A reference board is more useful when it answers questions than when it simply collects attractive images. For an illustration of a night market, you might need to understand stall construction, the distribution of artificial light, and how customers carry purchases. Organize the board around those uncertainties so research leads toward decisions you can actually use in the drawing.
Give each group of references a job
Read the brief and list the information you cannot confidently invent. Separate physical questions from visual-direction questions. How does a folding stall support its roof is different from how quiet should the palette feel. Use clear board sections for these needs, and write a short question above each section. The labels make gaps visible before rendering begins.
Select references that show the relevant evidence clearly. A beautifully graded street photograph may hide the stall's joints, while an ordinary daylight image reveals its construction. Include both if they answer different questions. Do not let presentation quality determine usefulness. A reference earns space because it resolves a specific uncertainty, not because it matches the mood of every other image.
Record source links and any usage or license information needed for the project. Reference use and direct asset reuse are different activities, and permission requirements depend on what you actually do with the material. Keep enough context to identify the source later. An untraceable screenshot may become a problem when someone asks where a distinctive design or incorporated resource originated.
Turn the board into decisions and close it deliberately
Annotate a few observations in your own words. For example, roof supports meet at these joints, bright lamps create small local pools, or customers leave a clear walking channel. These notes translate reference into structural knowledge. Avoid simply writing use this beside an entire image, which encourages imitation without identifying what the reference contributes to the new illustration.
Make a small sketch that applies the answers in a different arrangement. Construct your own stall and lighting setup rather than tracing one source composition. Check whether the board now provides enough information to resolve the main forms. If a key mechanical relationship remains unclear, search specifically for that gap instead of adding more atmospheric photographs to a section already well supported.
Limit further collection once the important questions have usable answers. Save unresolved items as specific follow-up tasks, such as find a clear underside view of the canopy hinge. A board can remain compact and still be thorough for its purpose. Endless accumulation makes the most relevant evidence harder to locate and can postpone drawing without improving the decisions it depends on.
Keep the final board with the project and update it only when a new practical uncertainty appears. During review, point to the relevant evidence when explaining a choice without implying the reference proves every artistic decision. The board should function as a working map from questions to observations, supporting an original composition while reducing guesswork about the particular objects, setting, and visual behavior the project needs.