Getting Your Local Obituary Records Into a Digital Preservation System
Paper obituary files degrade. Ink fades, paper yellows, and boxes get moved around when families downsize. I spent three years working with municipal records in the Sheboygan area and watched at least a dozen collections get damaged by poor storage conditions before I figured out a workflow that actually sticks. This is a practical walkthrough for taking physical or existing digital obituary records and getting them into a digital preservation setup that won't lose data when someone forgets about it for a few years. The process breaks down into four phases: gathering what exists, digitizing it cleanly, formatting it properly, and storing it where it will survive. The hardest part is never the scanning. It is the metadata and the file naming conventions that make future access possible. Start by pulling together every obituary record you can find for the area you are working with. That means physical copies from funeral homes, newspaper archives, family Bibles, cemetery records, and any digital folders someone already has sitting on a desktop somewhere. In my experience, the biggest time sink is not digitization itself. It is tracking down records that are already "digital" but trapped in JPEGs inside a folder named "Photos 2019" on a computer nobody looks at anymore.
Digitization Setup
You do not need a top-tier scanner for most obituary work. A flatbed scanner at 300 DPI is the baseline. Go to 400 DPI if the text is small or the original print is on newsprint that will blur at higher magnification. I found that using a document mat on the flatbed keeps pages flat and reduces the post-processing cleanup time significantly. Pages that curl from being in old folders take twenty minutes each to clean up in software if you skip the mat. With the mat, it takes about two minutes. If you are working with already-printed obituaries that are damaged or handwritten, a DSLR camera with a copy stand and consistent lighting works better. The key is keeping the camera perfectly perpendicular to the page and using a tripod. Handheld photos create perspective distortion that OCR software struggles with, and you end up spending more time on manual transcription than you would have on scanning. I once spent six hours correcting perspective-warped images that could have been avoided with a fifty-dollar copy stand from Amazon.
File Naming and Organization
This is where most people fail. A consistent naming convention matters more than scan quality because a poorly named good scan is useless to anyone who comes after you. Use a format like: LastName_FirstName_YYYYMMDD_Obit.pdf or .tiff. Include the date of death or publication date, not the date you scanned it. The scanning date is irrelevant to researchers and creates confusion when you have multiple scans of the same document over time. Organize folders by year of death, then by last name within each year folder. Do not create separate folders for each source like "Newspaper" or "Funeral Home." That fragments records. If a person has an obituary in the newspaper and one from the funeral home, they belong in the same folder. The source information goes in the metadata, not in the folder structure.
Get the Full Details
Metadata Standards
For preservation purposes, you want to capture enough structured data to make searching possible without drowning in fields nobody fills out. The minimum viable metadata set includes: full name, date of birth, date of death, place of death, names of surviving family members, funeral service details, cemetery location, source of the obituary, and scanner operator. That is it. More fields than that lead to incomplete records because people skip the optional ones and then the whole entry becomes unreliable. I used Dublin Core as the metadata framework for a project with the Sheboygan County historical society. It is simple enough that volunteers can fill it out accurately and detailed enough that a researcher can filter by multiple criteria. The downside is that some of the custom fields we wanted to add were not native to the platform we used, which meant exporting to CSV and merging metadata manually. That added about eight hours to a project that otherwise took two weeks.
Storage Strategy
A single hard drive is not a preservation strategy. It is a temporary holding method. For actual digital preservation, you need redundancy. The standard approach is the 3-2-1 rule: three copies of your data, on two different media types, with one copy offsite. That means at least one copy on a network-attached storage device, one on external hard drives that rotate, and one stored offsite either through a cloud service or a physical drive at a family member's house. File format matters for long-term preservation. TIFF or PDF/A are the standards. JPEG is fine for access copies but you should never store your master files as JPEG because the compression is lossy and every time you open and resave the file you lose quality. I learned that the hard way when I tried to work with a JPEG master from an earlier project and the text became illegible after the fifth resave. Starting over took two days. Check your files annually. Hard drives fail. I lost a collection of about forty thousand scanned documents when a NAS drive started developing bad sectors. I had a cloud backup but restoring from it took fourteen hours and a lot of checking because the sync had failed silently for three months. That incident taught me to run automated integrity checks and verify backups monthly, not just trust that something is working because it existed once.
Common Pitfalls
Here are the specific problems I encountered that beginners usually miss. First, not all obituaries have clear dates. Some old newspaper clippings have the publication date but no death date, and sometimes the dates conflict. Cross-reference with cemetery records or Social Security Death Index records when possible. Second, duplicate records appear constantly. The same person will have an obituary in the local newspaper, the Milwaukee Journal Sentinel, and sometimes a paid version from the funeral home with additional details. Decide early whether you want all versions or just the most complete one, and stick to that rule consistently. Third, OCR on handwritten or poorly printed obituaries is unreliable. Budget time for manual verification of OCR output rather than assuming the automated transcription is accurate. I check about thirty percent of my OCR work manually and catch errors in roughly forty percent of those samples. That error rate means you cannot trust raw OCR output for anything beyond a first draft. For scanning batch jobs, NAPS2 (Not Another PDF Scanner) is free and handles large batches without choking. For OCR, Tesseract is free and decent but requires some command-line comfort. ABBYY FineReader is the professional choice if you have budget. For metadata management, a simple SQLite database paired with a spreadsheet export covers most needs without expensive software. For long-term storage management, ArchiveBox or similar self-hosted tools can help track file integrity and create checksums automatically. The overall cost for a solo operator working on a small community collection runs about two hundred dollars for a decent flatbed scanner, plus whatever external storage you need. Time investment varies widely depending on volume and condition of source material, but a well-prepared collection of five thousand records typically takes between forty and eighty hours of actual work spread over several weeks. The metadata phase alone can consume half of that time if you are doing it thoroughly.
