Density Depends on How You Look: Geometry and Spectral Response

Hand one negative to two laboratories and ask each for the density of step 14. One answers 1.42 and the other 1.61. Neither has made a mistake, neither instrument is faulty, and the difference is not noise. They collected the transmitted light differently, and the number a densitometer reports is a property of the film and the instrument together.

That is not a defect in the measurement. It is what density is. A length is a property of a rod; a density is the answer to a question, and the question has to be stated. This page states it — in geometry first, then in colour — and ends by saying which questions the head you are about to build can and cannot ask.

Silver scatters, and scattering is where the trouble starts

A developed silver image is not a dye. Development leaves tangled filaments of metallic silver a few hundred nanometres across, sitting in gelatin, and light meeting one of those filaments is not simply absorbed. Some of it is absorbed, and some is scattered — thrown sideways out of the direction it was travelling in, mostly forwards but through a wide cone. The finer the silver, the more of the beam is scattered rather than absorbed, which is why the effect is stronger in a fine-grain emulsion at a given density than in a coarse one.

Now put a detector above the film and ask what it sees. If the detector is large and close, it catches the scattered light along with the light that came straight through, and reports a low density. If the detector is small and far away, or behind a narrow stop, the scattered light misses it entirely and is counted as absorbed, so the same patch reports a high density. The film has not changed. Only the collection has.

The same grain, two collections, two densities

1. 1Source and opal diffuser — identical in both panels; the illumination is not what differs
2. 2One scattering site in the emulsion — silver filaments a few hundred nanometres across; some light absorbed, some thrown sideways
3. 3Wide detector, close — collects the scattered light too, so it is counted as transmitted — diffuse density, the smaller number
4. 4Aperture stop — the component that decides the answer; it throws away everything outside a narrow cone
5. 5Small detector behind the stop — scattered light never arrives, so it is counted as absorbed — projection density, the larger number

Nothing about the film differs between the panels. The ratio of the two readings is the Callier coefficient Q, and it is greater than one for any real silver image.

The Callier coefficient, and why this page prints no value for it

The ratio of the two readings has a name and a symbol.

Q = D_specular ÷ D_diffuse

The Callier coefficient

D_specular is the specular density, read down a narrow cone; D_diffuse is the diffuse density, read with the scattered light collected. Because scattered light can only ever be added to the diffuse reading, Q is never less than one, and it rises with the amount of scattering — which means it depends on the density itself, on the grain, and on the developer.

André Callier is the name attached to it because of two papers of 1909, catalogued in the bibliography appended to the Hurter and Driffield memorial volume: La diffusion de la lumière par les clichés, in the bulletin of the Belgian photographic society, and The Absorption and Scatter of Light by Photographic Negatives Measured by means of the Marten’s Polarization Photometer, in the Photographic Journal. Those papers are not in this course’s research corpus.

!This course quotes no value of Q, and that is deliberate

Tables of “typical” Callier coefficients circulate widely — a number for fine grain, a number for coarse, a number that rises with density. This course has verified none of them against a Tier 1 or Tier 2 source and therefore prints none. What it can say is what the physics forces: Q is at least one, it increases with scattering, and it is a property of your film in your developer rather than a constant.

That is not a hole in the teaching. It is an experiment. Once your densitometer exists you can measure your own Q directly, on your own negatives, by reading the same step twice — once with the detector close and wide, once behind a stop that admits only a narrow cone — and the further experiments at the foot of this page say how. A measured coefficient for your material is worth more than a quoted one for somebody else’s.

There is one place the effect vanishes altogether, and Mike Ware states it plainly in the Platinomicon when reading a step tablet for a printing exposure scale: there is no Callier effect in contact printing. The reason is geometry again. In contact the negative sits against the paper with no lens and no stop between them, so scattered light travels a fraction of a millimetre and lands essentially where the unscattered light landed. Nothing has been thrown away, so nothing is counted as absorbed. A contact printer is, optically, a diffuse-collection instrument.

The two geometries a standard defines, and why opal replaced the sphere

ISO 5-2 is the standard that fixes this, and this course cites it by number without reproducing it, as the course’s measurement convention requires. What its freely published introduction records is enough to teach from.

Standard diffuse transmittance density is the measurement in which one side of the film is diffuse — the illuminator radiates into the whole hemisphere, or the receiver collects from it — and the other side is directional. The introduction says what it is for: it is the quantity relevant to contact printing and to rating films on a viewing box, because those are the conditions under which the film is actually used.

