PDF Tools — Tips & Tricks

Back to the overview: PDF Tools · Open the tool live: www.jpkc.com/tools/pdf/

The Manual explains every feature, the Examples show the workflows. This page is about what both assume but rarely state: why privacy is the real argument here, which limits you should know, and when which conversion path is the right one.

Privacy is the selling point here

The biggest difference from most online PDF services: your files never leave your device. There is no upload and no server processing your PDFs — every operation runs locally in your browser via JavaScript.

Sensitive documents are safe. Contracts, payslips, HR records, internal reports, completed applications — what is never uploaded cannot sit on someone else's server or be intercepted. With classic "merge PDF online" services, your document lands on their infrastructure.
No server for the processing. The actual work needs no server. Offline, however, it only works as long as your browser can serve the scripts from its cache — the viewer and Rich HTML reload their libraries on every run.
No account, no install. No login, no desktop software, no data trail.

If someone asks whether an online PDF tool is "okay" for a confidential file, that is precisely the answer: a purely client-side tool like this one is, an upload service is not.

The 50 MB limit and large files

PDFs in the Viewer & Editor, in Merge and Split, and images in Images → PDF may be up to 50 MB; larger ones are rejected. This is not a server limit but a safeguard for your browser's memory — all processing happens locally, after all.

Very large or page-heavy PDFs may take noticeably longer to process in the browser or, depending on the device, hit memory limits. An older laptop struggles with a 300-page scan more than a current desktop.
Workaround via Split: if you only need to work on part of a huge PDF, first extract the relevant pages (Split) and continue with the smaller excerpt.
Shrink images first. High-resolution photos bloat an Images → PDF heavily. For non-PNG/JPEG formats (WebP, GIF, BMP, TIFF) the tool rasterises via a canvas anyway and caps the edge length at 4000 px — so very large images add no value but cost memory. TIFF only works in browsers that can display TIFF themselves (Chrome and Firefox, for example, cannot); otherwise convert the files to PNG first.

Simple HTML or Rich HTML? The decision that matters

The Convert tab offers two HTML paths that differ fundamentally — the choice decides the quality of your result:

Simple HTML (pdfmake) translates the markup into real PDF text. The result is searchable, copyable and compact — but CSS styling has no effect, and the content of <style> or <script> blocks ends up as text in the PDF; remove them first. Use this when the content matters and the layout can be plain.
Rich HTML (html2pdf.js) renders the page visually, like a browser — with CSS, colours, flexbox, images. In return the output is raster: the page goes into the PDF as an image, the text is not selectable or searchable, and the file gets larger.

Rule of thumb: searchable text → Simple HTML or Markdown. Pixel-perfect look → Rich HTML. For Rich HTML, the Quality Scale option controls the resolution: 1× is fast, 2× is the default, 3× is for print — higher scaling means sharper, but slower and larger.

Do not count on JavaScript in Rich HTML: the tool strips all scripts and event handlers before rendering. Anything a script would write into the document (charts, content loaded later) is missing from the PDF. Produce such content beforehand as static HTML or as an image.

Searchable text, OCR and scans

Markdown and Simple HTML produce real text — you can search and select text in the result later. That is the invisible but big advantage over image PDFs.
There is no OCR. The Text feature (in the Viewer & Editor) and text extraction only read an existing text layer. A pure scan PDF without a text layer therefore returns almost nothing — that is not a bug but the absence of text recognition. For such cases you need a dedicated OCR tool.
The Info "Tagged" field comes from the PDF's AcroForm flag and says more about the presence of form fields than about a true tagged-PDF structure. Do not read too much real accessibility into it.
Text from password-protected PDFs: Info and Text in the Viewer & Editor cannot read files with an open password. The detour goes through Split: load the file, enter the password, use 1-end as the page selection, click Split — the downloaded copy is unprotected, and Text reads it in the viewer without trouble.

Missing characters: arrows, check marks and emoji

The Creator, Markdown and Simple HTML set everything in Roboto. The font covers Latin, Greek and Cyrillic, but no arrows, no ✓ or ☐, no other symbols, no CJK characters and no emoji. The PDF library would silently drop such characters — so after Generate PDF the tool warns you with "Not in the PDF font, left out: …" and names the affected characters.

Replace rather than lose: → becomes ->, ✓ becomes [x]. Markdown task lists (- [ ], - [x]) are set as [ ] and [x] anyway.
If special characters are a must, use Rich HTML: there the browser renders the text with the fonts available on your system — at the price of a raster PDF without searchable text.

Page selections and passwords in Merge & Split

The page syntax in Use pages (Merge) and Extract page ranges (Split) can do more than the input field suggests:

Reorder a document: Split with 3, 1, 2, 4-end moves page 3 to the front and keeps the rest behind it — a way to rearrange a single PDF without merging.
Reverse a document: end-1 outputs all pages backwards, handy for stacks that were scanned the wrong way round.
Several excerpts in one go: 1-5; 6-10; 11-end produces three files when splitting, which you fetch together with Download all (ZIP).
Duplicate pages: 1, 1 takes a page twice — say, a cover sheet that should appear at the front and the back.
Passwords are dropped: Merge and Split always save their result without password protection, even if the source was protected. Convenient for further processing, but keep it in mind before you pass the file on.
Bookmarks as a table of contents: when merging, every source file gets a bookmark. It carries the document title; only if none is set does the tool use the file name without .pdf. A meaningful title or file name therefore pays off — "chapter-3-measurements" reads better in the sidebar than "scan0042". If the source brings a meaningless title along (say, from the scanner), renaming the file does not help.

Small tricks

Reuse Creator templates. If you regularly build similar documents (quotes, minutes), save the structure once via Export JSON and load it next time via Import JSON — including embedded images. That saves re-clicking it all.
Duplicate blocks instead of rebuilding. In the Creator every block has a Duplicate button; a fully formatted table or paragraph is there a second time in seconds.
Structure first, then fill in. In the Creator, Add Block, Move up/Move down and Delete discard everything you have typed into the blocks since the last sync (a sync happens on Generate PDF, Duplicate or Export JSON, for example; the full list is in the Manual). So create all blocks in the right order first and fill them in afterwards — or click Generate PDF briefly before each rearrangement.
Order matters when merging. The merged PDF strictly follows the table order — sort before clicking Merge All, not after.
Fit Mode per image — set it last. In Images → PDF you set the Fit Mode per image in the image list — handy when a landscape photo needs "Fill" and a portrait scan needs "Fit". The Default Fit setting has no effect in the current version, and every change to the list resets all images to "Fit". So add all images first, then set the Fit Modes.
Rotate landscape scans when merging. If a file's pages lie sideways, Rotate pages by 90° in the merge table turns all pages of that file — before merging, without a detour through another program.

Combining with other JPKCom tools

PDF Tools convert existing input into PDF — and you best produce that input in the matching specialist tool:

Markdown Editor — write Markdown with a live preview, then convert it to PDF here in the Convert tab (searchable text).
WYSIWYG Editor — build HTML visually, then convert it as Simple HTML (searchable) or Rich HTML (pixel-perfect).
Graphic Editor — crop, scale and convert images to the right format before bundling them with Images → PDF.

More context: the overview for the big picture, the Manual for every feature, and the Examples for the step-by-step workflows. You can try everything directly in the tool.