A response curve describes how image input maps to image output.

In photography, the input is usually the amount of light hitting film or a digital sensor, and the output is the recorded signal or brightness value. In an ideal perfectly linear system, doubling the light would double the recorded signal, so the graph would be a straight line.

Real capture systems are not perfectly linear across the whole range. At very low exposure levels, response can flatten off, and at the high end it can also taper or “roll off” instead of increasing cleanly. With film this high-end taper is often called the shoulder.

This matters for dynamic range because the usable range depends on how well the medium can separate shadows, midtones, and highlights before noise, compression, or clipping reduce detail.

It also gets more complicated after capture: RAW sensor data and processed image files are not mapped the same way. JPEGs, for example, apply a non-linear tone curve, so the final image brightness does not directly match the sensor’s original linear response.

So, the response curve is basically the graph showing how faithfully and in what shape a film/sensor/image pipeline turns incoming light into recorded image values.