Standard projection transmittance density is the measurement in which both sides are directional, and the standard defines two of them, identified not by an angle but by an f-number, because an f-number is what is engraved on a projection lens. The introduction names f/4.5 as representative of microfilm readers and f/1.6 as representative of motion-picture projectors.

The historically interesting change is in the same introduction. The first edition of 1974 specified diffuse density by an integrating sphere; the second edition of 1985 replaced it with a diffuser, typically opal glass, and the introduction gives the reason, which is not convenience. When film lies on a light box or sits in contact with a sheet of paper, light is inter-reflected between the film and the surface behind it, and that inter-reflection changes the density. An opal diffuser in contact with the specimen reproduces that; an integrating sphere does not. The introduction notes the consequence honestly: slightly smaller density values are generally obtained by the opal method.

⚗Why a densitometer that mimics a contact printer is the right instrument for this course

Almost everything this course measures is destined for contact — a step wedge printed by contact in Part XIII, contact-printed alternative processes in the parts that follow, control strips read against each other. The standard’s own reasoning says that the diffuse, opal-in-contact geometry is the one that predicts what happens in contact. So the geometry you are about to build is not a compromise forced by cheapness. It is the geometry the work actually calls for, and the projection geometries are the ones this course would have to justify.

The sampling aperture

Both kinds of measurement need a third thing: the area being read. ISO 5-2’s introduction calls it the sampling aperture — a small opening in an otherwise opaque sheet, defining the patch whose density is reported.

The size matters in two directions at once, and the tension is the whole of the design decision. Too large, and the aperture straddles a boundary and averages two densities into a number that describes neither. Too small, and the grain of the film itself starts to show up as scatter in repeated readings, because you are sampling a few hundred clumps of silver instead of a few hundred thousand.

Put a number on it with the wedge you own. A Stouffer T2115 is half an inch by five inches — 12.7 mm by 127 mm — divided into 21 steps, so each step is 127 ÷ 21 ≈ 6.0 mm long and 12.7 mm wide. A 2 mm aperture centred on a step leaves 2 mm of margin on the short axis and more than 5 mm on the long one, which is generous even with a hand-registered wedge. A 3 mm aperture still fits and gathers more than twice the light. A 6 mm aperture does not fit at all. That arithmetic, and nothing more sophisticated, is why the head you build uses 2 mm and offers 3 mm as an alternative — the same two sizes the X-Rite 361T offers, alongside 1 mm and an optional 0.5 mm for finer work.

Which geometry matters where

What you are doing	The geometry that predicts it	Why
Contact printing a negative	Diffuse	Scattered light lands where the direct light lands; nothing is thrown away
Printing under a diffusion enlarger or cold-light head	Diffuse, or close to it	The illumination is already scattered from all directions before it reaches the negative
Printing under a condenser enlarger	Projection	Condensers deliver a directional beam and the lens accepts a limited cone, so scattered light is lost
Projecting a slide or a motion picture	Projection, at the lens’s own f-number	The instrument is a projector
Scanning	Neither exactly	A scanner’s optics are its own; a scanner’s numbers are its own until calibrated
Comparing with a manufacturer’s published curve	Diffuse visual	Because that is what the manufacturer used, and says so

That last row is not a convention this course has adopted. It is printed on the figures themselves. Kodak’s T-Max 100 and Tri-X datasheets both label the axis block of every characteristic curve “Densitometry: Diffuse visual”, next to the exposure and process conditions. If you want your curve to be comparable with theirs, that is the geometry and that is the colour, and any other choice needs a reason and a note.

⁇The condenser-contrast claim, and what the course will and will not assert

The standard explanation for why a negative prints with more contrast under a condenser enlarger than under a diffusion head is the Callier effect: the condenser reads the negative in projection geometry, so the denser highlights — which scatter more — are effectively denser still, and the tonal scale stretches. The claim is old, widely repeated, and physically plausible for exactly the reason this page has just set out.

This course has not sourced it to a manufacturer publication or a standard reference work, and so it states it as the received explanation rather than as an established figure. Nobody in the corpus this course reads gives the size of the effect for a named film and a named enlarger. What is not in dispute is the direction: if the effect exists at all, a condenser prints harder, never softer.

And it is measurable by you. Print the same negative to the same shadow density under both heads, read the two prints, and the difference in their highlight densities is your own answer for your own equipment — and the break/fix and characterisation pages that close this part give you the instrument to obtain it with.

