Ideal Scanner Settings & Tips for Bulk Uploads
This article will go through ideal scanner settings and other tips around the optimal settings for your bulk sheet scanner.
File Format (JPEG & PDF)
PaperScorer accepts scanned answer sheets in JPEG and PDF formats. When configuring your scanner, select one of these two output formats. JPEG is ideal for single-page answer sheets, while PDF is the better choice when scanning multiple answer sheets in a single batch through a document feeder. Each uploaded file must be 250 MB or smaller. If your scanned files exceed this limit, try reducing the DPI slightly or switching from color to grayscale output in your scanner settings.
Resolution (DPI - 300)
Set your scanner to 300 DPI or higher for the best recognition accuracy. DPI (dots per inch) controls how much detail is captured in the scanned image. At 300 DPI, PaperScorer can reliably detect filled bubbles and distinguish them from stray marks or incomplete erasures. Lower resolutions may result in misread answers or processing errors. While scanning above 300 DPI is acceptable, it will produce larger files and slower upload times without a significant improvement in accuracy.
Color Mode (grayscale)
Scanners should be set to grayscale mode, not full color or black-and-white. Grayscale provides the brightness information the recognition engine needs to accurately detect filled bubbles, read student IDs, and distinguish light pencil marks from unfilled bubbles — all of which depend on measuring how dark each bubble region is relative to its surroundings. Full color works but produces files approximately 3x larger with no recognition benefit, since the engine converts to grayscale immediately upon loading. Black-and-white (1-bit binary) mode should be avoided entirely, as it discards the brightness gradients the engine relies on, making it impossible to detect light fills or apply adaptive thresholding. For resolution, 300 DPI is recommended — higher DPI increases file size and processing time without meaningfully improving accuracy. Ensure the scanner's auto-contrast or auto-exposure features are disabled if they aggressively lighten or darken the image, as consistent scan brightness produces the most reliable results.
Auto-Deskew and Auto-Crop
Disable the scanner's built-in auto-deskew and auto-crop features. The recognition engine performs its own perspective correction using fiducial markers on the answer sheet, which is more accurate than the scanner's generic algorithms. Scanner-level deskew can introduce interpolation artifacts and slightly alter the geometry of marker squares, making them harder to detect. Auto-crop can cut into the margins where markers and QR codes are located, especially on slightly rotated sheets. If the scanner offers an "overscan" or "extend scan area" option, enabling it can help ensure no edge content is clipped.
Brightness, Contrast, and Auto-Exposure
Set brightness and contrast to their default/neutral values, and disable any auto-exposure, auto-contrast, or background removal features. The engine's adaptive thresholding is designed to work across a wide range of scan brightness levels and paper shades — it normalizes internally. Scanner-level adjustments that boost contrast can clip light pencil fills to white (making them invisible) or darken paper texture to look like filled bubbles. Background removal features are particularly problematic because they attempt to whiten the paper, which destroys the subtle brightness differences the engine uses to detect light fills.
Duplex and Blank Page Handling
If using a duplex (double-sided) scanner, either set it to simplex (single-sided) mode or enable blank page detection/removal so the backs of sheets are automatically discarded. Answer sheets are single-sided, and blank back pages double the number of images the system must process and store. If blank page detection is not available, the engine will handle blank pages gracefully (they'll fail QR reading and be flagged), but it wastes processing time and storage.
Multi-Feed Detection
Enable multi-feed (double-feed) detection if the scanner supports it. When two sheets pass through together, the resulting scan contains overlapping forms that cannot be recognized. Most document scanners offer ultrasonic multi-feed detection that stops the scan immediately when overlapping pages are detected, allowing the operator to re-feed. This is especially important for bulk runs of hundreds of sheets where a missed double-feed would require identifying and re-scanning the affected students manually.
Page Orientation
The scanner can be set to feed sheets in any orientation — the engine automatically detects and corrects portrait vs. landscape rotation and upside-down sheets using QR code position. However, feeding sheets consistently in the same orientation (typically portrait, QR code at the bottom) minimizes the orientation correction step and produces marginally faster processing.
Page Alignment and Skew
Ensure the entire answer sheet is visible in the scan and positioned as straight as possible. Skewed or rotated pages can interfere with PaperScorer's ability to locate the bubble grid and timing marks on the answer sheet. If using a flatbed scanner, align the sheet edges with the scanner guides. If using an automatic document feeder (ADF), make sure sheets are fed squarely and not at an angle. A small degree of skew is tolerable, but keeping sheets straight will produce the most reliable results.
Lighting and Shadows
Use good, even lighting when scanning and avoid shadows falling over the bubble areas of the answer sheet. This setting is especially important when using a mobile device camera or a portable scanner rather than a traditional flatbed. Uneven lighting or shadows can cause certain bubbles to appear filled when they are not, or make legitimately filled bubbles harder to detect. If your scanner has brightness or contrast settings, keep them at their defaults unless the output appears noticeably too dark or too light.
Sheet Condition
Avoid scanning answer sheets that have folds, creases, or stray marks near the bubble areas. Physical damage to the sheet — such as wrinkles from folding or dog-eared corners — can cast small shadows that PaperScorer may interpret as filled bubbles. Similarly, stray pencil marks, smudges, or doodles near the answer grid can lead to misreads. If a sheet is damaged, flatten it as much as possible before scanning. Instruct students to keep the bubble area clean and to erase fully when changing an answer.
Multi-Page PDF Processing
When you upload a PDF file, PaperScorer processes each page individually as a separate answer sheet. This makes PDF the ideal format for high-volume scanning with a document feeder — you can load a stack of answer sheets, scan them into a single PDF, and upload that one file. Each page will be matched to a student and scored independently. If a particular page fails to process, the remaining pages in the PDF will continue to be scored normally.
Batch File Selection
You can select multiple files at once when uploading to PaperScorer. In the file selection dialog, hold Ctrl (Windows) or Command (Mac) to select multiple individual files, or hold Shift to select a range. This allows you to upload an entire folder of scanned answer sheets in a single operation. Each file is uploaded and processed sequentially, and a summary table showing the status of every file — including any errors — is displayed when the batch is complete.
If you have other questions about scanner settings, please reach out to our support team with questions about settings. We want to keep this page updated with all the latest possible settings.