Sanitizing a property deed PDF before e-discovery means physically removing hidden metadata, OCR text layers, comments, and embedded files from the document — not just covering visible names with a black box — and a recommended workflow keeps the original deed on your own device so the file never reaches a third-party server.
When a deed enters production, opposing counsel can inspect far more than the visible page. Metadata can expose the seller’s attorney, the software used to draft the document, and the exact creation date. An invisible OCR layer on a scanned deed can duplicate every word of the grantor’s name. A single margin comment can reveal attorney work product. This guide walks through a nine-step sanitization procedure built around the failure points that cause privilege waivers, with explicit warnings at each stage where attorneys commonly make a costly mistake.
What You Will Need
- The original property deed PDF (the recorded or unrecorded version you intend to produce)
- A working copy of the deed — never sanitize your only original
- A PDF tool with true redaction and metadata sanitization:
- Adobe Acrobat Pro (desktop; Redact tool and Remove Hidden Information feature)
- Sejda Desktop (local PDF editor with a redaction tool)
- One option is OctopusPDF Redact/Sanitize, a browser-based tool that processes files 100% locally
- A metadata verification tool: ExifTool (command line) or the Document Properties panel in Acrobat
- A written inventory of sensitive strings to redact (grantor/grantee names, addresses, legal description, parcel or tax IDs, loan amounts, notary details)
- A clean output folder for the produced file, separate from your working files
Step-by-Step Instructions
Step 1. Inventory the deed’s sensitive material before opening any tool
- Action: List every confidential item contained in the deed.
- Explanation: A recorded deed typically includes grantor and grantee full names, mailing addresses, a full legal description, parcel or tax ID numbers, a notary block, and possibly a purchase price or loan amount. Write this list down now — you will use it in Step 4 for redaction and again in Step 9 for verification.
- Watch out: Do not forget handwritten margin notes, stamps, or bates-number overlays on scanned deeds. They are part of the image layer and must be redacted like any other visible text.
Step 2. Create a working copy of the original deed
- Action: Duplicate the original PDF file and name the copy
Deed_working.pdf. - Explanation: Redaction and sanitization are physically destructive operations that cannot be undone after saving. The untouched original stays in a separate folder as the authentic record for your file.
- Watch out: If the deed is a scan with an OCR text layer, the working copy inherits that layer. You will remove it in Step 5.
Step 3. Strip document-level metadata
- Action: Open the working copy and remove all metadata fields: author, title, subject, keywords, creator, producer, and creation/modification dates.
- Explanation: Metadata survives normal text deletion. In Adobe Acrobat Pro, go to File > Properties > Description, delete every field, then use File > Save As with a new filename. In a local browser sanitizer, the sanitize function deletes metadata in one action rather than leaving it inside the document information dictionary.
- [Insert screenshot: Document Properties panel with all metadata fields cleared]
- Watch out: Using “Save As” with a new filename is not enough by itself. You must clear the fields themselves; otherwise, the metadata travels into the new file.
Step 4. Redact visible text with a real redaction tool, not a black rectangle
- Action: Select your PDF tool’s redaction feature, mark every sensitive string from your Step 1 list, then apply the redactions.
- Explanation: Drawing a black rectangle or using a highlighter covers text only visually. The underlying characters remain in the PDF content stream and can be extracted by any parser. True redaction removes the characters permanently and renders a solid block in their place. In Adobe Acrobat Pro, use Tools > Redact > Mark for Redaction, then Apply Redactions.
- Watch out: If your PDF viewer lets you select and copy text from underneath the black box, the document is not redacted. The tool you used was not a redaction tool.
Step 5. Sanitize the hidden layers: OCR text, comments, annotations, and embedded files
- Action: Run the sanitize function that strips metadata, comments, hidden content, OCR layers, and embedded files in a single pass.
- Explanation: Scanned deeds frequently carry an invisible OCR text layer that duplicates every word on the page. Comments and sticky notes can contain attorney work product. A proper sanitize function physically deletes these layers rather than hiding them from view.
| Sanitization removes | Example in a deed |
|---|---|
| Metadata | Author “Jane Doe – Seller’s counsel”, creation date, producer software |
| Comments and annotations | Margin note such as “argue adverse possession” |
| OCR / hidden text layer | Searchable copy of the grantor’s name under the scanned image |
| Embedded files and attachments | A title report or survey attached to the PDF |
- Watch out: If your tool presents separate checkboxes for metadata, comments, and hidden text, select all of them. Selecting only “metadata” leaves the OCR text layer intact and searchable.
