Bionic Reading Converter

How does the converter turn ordinary text into bionic reading?

Paste Reading is useful. into Input text. With the initial 40% setting and the short-word threshold at 3, the preview becomes Reading is useful. Conversion updates as you type or change an option; there is no separate convert button. Spaces and punctuation remain in the output.

The percentage determines a prefix length, rounded upward: bold characters = ceiling(word length × percentage ÷ 100). The short-word rule runs first, so “is” is fully bold rather than partly bold. The table shows the seven-character word “reading” at each preset, with the threshold left at 3.

Bold prefixes for “reading” at the available percentage presets

Setting	Calculation	Preview
30%	ceil(7 × 0.30) = 3	reading
40%	ceil(7 × 0.40) = 3	reading
50%	ceil(7 × 0.50) = 4	reading
60%	ceil(7 × 0.60) = 5	reading
70%	ceil(7 × 0.70) = 5	reading

Adjacent settings can therefore produce identical prefixes. Raising the percentage does not necessarily add another bold character to every word. Copy as HTML exports the first example as <b>Rea</b>ding <b>is</b> <b>use</b>ful., while Copy as Markdown inserts double asterisks around those same prefixes.

How do you choose the bold percentage and short-word threshold?

Percent of word scales emphasis with word length. The preset buttons offer 30%, 40%, 50%, 60% and 70%; Bold percentage extends from 10% to 90% in steps of five. Fixed first letters instead offers one through six leading letters, capped at the word's length. This makes the prefix length predictable across longer words.

Min word length to bold is a full-bold threshold, despite its name. Its choices run from 1 to 5 characters, with 3 selected initially. A threshold of 4 fully bolds “text”; lowering it to 3 lets the percentage or fixed-letter rule split that word. This threshold applies in both modes. Change one setting at a time so you can see which rule produced the difference.

What does Clean pasted PDF text actually change?

This option cleans text already copied from a document; it does not open a PDF or recognize text in an image. It expands the ligatures fi, fl, ff, ffi and ffl into ordinary letters. It also joins single line breaks within paragraphs and removes a hyphen followed by a line break between letters. Blank lines retain paragraph separation.

Header and footer removal uses repetition rather than PDF layout information. A matching line must occur at least three times, touch a document edge, and have roughly regular gaps averaging at least four lines. Page-number removal requires at least three sequential number lines beside detected repeated margins. An isolated number or a header appearing twice does not meet those rules.

Review poems, lists and deliberately hyphenated text before exporting: joining lines can change their intended layout. Cleanup affects the converted output without replacing Input text. Show original displays the untouched input, so with cleanup enabled the comparison includes both cleanup and bolding differences. Switch cleanup off to compare bolding alone.

Which words and tokens can stay unbolded?

Every exclusion starts unchecked. URLs and emails recognizes scheme-based addresses, addresses starting with www., and email-like tokens. Numbers skips tokens containing numbers but no letters; a mixed token such as “A12” still follows the bolding rules. #hashtags and @mentions keeps those tokens plain. Text inside `backticks` protects paired inline backticks on one line, rather than acting as a complete code-block parser.

Custom skip words (comma separated) matches whole tokens without regard to case after trimming outer punctuation. Adding “however” also skips “However,” but does not skip “howeverlong”. Exclusions run before the short-word threshold. Ordinary whitespace is preserved; hyphenated and apostrophized words normally remain a single token, with internal punctuation included when calculating the prefix length.

How do the copy and download formats differ?

Copy as HTML produces escaped text with b tags and line breaks represented by br tags. Typed HTML is treated as text rather than executed. Copy as Markdown simply wraps bold parts in double asterisks; it does not escape existing Markdown syntax. Review the result when your source already contains asterisks or other markup.

Copy rich text writes HTML plus a plain-text fallback, letting the receiving editor choose what it accepts. If formatting disappears, inspect that editor's paste mode or use HTML source where supported. Copy Unicode bold replaces only emphasized ASCII letters with Mathematical Bold characters. Other characters remain unchanged; this is a character substitution rather than ordinary font styling.

Download .html creates bionic-reading.html with the chosen display CSS embedded in the document. Copied HTML, Markdown and rich text do not include those display settings. Show original changes only the preview: the copy buttons and download still export the bionic result. Choose the export format deliberately before moving text into another application.

How can you adjust the reading display and compare comfort?

Reading display offers Sans-serif, Serif and Dyslexia-friendly (Verdana) font choices, plus Light, Dark and High contrast themes. Font size ranges from 14 to 28 pixels, line height from 1.4 to 2.2, letter spacing from 0 to 0.2em, word spacing from 0 to 0.6em, and column width from 40 to 100ch. Reset display restores these presentation controls to their initial values.

Bionic reading here means partial-word emphasis. The converter measures neither reading speed nor comprehension and offers no clinical assessment. Compare a familiar passage with Show original and choose a comfortable presentation. The font option's label is a display choice, not a treatment claim. Use the Reading Time Calculator for a separate duration estimate or the Word Counter to inspect text length.

What are the language, size and privacy limits?

The converter uses Unicode grapheme segmentation when available to keep combined characters together, with a code-point fallback. Han, Hiragana, Katakana and Thai text additionally uses word segmentation when available; results follow the browser's segmenter. The displayed word count instead splits the original input on whitespace, so it is not a language-aware count of those segmented words.

The input field sets no fixed character cap, but conversion runs synchronously as text changes, so large passages depend on device resources. Processing stays in the browser. Only display preferences are written to localStorage; input text stays in component memory. Clear empties that input, while copied clipboard text and downloaded documents remain separate copies under your control.