About

Memory cards are a spaced-repetition tool grounded in the testing effect: retrieving information from memory strengthens long-term retention more effectively than passive review. A poorly designed card set - inconsistent sizing, illegible fonts, ambiguous pairings - degrades recall accuracy. This generator produces matched card pairs where each front face maps to exactly one back face, enforcing a bijective relationship across n pairs. Cards render at 300 DPI for print clarity. The tool handles text wrapping, image scaling with aspect ratio lock, and grid pagination for A4 output.

Note: this tool approximates print bleed at 0 mm. For commercial print runs, add 3 mm bleed manually. Maximum recommended set size is 50 pairs to keep export file size under 10 MB. Image-heavy cards consume more storage. The localStorage budget caps at roughly 4 MB of base64 image data before the browser may reject writes.

Formulas

Cards per A4 page is calculated from the printable area divided by card area plus gutter spacing:

N = floor(W_pagew_card + g) × floor(H_pageh_card + g)

Where W_page = 210 mm (A4 width), H_page = 297 mm (A4 height), w_card and h_card are card dimensions, and g is the gutter gap (default 5 mm).

Total pages required for n card pairs (front sheet + back sheet):

P = 2 × ceil(nN)

The factor of 2 accounts for separate front-face and back-face sheets. Image scaling preserves aspect ratio: s = min(w_targetw_img, h_targeth_img).

Reference Data

Card Size	Dimensions	Cards per A4	Best For
Mini	50 × 50 mm	20	Vocabulary drill
Small	60 × 80 mm	12	Language flashcards
Standard	70 × 95 mm	8	General study
Large	90 × 120 mm	6	Children's matching games
Poker Size	63.5 × 88.9 mm	8	Standard playing card format
Tarot Size	70 × 120 mm	6	Illustrated cards
Index Card	76.2 × 127 mm	6	Detailed notes
Square S	60 × 60 mm	15	Symbol matching
Square M	80 × 80 mm	8	Picture memory
Square L	100 × 100 mm	6	Toddler cards
Half A6	52.5 × 74 mm	12	Compact quiz decks
A7	74 × 105 mm	8	Exam revision

Frequently Asked Questions

Each card renders at 300 DPI. A standard 70 × 95 mm card produces a canvas of 827 × 1122 px. Smaller card sizes yield fewer pixels but remain sharp for text. Image-heavy cards on mini sizes (50 × 50 mm) may lose fine detail below 150 px source width.

Consumer printers vary in duplex alignment. Separating front and back sheets lets you manually align them. Print the front sheet, flip the paper, feed it back, then print the back sheet. The card order on the back sheet is horizontally mirrored per row so fronts and backs align when the paper is flipped along the long edge.

The renderer uses a word-wrap algorithm that measures each word via Canvas measureText(). If total text height exceeds 85% of card height, the font size auto-reduces in 1 px decrements until it fits. A minimum floor of 8 px prevents illegible output. If text still overflows at the minimum, it truncates with an ellipsis.

The tool rejects files above 5 MB to prevent localStorage quota exhaustion. Base64 encoding inflates file size by roughly 33%. A 5 MB source becomes approximately 6.7 MB in storage. With multiple cards, you can hit the browser's 5 - 10 MB localStorage limit quickly. Compress or resize images before uploading.

The algorithm iterates from index n − 1 down to 1, swapping each element with a randomly chosen element at index 0 to i (inclusive). This guarantees each permutation has equal probability 1n!. The tool uses Math.random() which is sufficient for non-cryptographic shuffling.

Yes. The back-sheet renderer reverses card order within each row. This accounts for long-edge flip on A4 portrait. If you flip along the short edge instead, the alignment will be incorrect. The export labels specify "Flip on long edge" as instruction text in the margin area.