Reference Quality v. Digitization Quality

You might have seen it mentioned throughout the website and other Sourcery channels that we offer “reference quality” scans and not “digitization quality” scans. But what does that mean, in practice? 

Digitizing materials is a long process that can get costly, time consuming, and require large amounts of staff effort. Digitization projects generally require specialized or high-quality scanning equipment, particularly projects that are digitizing bound materials like books and manuscripts or particularly fragile material, like old letters or pieces of ephemera. The result is images that are high resolution, and have detailed metadata.  

Reference quality scans, on the other hand, are simply meant to be used as the name suggests – to review and evaluate the scanned content. Reference scans should be clear and legible, however, they don’t need to be of the highest quality for digitizing or publication. You should be able to use a reference quality scan to create a good transcription by hand or via OCR or AI such as Transkribus or Gemini, but not to submit for publication in a book or a journal.  

More specifically, reference quality scans sent and received through Sourcery should be: 

1. Legible. All areas of the scanned material should be readable with minimal manipulating by the Researcher.  

2. Clear. The scans should be minimally blurry and generally noiseless.  

3. Straight. While sometimes this cannot be helped, Sourcerers will try to keep the scans as straight as possible.  

4. Well lit. Sourcerers will try to take scans with minimal shadows. If this is not possible, they take multiple scans of pages at different angles, so the entire document is visible. 

5. Complete. Sourcerers will ensure all pages are included and that no edges or sides are cut off. 

👀 You can view examples of ideal, acceptable, and unacceptable scans via the Sourcerer and Researcher handbooks.