How to enlarge a photo with the grid method

Use an 8 × 10 in reference and a 16 × 20 in canvas, each divided into 4 columns × 5 rows. Reference cells are 2 × 2 in (8 / 4 and 10 / 5). Canvas cells are 4 × 4 in (16 / 4 and 20 / 5). Both dimensions double: 16 / 8 = 20 / 10 = 2. This is a uniform 2× enlargement.

Reference cell width = reference width / columns. Reference cell height = reference height / rows. Canvas cell width = canvas width / columns. Canvas cell height = canvas height / rows. Compare scale factors in the same physical unit. Only equal width and height ratios give a uniform scale.

A coarse grid is usually better than a dense one. Begin with four to eight cells across the shorter dimension. Simple landscapes and still lifes may need only a 4 by 5 or 5 by 7 grid. A portrait with small changes around the eyes and mouth may benefit from eight to twelve cells, but adding dozens of cells too early creates visual noise and encourages you to copy fragments without seeing the head as a whole.

You can also subdivide only where needed. Keep the broad grid for the full composition, then draw a temporary diagonal or half-cell inside one square that contains a difficult hand, eye, or architectural junction. Local subdivision preserves the clarity of the main scaffold while giving extra reference for a genuinely complex area.

Use a sharp hard pencil such as 2H on drawing paper, watercolor pencil in a compatible color on watercolor paper, or a light chalk line on a toned or painted ground. Test the material on a scrap or hidden edge first. The grid should be visible while you work but easy to erase, cover, or integrate later. Heavy graphite can groove paper, dirty pale paint, and remain visible beneath transparent washes.

Measure from one fixed edge rather than stepping a ruler from cell to cell. Small errors accumulate when each new mark depends on the previous mark. Match Photo Grid labels: letter columns A–D left to right and number rows 1–5 top to bottom. A1 is the upper-left cell; D5 is the lower-right. Clear labels reduce the chance of transferring a correct shape into the wrong square.

Do not begin by polishing the contents of the first square. First mark the highest, lowest, leftmost, and rightmost points of the subject. Add major landmarks such as the eye line, shoulder angle, horizon, or corners of a dominant object. Check where the silhouette crosses grid lines and estimate each crossing as a fraction of the cell width or height.

Connect landmarks with simple straight or gently curved lines, then inspect the negative spaces around the subject. The gap between a head and shoulder, or between a vase handle and its body, often reveals proportional drift more clearly than the named object. Work across several neighboring squares so contours join naturally instead of becoming a patchwork of disconnected segments.

Step back until the grid becomes less noticeable and compare the large envelope, tilt, and spacing with the reference. Look for gradual drift: a contour may cross every individual cell reasonably well yet move farther from the correct position over several rows. Flip the reference and drawing digitally or view the physical drawing in a mirror to expose familiar errors.

Once the large structure works, add internal shapes and a simple value map. Erase or soften the grid before delicate rendering, but keep enough landmarks to prevent later drift. If paint will cover the transfer, fix loose graphite according to the material manufacturer’s guidance and avoid sealing a surface in a way that harms adhesion.