Your PDF is too large for UYAP: what to do

Two very different kinds of document travel under the name PDF. The first is a text-based PDF, produced by a word processor, by the UYAP editor, or by case management software. In this file the letters are still letters: it stores characters, fonts, and layout rather than a picture of the page. Size grows slowly with page count, and the text can be searched.

The second is a scanned PDF, where every page is effectively a photograph. The file does not know what the paper says; it only carries an image of it. Three settings decide how much room that image takes. Resolution, measured in dpi, matters most: doubling the scanning resolution quadruples the number of pixels on the page. Colour mode comes next, since a colour scan stores noticeably more data per pixel than a greyscale one. Third is how the image itself is compressed, and some scanners save with almost no compression by default.

The gap between the two versions of the same content is large. A petition typed on a computer uploads without trouble, while the same petition printed out and scanned in colour at high resolution can push against the limit. There is a quick way to tell which one you have: open the document and drag the cursor across a line of text. If the text highlights, the file is text-based. If nothing selects and the whole page behaves as a single block, it is a scan.

When a file exceeds the limit, the first move is to make it smaller, not to break it up. In Dosya Kalemi, the Compress PDF tool has a target-size mode with UYAP as a ready option pointing at 40 MB; that figure is written into the tool's source as the single-document limit. In target mode the file is reduced step by step until it falls under the limit you chose.

Two approaches run behind the scenes and the smaller result wins. The first is a lossless rewrite that clears out redundancy while leaving text and structure untouched. The second redraws each page as an image at a given resolution and embeds it again. In target mode the quality rungs start at 200 dpi and step down through 150, 120, 96, 72 and 60 dpi, stopping at the first rung that fits. If you pick quality mode instead, three of those rungs are named: light is 150 dpi, balanced is 120 dpi, strong is 96 dpi. One detail matters a great deal: if the lossless rewrite alone gets the file under the limit, no page is ever turned into an image and the document stays as text.

It is worth being honest about the limit itself. 40 MB is the value encountered in practice for a single document, and it can differ with the portal version, with the type of filing, and with changes made over time. Aiming comfortably below the limit rather than right at it is safer, and the warning text on the upload screen is always the most current source.

The redraw path is lossy. What you lose starts in details you barely notice and then reaches legibility itself. The first casualties are usually small-print footnotes, handwritten annotations and signatures, the fine lettering around a seal or office stamp, thin table rules, and pages that already arrived degraded by fax. Printed body text is usually comfortable at 150 dpi; at 96 dpi small digits start to run together; 72 and 60 dpi mainly serve to prove the page existed.

So do not send the result without opening it. You know where to look: dates, figures such as amounts and case numbers, signatures, the wording on a seal, notes added by hand. If those read at one hundred percent zoom, the result is usable. If they do not, stepping back to a gentler setting and covering the difference by splitting is better than filing a document nobody can read.

One point deserves emphasis: compression is not redaction. Information that someone believed was hidden by drawing a black rectangle over it is still inside the file, and shrinking the file does not remove it. You also cannot predict which path will win; if the lossless rewrite wins, the text layer survives intact and can be copied out. When something genuinely has to be removed, that is what the Redact PDF tool is for. Never use size reduction as a way of concealing content.

Splitting belongs at the point where compression can no longer get under the limit without destroying legibility. The typical case is a scan of several hundred pages that was low quality to begin with; pressing harder there buys size at a cost you cannot afford.

When you do split, put the cuts on the document's own boundaries rather than at even page counts. The petition as one file, exhibits as another, an expert report as another, bank records as another: this way the court and opposing counsel can find what they need. Halves of a single document cut mid-way through only slow readers down. The Split PDF tool takes either comma-separated page ranges you type or cut points you place by clicking page thumbnails, so you can see where you are before committing.

Naming is part of the job. Put a sequence number first so the files sort correctly in a list, then a short description of the content, then the page range the part covers in the original. Using hyphens instead of spaces and non-ASCII characters keeps the names intact across different systems. If the parts together make up a case file, an index listing what sits in which part helps the reader considerably; the Case Bundle tool can generate that index, known in Turkish practice as a dizi pusulası, a numbered schedule of the documents in a case file.

The lasting fix lives in the scanner settings. Unless colour is itself evidence, meaning a mark made in coloured ink or a stamp whose colour carries meaning, scanning in greyscale cuts the file substantially without harming readability. There is also no need to reach for high resolution: 200 to 300 dpi is enough for ordinary printed text in practice, and 600 dpi is better saved for photographs or unusually fine print. Setting the scanner's mode to document or text rather than photo has the same effect.

Pages captured with a phone are a separate source of trouble. A camera records in full colour at high resolution, the paper sits crooked, and the lighting is uneven. Using a scanning app instead of the plain camera crops the edges and shrinks the file. If pages did come out crooked, the Straighten Scan tool measures and corrects them, and because it does not convert pages to images it leaves any existing text layer alone.

Best of all is not scanning at all. If the document is already digital, a UDF file from UYAP or a word processor document, converting it straight to PDF keeps the file small and keeps the text searchable. Printing it out and scanning it back in throws away both properties at once.