Image to line art: design a sampling study

An image as a source of instructions

An image can become a source of instructions rather than a background to decorate. Sample brightness and let it control a mark's length, density, or angle. Choose one relationship first. The resulting drawing should reveal which image features your rule preserves and which it deliberately discards.

Map image samples to output positions

Use an image you made or have permission to adapt. Sample it on a coarse grid, taking care to map output positions back to image coordinates. Start with a simple brightness formula and draw a line whose length increases as the sampled value becomes darker.

The useful adjustable quantity in this study is sampling interval; give it a name and a clearly recorded starting value. Break the construction into a single working unit and a repeated or extended form. Verify the unit on its own first. Only multiply it after its boundaries, movement, and proportions match the proposed rule. A useful acceptance check is whether the main light and dark masses survive without requiring every detail. Judge that at the intended viewing size or working distance, where tiny construction mistakes and exaggerated effects can have very different consequences.

Check brightness ranges and coordinate mapping

If the drawing appears mirrored or stretched, inspect the coordinate mapping before changing the mark. If everything is dark, check the range of sampled values. Compare a handful of printed numeric samples with their visible locations in the original image.

Use the plain reference to decide whether sampling interval caused the problem or merely made it easier to notice. Temporarily show construction points and intermediate values. Reduce repeated elements to a handful and pause anything that changes over time.

Compare mark length with mark spacing

Try a still life with one light object and one dark object against a plain background. Produce one version controlled by mark length and another controlled by spacing. Compare where each version loses edges or invents texture.

Build a comparison around these settings: 8, 16, and 32 output pixels. Use the first trial as an anchor, the second as a moderate departure, and the third to reveal the limit of the idea. Return to the anchor before deciding. The final choice may sit between the tested values; use the comparison to identify that smaller next adjustment.

Respect the sample grid's alignment

When comparing sampling intervals, keep the grid origin fixed. Shifting the origin can cause narrow features to land between samples and disappear, even though the interval is unchanged. Use a simple black bar on white as a diagnostic image before the folded-paper photograph. If the bar appears at the wrong scale or position, repair the coordinate conversion there rather than trying to interpret the error through complex photographic detail.