token highlighting
subtitleStyle.nameMatchColor #f5bde6 Color used for character-name matches

For full details on dictionary generation, name variant expansion, auto-sync lifecycle, and configuration, see the dedicated Character Dictionary page.

Frequency Highlighting

Frequency highlighting colors tokens based on how common they are, using dictionary frequency rank data. This helps you spot high-value vocabulary at a glance.

Modes:

  • Single — all highlighted tokens share one color (singleColor).
  • Banded — tokens are assigned to five color bands from most common to least common within the topX window.

SubMiner looks up each token's frequencyRank from term_meta_bank_*.json files. Only tokens with a positive rank at or below topX are highlighted.

Key settings:

Option Default Description
subtitleStyle.frequencyDictionary.enabled false Enable frequency highlighting
subtitleStyle.frequencyDictionary.topX 1000 Max frequency rank to highlight
subtitleStyle.frequencyDictionary.mode "single" "single" or "banded"
subtitleStyle.frequencyDictionary.matchMode "headword" "headword" or "surface"
subtitleStyle.frequencyDictionary.singleColor Color for single mode
subtitleStyle.frequencyDictionary.bandedColors Array of five hex colors for banded mode
subtitleStyle.frequencyDictionary.sourcePath Custom path to frequency dictionary root

When sourcePath is omitted, SubMiner searches default install/runtime locations for frequency-dictionary directories automatically.

::: info Frequency highlighting skips tokens that look like non-lexical noise (kana reduplication, short kana endings like ), even when dictionary ranks exist. :::

JLPT Tagging

JLPT tagging adds colored underlines to tokens based on their JLPT level (N1N5), giving you an at-a-glance sense of difficulty distribution in each subtitle line.

How it works:

SubMiner loads offline term_meta_bank_*.json files from vendor/yomitan-jlpt-vocab and matches each token's headword against the bank entries. Tokens with a recognized JLPT level receive a colored underline.

Default colors:

Level Color Preview
N1 #ed8796 Red
N2 #f5a97f Peach
N3 #f9e2af Yellow
N4 #a6e3a1 Green
N5 #8aadf4 Blue

All colors are customizable via the subtitleStyle.jlptColors object.

Key settings:

Option Default Description
subtitleStyle.enableJlpt false Enable JLPT underline styling
subtitleStyle.jlptColors.N1N5 see above Per-level underline colors

::: tip JLPT tagging requires the offline vocabulary bundle. See JLPT Vocabulary Bundle for setup instructions and file locations. :::

Runtime Toggles

All annotation layers can be toggled at runtime via the mpv command menu without restarting:

  • ankiConnect.nPlusOne.highlightEnabled (On / Off)
  • subtitleStyle.nameMatchEnabled (On / Off)
  • subtitleStyle.enableJlpt (On / Off)
  • subtitleStyle.frequencyDictionary.enabled (On / Off)

Toggles only apply to new subtitle lines after the change — the currently displayed line is not re-tokenized in place.

Rendering Priority

When multiple annotations apply to the same token, the visual priority is:

  1. N+1 target (highest) — the single unknown word in an N+1 sentence
  2. Character-name match — dictionary-driven character-name token styling
  3. Known-word color — already-learned token tint
  4. Frequency highlight — common-word coloring (not applied when N+1/character-name/known-word already matched)
  5. JLPT underline — level-based underline (stacks with the above since it uses underline rather than text color)