Blog · 6 min read

The drawing set arrives as one 300-page PDF

· PDFeyes Blog

Ask anyone who handles drawing sets what they do when a 300-page PDF arrives and the answer is usually some version of: scroll, screenshot the page numbers into a note, and scroll again next week.

IFC_Set_Rev3 312 pages A-101 FIRST FLOOR PLAN A-102 SECOND FLOOR PLAN S-101 FOUNDATION PLAN M-101 MECHANICAL PLAN E-101 LIGHTING PLAN named from the title block
The number is already printed on every sheet. Splitting by page number throws it away.

The information is already there

This is what makes it irritating rather than merely tedious. Every sheet in that PDF is labeled. The architect printed A-101 on it, the structural engineer printed S-201 on theirs, and the number sits in the same place on every sheet because that is what a title block is for.

The file just does not carry it. A PDF page has an index, not a name, so the moment the set is assembled the numbering exists only as ink on the page.

Why the obvious tool does not exist

Splitting a PDF is trivial, and dozens of sites do it. Reading the sheet number is the part nobody bothers with, and there is a specific reason it is harder than it looks.

A floor plan is full of numbers shaped exactly like sheet numbers. Room 101. Detail 3/A-501. Callout 2/S-201. Scan the whole page and you find a dozen candidates, and picking the wrong one gives you a folder that looks organized and is not, which is worse than no folder at all.

The answer is to look only where the number is actually printed. The title block is a boxed strip in a predictable place, and everything outside it can be ignored.

The confidentiality problem

Drawing sets are frequently confidential before bid. Uploading a client’s set to a free website to get it split is a conversation nobody wants to have with a project manager, and it is a real reason people keep doing this by hand.

A tool that runs in the browser removes the question. The set is opened into your own computer’s memory, split there, and never sent anywhere. You can prove it by disconnecting from the internet and splitting anyway.

What it does not solve

Scanned sets have no text to read. Sets drawn without a title block convention have no number to find. And splitting the wrong revision is still splitting the wrong revision, faster.

But the part where somebody spends an afternoon renaming page-147.pdf to M-301.pdf was never necessary. The name was on the page the whole time.

Drop a set and see every sheet name before anything is split. Nothing is uploaded.

Open the drawing set splitter