Colour is the other half of the definition

A geometry alone does not define a density, and ISO 5-3 is the standard that supplies the rest. Its freely published introduction sets out the idea in a form that is worth understanding whether or not you ever read the normative text.

The spectral response of an instrument is not the response of its detector. It is the response of the detector multiplied, wavelength by wavelength, by the transmission of every filter, every lens and every window in the path, and by the spectral power distribution of the source. Multiply all of those together, tabulate the result, and you have what the standard calls the spectral products — a table of numbers that says what the instrument weights each wavelength by. Two instruments with different lamps, different filters and different photodiodes give the same reading if, and only if, their spectral products agree.

Given a spectral product, the same density can be computed from a measured spectral transmittance rather than read off a filter instrument, and the 2009 revision added spectral weighting factors — the products interpolated to 1 nm intervals and normalised — for exactly that purpose, because simple filter instruments are in decline. The introduction is candid that summing coarse 10 nm products against measured spectra was never strictly accurate, with the errors probably in the third decimal place.

Two more things from that introduction bear directly on the build. The traditional illumination for both transmittance and reflection density is based on Planckian radiation at approximately 2856 K — CIE standard illuminant A, adopted in 1931, approximated by a tungsten filament. And the introduction records that process-control instruments using an LED source are now being made, with a warning attached: care is advised when comparing measurements made under different illumination conditions, particularly across colorants with different spectral characteristics.

That warning is, almost word for word, the limitation of the instrument you are about to build.

Named conditions: visual, printing, and the status densities

Out of that machinery come the named spectral conditions. Three matter here.

Visual density weights wavelengths roughly as the eye does, peaking in the green. It is the condition under which black-and-white film curves are published, and it is what “diffuse visual” on a Kodak datasheet means.

Printing density weights wavelengths as the print material sees them. Ordinary silver-gelatin enlarging paper is blue-sensitive, so a printing density is essentially a blue reading. For a neutral image the two conditions barely differ. For a stained one they differ a great deal, which is the next section.

Status densities — A, M, T, E and others — are the standardised conditions for colour work and graphic arts. They are named in ISO 5-3 and none of them is what a black-and-white worker needs, but the names appear on every commercial instrument and it is useful to know that they are spectral products with labels rather than mysteries.

What the detector sees, what the eye sees, and where three sources sit

• UVA source band (320–400 nm) — alternative-process printing; maximum output around 365 nm
• blue LED (440–470 nm) — approximates a printing density on blue-sensitive paper
• green LED (515–545 nm) — the course's transmission channel
• photopic peak region (545–570 nm) — where a visual density is weighted most heavily
• infrared, invisible and unwanted (700–1120 nm) — the detector is most sensitive here and the eye is blind
• Silicon photodiode response (BPW34 family) — peak 920 nm; 10 per cent points at 420 and 1120 nm

The photodiode curve is constructed from the datasheet's stated maximum at 920 nm and its stated ten per cent limits at 420 and 1120 nm; it is not digitised from the manufacturer's plot, and the band positions are nominal LED and source ranges rather than measured emission spectra. The lesson of the drawing does not depend on the exact shape: a bare silicon detector is an infrared detector that happens to see a little visible light, and a visual density read with one is a claim about a filter you have not fitted. The coloured strip approximates where the visible spectrum falls and is a reading aid only; the wavelengths in the labels carry the information. The bands and curves are drawn to show the relationship, not measured.

A neutral image reads alike in any band; a stained one does not

Metallic silver, finely divided, absorbs across the visible more or less evenly. That is why a black-and-white negative looks grey rather than coloured, and it is why the whole apparatus of spectral conditions matters less for ordinary silver work than a reader of ISO 5-3 might fear. Read a neutral silver step in green, in blue or in red and the three numbers are close. Close is not identical, and the residual difference is one of the things your instrument can measure once it exists.

Then there are the staining developers, and they break the equivalence completely. Bergger’s own datasheet for PMK describes what pyrogallol does: a yellow-green tint surrounds each silver grain and fills the space between grains, becoming an intrinsic part of the image, so that the density of a pyro negative is “the conjunction of two densities, that of the silver and that of the coloration” — and the maker states that the film’s printing qualities are increased as a result.

Follow that through. A yellow-green absorber is yellow-green because it absorbs the blue end of the spectrum more strongly than the green. So the stain contributes a large density to a blue reading and a small one to a green reading. A pyro negative therefore has a printing density substantially higher than its visual density, and the gap grows with the amount of stain — which is to say, with the highlights, which is precisely the mechanism behind the compensating behaviour those developers are prized for.

!What is sourced here and what is inference

The yellow-green tint, the statement that the stain is a second density, and the claim that printing qualities are improved all come from the manufacturer’s datasheet. The step from “yellow-green” to “the blue density exceeds the visual density” is this course’s own inference from elementary colour physics, not a measured figure from any source in the corpus. Nobody in that corpus gives the size of the gap for a named film and developer.

So: measure it. A pyro-developed step wedge read once with a green LED and once with a blue one gives you the two curves and the difference between them, and the characterisation assignment that closes this part is where that becomes a plotted result rather than a paragraph. Anyone printing with a staining developer who reports a contrast index without saying which channel they read it in has reported a number that cannot be checked.

Ultraviolet density, and the negatives that live or die by it

The alternative processes ahead of you in this course are printed by ultraviolet. Ware’s Platinomicon sets out the source requirement: what serves is long-wave ultraviolet, UVA, 320 to 400 nm, with a maximum output around 365 nm, which he gives as the calculated optimum for the siderotype processes, and he warns against short-wave mercury lamps, which damage eyes and living tissue rapidly and offer no advantage.

A negative destined for platinum, palladium or cyanotype is therefore judged on its UV density, and a visual reading can mislead badly in either direction. Film base itself is often not transparent in the near ultraviolet even when it looks water-clear; the X-Rite 361T brochure makes exactly this point as a selling feature, saying its built-in UV response measures film-base fog that an ordinary densitometer cannot see, because UV blocking can occur in a seemingly clear base. A stain that reads modestly in green may be nearly opaque at 365 nm. And a digitally printed negative on inkjet film — the route many alternative printers take — is built from dyes or pigments chosen for one purpose, with UV transmission that no visual reading predicts at all.

What this course can honestly say is therefore limited, and worth stating precisely:

• Established. UVA from 320 to 400 nm with a maximum near 365 nm is the printing radiation for the iron-based processes, from a Tier 2 conservation source.
• Established. A commercial instrument sells a separate UV response on the grounds that visual and UV readings of the same film base differ.
• Not established here. The UV transmission of any specific film base, stain or inkjet negative material. No number in this course’s corpus supports one, and none is printed.

The optical head has a UV variant for this reason, and the build page describes it and does not build it — a 365 nm source is a hazard class of its own, needing an enclosure and an interlock, and it belongs with the alternative-process cluster rather than smuggled into a Level A afternoon.

Reflection density, briefly, because prints are the other half of the course

Everything so far has been about light passing through film. A print reflects, and the same two questions — what geometry, what colour — return with different answers.

ISO 5-2’s foreword lists the four parts of ISO 5, and the fourth is Geometric conditions for reflection density. This course has not obtained ISO 5-4 or its preview, so it reproduces nothing from it, states no tolerance and claims no conformance. What the build page does instead is derive the arrangement from first principles and declare its scale relative.

The principle is one sentence. A glossy print returns two quite different things: a specular reflection from the surface, which carries no image information at all and is simply a picture of the lamp, and a diffuse reflection from the silver and the paper beneath, which is the image. Illuminate at 45° to the print’s normal and collect along the normal, and the specular component leaves at 45° on the other side and misses the detector entirely — the arrangement conventionally written 45/0. Get the angle wrong and a glossy and a matt paper at the same silver density read differently, which makes the measurement a measurement of surface finish.

The second half is the scale. Transmission density has a natural zero: air. Reflection density has none, because there is no such thing as perfectly white paper, and every reflection reading is therefore relative to a reference. Paper white and Dmax are the two ends, and both belong to the paper rather than to the image — which is why a print’s tonal range is a property of the material before it is a property of the printing.

What a real instrument specifies, item by item

The X-Rite 361T is a desktop transmission densitometer of the kind that sat in every process laboratory, and it is worth reading its specification not as a shopping list but as a template for what you will have to say about your own.