Step 6. Confirm the file is processed locally, not uploaded
- Action: Check that your chosen tool runs in the browser tab or on the desktop and that no upload step occurs before processing.
- Explanation: Many free online PDF sites upload your file to their server first. Even when they promise to “delete after 1 hour,” you cannot verify it. A deed headed for e-discovery is exactly the kind of confidential document that should not spend time on infrastructure you do not control. Local processing means the servers physically cannot receive the file. Real-world testing of a local browser sanitizer recorded 0 bytes uploaded across 113 test documents.
- Watch out: If the tool shows a progress bar after a file-selection step, ask whether that bar represents local processing or an upload. If you are unsure, switch to a desktop tool or a documented local-processing browser tool.
Step 7. Save the sanitized file as a new PDF
- Action: Save the output under a new name, such as
Deed_Sanitized_2026-08-12.pdf, in your clean output folder. - Explanation: A new filename preserves the chain of custody and prevents accidental overwrite of the working copy. Keep the working copy untouched until verification in Step 9 passes.
Step 8. Verify metadata and hidden content with a second tool
- Action: Open the sanitized output in a separate tool and check both Document Properties and full-text search.
- Explanation: In Acrobat, inspect File > Properties and confirm all fields are empty. If you use ExifTool, run it against the output and confirm no author, producer, or date fields remain. Then search the PDF for every sensitive string from your Step 1 list; the number of hits must be zero.
- Watch out: If the verification tool is the same tool that performed the sanitization, open the file in a second, independent reader to get a true check.
Step 9. Re-read the final PDF page by page
- Action: Page through the entire document, zoom to 200–300% over each redacted block, and confirm the redaction covers the full width of the text.
- Explanation: Production failures usually appear at the edges: redaction boxes slightly too short, stamps partially visible, or a bates-number label overlapping a redacted region.
Common Mistakes / Troubleshooting
| Problem | Likely Cause | Fix |
|---|---|---|
| Text under a black box can still be copied | A rectangle or highlighter was used instead of the redaction tool | Apply the real redaction tool, then search for the same string; if it is found, the redaction must be redone |
| Author name still appears in Document Properties | Metadata was never cleared from the file | Clear all fields in File > Properties and re-run the sanitize step |
| PDF fails to sanitize or will not open | DRM or permission locks on the file | Real-world tests show roughly 8.8% of PDFs fail sanitization, typically DRM-locked files; unlock the document or use a desktop tool with permission handling |
| Search still finds redacted grantor names | The OCR / hidden text layer was not included in sanitization | Re-run sanitization with the OCR and hidden-text options selected |
| “File too large” error | Free-tier file-size cap of 20 MB | Use a paid tier that supports up to 100 MB and batch mode, or switch to a desktop editor |
How to Verify Success
- Document Properties shows no author, producer, creator, or creation date
- Full-text search returns zero results for every name, address, and figure on your Step 1 list
- The Comments, Annotations, and Attachments panels are empty
- Each redacted block is solid black with no text visible at 300% zoom
- The output is a newly saved PDF with a distinct filename; the original deed remains untouched and intact
- The file was processed without any upload (0 bytes sent, if your tool reports this statistic)
FAQ
Q1. Can I just print the deed to PDF and call it sanitized?
No. Printing flattens some annotations, but metadata such as author and producing software can survive, and the OCR text layer usually remains searchable. Sanitization must be an explicit operation that physically deletes metadata, comments, and hidden content.
Q2. What file sizes can I process?
Free local browser tools typically allow files up to 20 MB and a limited number of conversions per day. Paid tiers commonly raise the limit to 100 MB and add batch mode. A deed with many high-resolution scanned pages may require the paid tier or a desktop application.
Q3. Why does local processing matter for a real estate deed?
A deed in e-discovery is confidential and potentially privileged. Uploading it to a third-party server means a copy of the document exists on infrastructure you cannot control or audit. Local processing ensures no copy ever leaves your device, which strengthens defensibility if the production is later challenged.
Key Facts
- Sanitization goes beyond visible redaction: it physically deletes metadata, comments, and hidden content rather than covering them.
- Local processing keeps the source PDF on your device, so third-party servers cannot receive the file.
- Real-world testing of a local browser sanitizer achieved a 91.2% success rate on 113 PDFs; the remaining files were DRM-locked.
- The median processing time for sanitization operations on that test set was 420 ms across 7 core operations.
- Free browser-based sanitization typically caps at 20 MB and 3 conversions per day; paid plans support 100 MB files and batch processing.