Part XV Overview: Build a Densitometer

Part XIII read its step wedge with a lux meter, a spot meter and a set of matched grey patches, and measured how badly each did it. Part XIV then built a machine that puts a known quantity of light on film. Between the two is a hole that everything downstream falls into: the exposure axis of a characteristic curve is now under control, and the density axis is still guesswork with an error bar around it.

This part closes it. What it produces is not a gadget but a stated measurement — a number with a geometry, a colour and an uncertainty attached — because seven later parts read that number and none of their conclusions survive without it.

The measurement has a geometry and a colour before it has a value

A negative does not possess a density the way it possesses a length. It scatters light as well as absorbing it, so how much of the scattered light your detector collects is part of the answer. A reading through a diffuser in contact with the film and a reading down a narrow cone disagree, on the same film, by an amount that grows with density and with grain — the Callier effect. ISO 5-2 exists because of that disagreement; ISO 5-3 fixes the other half, the spectral condition, because a neutral silver image reads nearly alike in any band and a stained or toned one does not.

The course quotes neither standard. It states its own convention, names the standards as what that convention is modelled on, and marks every figure as its own measurement under its own stated criterion — the ruling governing every number since Part XIII.

The design, in one paragraph

A green LED on the same constant-current drive Part XIV built, behind an opal diffuser in contact with the film; a 2 mm sampling aperture defining the patch actually read; a silicon photodiode and transimpedance amplifier looking down at that aperture through a light-tight shroud; a 16-bit converter with programmable gain, because the Pico’s own 12-bit converter runs out of counts at 3.0 D and can be shown to; and firmware in the same MicroPython idiom, logging every reading with its dark value beside it. A 45/0 reflection attachment shares the photodiode and amplifier, so the instrument reads prints too — the only thing in this course that can.

What you build, and what you measure

Two builds: the optical head, where geometry and stray light are settled, and the electronics and firmware, where the signal chain is. Then a calibration experiment in both modes, which is what turns the object into an instrument, a break/fix page on the ways it misbehaves, and a full film-and-developer characterisation using both machines — the first measurement in this course whose two axes are both instruments rather than approximations. What you record is not the parts list but four numbers: repeatability, the stray-light floor that caps the range, drift, and linearity — which the course cannot source for you and therefore asks you to measure.

What it will honestly claim, stated before you buy anything

Diffuse, visual-like transmission density, to about 3.0 D, on a scale that stays relative until it is tied to a calibrated wedge. It will not conform to ISO 5-2 or 5-3, and the pages say which requirements it misses. It reads a green band rather than a spectral product, so a pyro-stained negative reads low against a printing density by an amount only your own measurement can give. The reflection scale is relative unless you buy a certified white tile — and the planner could not price one.

Against that, the benchmark this part uses throughout: X-Rite’s 361T brochure specifies repeatability of ±0.01 D, linearity of ±0.02 D from 0 to 5.0 D, apertures of 1, 2 and 3 mm, and zero stability of ±0.02 D per eight hours. Those four things are a specification, and yours will have all four by the end of the calibration page — measured, not claimed.

What you need first

Part XIII complete, with its step-wedge negative archived, because reading it again is how you learn what the interim method was worth. Part XIV complete: this instrument is its mirror, and its electronics primer is assumed in full and not taught again. A 21-step transmission wedge — the Stouffer T2115 — and here the calibrated CC part stops being a luxury, because an uncalibrated wedge gives you a repeatable instrument and no absolute scale at all.

⚠USB voltage only, and one hazard held back

Everything here runs from a certified USB supply or power bank at 5 V, for the reason HSE gives: limit the supply voltage to the lowest that will do the job, and 230 V AC can kill. Nothing mains is built, modified or opened anywhere in this part.

The optical head is Level A — cutting, gluing and blackening with hand tools. So is the electronics page, because its lamp runs at a few milliamps rather than Part XIV’s hundreds; soldering is declared there as a Level B step under the mixed-level rule, for the iron and its flux fume, and it is a step the breadboard route lets you decline. The UV channel is described and not built, because 365 nm is a different hazard class needing an enclosure and an interlock, and it earns its own treatment rather than a paragraph.

Four routes, and the last is not a failure

Build it, which is what the pages assume. Buy a light-to-digital sensor module — a TSL2591 or a VEML7700 puts photodiode, amplifier and converter in one part, and the pages say what that simplifies and what it hides. Buy a second-hand commercial instrument, then calibrate it here, because a twenty-year-old densitometer with an unknown history is not a certificate. Or use a scanner, good enough to see a curve’s shape and not to state a Dmax, as Part IX already showed.

The electronics are pocket money. The wedge and the reference tile cannot be improvised, and the planner records that it could price neither.

The pages of this part

Part15 of 28Level3 — AdvancedPages7Estimated time16.8 hoursHighest safety levelLevel A

✱1 of 7 pages in this part need a darkroom, a UV source or mains-powered equipment, marked below. Each says what can be improvised and, where one exists, gives an alternative route.

0 / 7 lessons in this part completed

• LessonDensity Depends on How You Look: Geometry and Spectral Response60 min
• LessonPhotodiodes, ADCs and Measuring Light Over Three Decades75 min
• BuildBuild: The Densitometer Optical Head (Transmission and Reflection)A180 min
• BuildBuild: The Densitometer Electronics and FirmwareA180 min
• ExperimentExperiment: Calibrating the Densitometer in Both ModesA180 min
• Break/fixBreak-Fix: Diagnosing a Misbehaving DensitometerA90 min
• AssignmentAssignment: Characterising a Film and Developer with Your Own InstrumentsA✱ Darkroom240 min