The histogram shows how values are distributed in your image. It's like an X-ray that reveals problems your eye might miss.
How to Read It
Left side: Dark values (shadows)
Middle: Mid-tones
Right side: Light values (highlights)
A tall spike means lots of pixels at that value. A flat area means few or no pixels at that value.
The histogram doesn't tell you if an image is "good" or "bad." It shows you the distribution of values so you can make informed decisions about your painting.
Common Patterns
Bunched in the Middle
Most values are mid-tones with few darks or lights.
Result: "Muddy" look with low contrast. Push your darks darker and lights lighter to add impact.
Spread Across the Range
Values distributed from dark to light across the histogram.
Result: Full tonal range. Dynamic and impactful. Most professional work uses this distribution.
Shifted to Darks
Most values are on the left side (dark end).
Result: "Low key" lighting. Moody and dramatic. Common in night scenes and film noir.
Shifted to Lights
Most values are on the right side (light end).
Result: "High key" lighting. Bright and airy. Common in fashion and beach scenes.
Fixing Muddy Values
The most common problem is a histogram bunched in the middle. Here's how to fix it:
- Check the histogram. If it's bunched in the middle, you've confirmed the problem.
- Use the eyedropper. Sample your darkest dark and lightest light. They should be near 1-2 and 9-10 on the scale.
- Push the extremes. Go back to your painting and darken your shadows, lighten your highlights. Leave mid-tones alone.
Result: The histogram spreads out, your painting has more contrast, and forms appear more three-dimensional.
Tips
- Check before you start painting. Load your reference photo and check the histogram. This helps you understand the value structure before you commit to a painting.
- Check your work in progress. Take a photo of your painting, load it in Value Study, and compare the histogram to your reference. Are they similar? Should they be?
- Don't chase perfection. The histogram is a tool for understanding, not a target to hit. Different subjects need different distributions.
Last updated: