Repository navigation
Reading Documents
| Format | Extensions | How it's read |
|---|---|---|
.pdf |
Text PDFs. Running headers, footers and page numbers are removed, and lines are joined back into paragraphs. | |
| Word | .docx |
Paragraphs and tables in document order; tables are read row by row. Macros and embedded objects are ignored. |
| EPUB 2 and 3 | .epub |
The book's files in reading order. Scripts and styles are ignored. |
| Markdown |
.md, .markdown
|
Links are read as their text and images as their alt text. Code blocks are skipped with a short spoken note. |
| Plain text | .txt |
UTF-8, UTF-16 and Windows-1252 text. |
| Pasted text | none | Treated like plain text. |
A file's contents must match its extension: a text file renamed to .pdf, for example, is refused.
Password-protected PDFs are refused too; remove the password and upload the file again.
TTSNative doesn't keep the uploaded file. It extracts the text on your computer, stores that in its library, and deletes the upload.
| Limit | Default | Environment variable |
|---|---|---|
| File size | 25 MB | LTTS_MAX_UPLOAD_MB |
| Extracted text | 2,000,000 characters (about 330,000 words) | LTTS_MAX_TEXT_CHARS |
| PDF pages | 2,000 | LTTS_MAX_PDF_PAGES |
| Documents in the library | 500 | LTTS_MAX_DOCUMENTS |
| Time to process a file | 90 seconds | LTTS_EXTRACTION_TIMEOUT_S |
The server enforces these limits; see Settings and Models for how
to change them. The web app also checks for 25 MB itself before uploading, so a higher
LTTS_MAX_UPLOAD_MB doesn't currently let larger files through the upload button.
Text-based books are small: a 400-page novel PDF is typically 1–5 MB. Files over 25 MB are almost always scanned images or image-heavy layouts.
Each document is split into sections, which appear in the Contents list.
| Format | Where sections come from |
|---|---|
| The PDF's outline (bookmarks), when it has at least two top-level entries. Each chapter starts at its heading line, even partway down a page. Pages before the first entry become "Opening pages". Without an outline, each page is a section ("Page 1", "Page 2" and so on). | |
| DOCX | Paragraphs styled Title, Heading 1 or Heading 2. Deeper headings are read as headings inside a section. |
| EPUB | Each file in the book's reading order, named from the table of contents (or its first heading). |
| Markdown |
# and ## headings. Deeper headings are read as headings inside a section. |
| Text, pasted text | Short lines on their own that start with Chapter, Part, Book, Section, Prologue or Epilogue (such as "Chapter 3" or "Part Two"), or a Roman numeral on its own, such as "IV". |
Within each section, the text is split into sentences (very long sentences are split into shorter parts). Each one is generated, played and highlighted on its own, which is why the highlight always matches what you hear.
If a chapter starts in the wrong place, for example with the last lines of the previous chapter, or has the wrong name, press Adjust this section above the text.
- Title renames the section.
- Starts at lists the end of the previous section and the start of this one. Pick the sentence the section should start with; the previous section then ends just before it. The first section always starts at the beginning of the document.
The change is saved with the document, and the player and the longer pause between chapters follow it at once. Documents added before chapter detection improved can be fixed this way without uploading them again.
TTSNative reads the text layer of a PDF. It doesn't support OCR (reading text from images) yet.
- If no page has text, the PDF is refused with a message that it has no text layer and looks like scanned images.
- If only some pages have no text, the rest is added and a warning at the top of the reader says how many pages were skipped.
To listen to a scanned PDF, export the text from the original document, or run OCR on it with another tool first.
Paste text in the Library opens a dialog with an optional Title and a Text box that counts characters. Add to library saves it like any other document. Blank lines separate paragraphs; text without blank lines is read one paragraph per line. Pasted text is stored only on your computer.
- Contents. On wide screens, the list of sections sits beside the text. On narrower screens, use the Contents button next to Play. A speaker icon marks the section being read. Choosing a section shows it and moves the playback position to its start.
- Section pager. The line above the text shows "Section N of M", with Adjust this section beside it (see Fixing a chapter); the buttons below the text go to the previous and next section.
- Click a sentence to play from there.
- Player bar. Previous and next sentence, previous and next section, and the position slider.
- Keyboard. See Keyboard Shortcuts.
While narration plays, the reader follows it into the next section.
Press the magnifying glass next to Play (Search this document) and type a word, name or phrase. Results appear as you type, in reading order, each with its section and the words found marked. Case and extra spaces don't matter. Choosing a result shows that sentence and moves the playback position to it. The first 200 matches are listed, with the total count above them.
-
Bookmark a sentence while listening with the bookmark button in the player bar or the
Bkey; press it again to remove the bookmark. Bookmarked sentences have a dotted underline in the text. - The Bookmarks tab, next to Contents in the side panel (or in the Contents dialog on smaller screens), lists them in reading order. Choose one to jump back to it.
- Notes. Add a note (or Edit note) keeps up to 2,000 characters with a bookmark.
- Export notes saves every bookmark and note as a Markdown file, grouped by section.
Bookmarks are stored with the document on your computer, are deleted with it, and stay on the right sentence when you fix a chapter.
The moon button in the player bar stops playback after 5, 10, 15, 30, 45 or 60 minutes, or at the end of the section being read. While a timer runs, the button shows the minutes left (or "End of section"). Playback always stops between sentences, ready to carry on from the next one; Off cancels the timer.
Settings → Appearance changes how the text looks on this device: Text size (Small to Extra large), Line spacing (Compact, Normal, Relaxed) and Typeface (Serif or Sans serif), with a preview.
Your position in each document is saved automatically as you listen. Library cards show how much you've heard ("New" if you haven't started), and the Continue listening card offers Resume for your most recent unfinished document. Opening a document puts you back where you stopped; press Resume to carry on.
TTSNative has two ways to change how fast you hear the text:
| Playback speed | Speaking pace | |
|---|---|---|
| What it changes | How fast the audio plays; pitch is preserved | How quickly the voice itself speaks |
| Range | 0.75×, 0.9×, 1×, 1.1×, 1.25×, 1.5×, 1.75×, 2× | 0.70× to 1.40×, in steps of 0.05 |
| Where | Speed button in the player bar, [ and ], Settings → Playback |
Settings → Playback |
| Applies to | Every voice | Kokoro voices |
| When it applies | Immediately | New audio is generated at the new pace |
With more than one document, the Library has a search box (titles and file names) and a Sort by menu: Last read or added (the default), Title, Date added or Least listened. The sort is remembered on this device.
Use the bin icon on a document's card in the Library and confirm. The document, its reading progress and its bookmarks are permanently removed from your computer.
Using TTSNative
- Installation
- Getting Started
- Reading Documents
- Exporting Audio
- Keyboard Shortcuts
- Settings and Models
- Other Devices
Voices
Help
Project