Question about the need for linearity in Piezography



I noticed when I use the Sekonic meter and their expensive test card for determining exposure, the curve generated off the sensor sometimes follows the S curve as in the film era in the better digital cameras, i.e. a nice toe and shoulder curve roll-off on both. The cheaper cameras may show no shoulder at all, but they may have a toe hence things like clouds and white getting blown out with those cameras.

I’m wondering if the need (fuss?) for linearity with step wedges and their linearization might be at odds with the camera’s sensor that is non-linear, perhaps by the sensor or the internal processing in camera? All the step tablets are good for out of camera work, but once the camera enters, the step tablet might work against the camera’s curve and then needs to be altered for that.

If this makes any sense…



One way of thinking about linearity in inkjet printing is to remove any “fingerprint” of the device on the final print. Basically the process is to understand how the device behaves, and then account for that character.

In theory an inkjet printer takes in data, and then squirts out ink accordingly. A profile (or curve) is made by a manufacturer (OEM or 3rd party) based on a statistically “standard” device to tell printer the ratios of ink to use for a given paper. The challenge is there are lots of variables to account for that can happen once that profile is our in the wild. Variance in manufacturing, wear-and-tear of components, variance in paper coating, and variances ink are examples of things that can cause non-linear behavior.

When you print out a step-tablet and measure it, you are characterizing the printer. When you take that data, along with the original profile (or curves) use to print the test, a “linearized” profile can be created.

Make sense?