Yet another sketch.

04. (Limited) success by measuring light absorption

Updated on January 23, 2015.








On the Intro page, I mentioned that a book is open to a spread if light falls onto that spread (approach VII). This is a conceptually simple observation, and I am curious to try it out if I can get access to knowledge and equipment to produce printed-electronics photo transistors.

However, printed electronics is not common amateur technology at the time of writing. But the concept of light spawned another idea in a discussion with Johannes Nilsson and Tony Olsson. If a number of consecutive pages in a book each has a hole covered with a semi-transparent material, then the current spread among those pages could be estimated by measuring the amount of light passing through the layers of semi-transparent material to the bottom of the hole. The more pages, the less light.

I made a test rig to try the idea, with a photoresistor in the bottom of a cardboard box and a number of paper cards with 5 mm holes in them. For semi-transparent material, I compared Japanese paper in two different weights: 9 and 30 grams per square meter. Japanese paper is commonly used for repairs and reinforcements in classical bookbinding; when it is pasted to another paper, it is more or less transparent, but on its own it is rather semi-transparent.

The test showed that the photoresistor provided repeatable values as long as the ambient light conditions were stable, and that the 30 g Japanese paper had the right amount of light absorption to provide nice differentiation.

Next, I made a book with pages from regular 80 g office paper, sewn on tapes and case bound. Two photoresistors are built into the inside of the front board, with wires coming out at the bottom of the spine. There are punched holes in the pages of the first eight spreads of the book, aligned with the photoresistors and covered with Japanese paper according to the following diagram.

Similar to CB01, an Arduino Uno is connected to the photoresistors and programmed to read the resistance values and send them to the serial port. A Processing program on the connected laptop then converts the resistance values to an estimate of which spread the book is currently open to. The first eight spreads of the book are labeled A–H, and the Processing program presents the two current resistance values and its guess of current spread (A to H, or blank for a closed book, shown in the top left of the laptop screen).

A few constants in the Processing code need to be calibrated manually for the current light conditions, but when that is done it turns out that the CB04 book actually reports reliably which of the first eight spreads it is open to.

On spread D, one of the resistors gets the light directly from the room while the other catches light through four layers of Japanese paper. Since the resistance of a photoresistor is inversely proportional to the amount of light it receives, the r1 resistor assumes its minimum value while r0 is at the top of the active measuring range.

When the pages are turned to spread F, r1 is in the middle of its active range while r0 gets practically no light and assumes its maximum value.

In other words, a book made like CB04 can report which spread it is open to. The way it works in shown in the February 2015 video snapshot.

This is definitely progress, as the approach could scale to more spreads by using more than four layers of translucent paper for each photoresistor and adding more photoresistors. The resulting book would be rather ugly, though, with the outer page margins perforated with holes.

The manual calibration could probably by sidestepped by adding another photoresistor that is constantly exposed to the light, and using the value of that resistor as a floating baseline in the calculations.