Specification	The 361T	What the number means
Measuring range	0 to more than 6.0 D	The span over which any figure below applies at all
Measuring area	1, 2, 3 mm; 0.5 mm optional	The sampling aperture, chosen for the work
Repeatability	±0.01 D, 0.0 to 5.0 D (Ortho and Visual, 2 and 3 mm)	Read the same patch twice and this is the spread
Linearity	±0.02 D over the same range	How far the scale bends away from a straight line
Zero stability	±0.02 D per 8 hours	How far the zero wanders once warmed up
Slope stability	±1 % per year	How fast the calibration ages
Interinstrument agreement	±0.02 D	Whether two of them agree with each other
Warm-up	2 minutes; 5 for UV	How long before any of the above is true
Ambient interference	D decreases by less than 0.25 %	How much room light gets into the reading
Geometry	Per ANSI PH2.19 and ISO 5/2; illumination at 0°, collection by a diffusing surface	The question the number answers
Response	X-Rite Ortho, X-Rite UV, optional Visual	The spectral condition
Source	About 2850 K	The illuminant behind the spectral product
Operating temperature	10 to 40 °C	Outside this, none of it is claimed

Three of those rows repay a second look. Repeatability and linearity are different claims, and the instrument’s linearity figure is twice its repeatability figure — being consistent is easier than being right, and any instrument you build will show the same ordering. The UV response is specified over a shorter range than the visual one, to 3.5 D rather than 5.0, because there is less light and more trouble at the short end. And ambient interference has a number, which tells you that stray room light was a design problem for a professional instrument in a metal case, and will be a much larger one for a box you have made.

What this means for the head you are about to build

Every decision on the optical head page follows from something above.

An opal diffuser in contact with the film, because the standard’s own reasoning says that inter-reflection between diffuser and specimen is what a contact print experiences, and contact is what this course does.

A 2 mm aperture, from the wedge arithmetic — 6.0 mm by 12.7 mm per step — with 3 mm offered where more light matters more than placement.

A green LED near the photopic peak, because the eye’s sensitivity is greatest there, a neutral silver image reads nearly the same in any visible band anyway, and green sits far from both the blue end where a stain would separate the reading from a visual one and the near infrared where the detector is most sensitive and the eye is blind.

No claim of conformance to anything. The head has a diffuser below and a large photodiode a millimetre above, which makes it diffuse on both sides — neither of the two modes the standard defines. It has a green LED and a bare silicon photodiode, which is a spectral product but not a specified one. It reads a relative scale until a calibrated wedge ties it down. Every one of those is written on the instrument’s certificate at the end of the calibration experiment, and an instrument that states its limits is worth more than one that claims a standard it has never been tested against.

Recap

A density is not a property of a negative alone. Silver scatters as well as absorbing, so a detector that collects the scattered light reports a smaller number than one that throws it away; the ratio is the Callier coefficient Q, always at least one, and this course prints no value for it because it has sourced none. ISO 5-2 fixes the geometry, distinguishing diffuse density — the quantity relevant to contact printing and light-box viewing — from projection density at f/4.5 and f/1.6, and the sampling aperture defines the patch. Opal replaced the integrating sphere in 1985 because inter-reflection between diffuser and specimen is what really happens under a contact print. ISO 5-3 fixes the colour through spectral products, and the named conditions follow from it; a neutral silver image reads much the same in any of them and a pyro-stained one does not, because a yellow-green stain is a large density in blue and a small one in green. Alternative-process negatives are judged in UVA around 365 nm, where this course can source the printing radiation and nothing about any particular base’s transmission. Reflection density needs the 45/0 arrangement to throw the specular component away, and a reference white to have a scale at all. And a commercial specification is nine or ten separate claims, of which repeatability and linearity are the two most often confused.

Check your understanding

Question 1. You read step 12 of a wedge twice on the same negative: 1.85 through a small aperture stop and 1.52 with the detector large and close. What have you measured, and what would the same pair of readings look like on a piece of clear film base?

• An instrument fault; the two numbers should agree and one channel needs recalibrating
• A Callier coefficient of about 1.22 for that step; on clear base the two readings would very nearly agree
• A Callier coefficient of about 0.82 for that step; on clear base the specular reading would be the lower
• The difference between a visual and a printing density

Show the answer and why

Answer: A Callier coefficient of about 1.22 for that step; on clear base the two readings would very nearly agree

Q is the specular density divided by the diffuse one, 1.85 ÷ 1.52 = 1.22. The direction is forced by physics: scattered light can only add to what the wide detector collects, so the specular number is the larger and Q cannot be below one, which rules out the third option. The clear-base check is the useful part of the question. Clear base scatters almost nothing, so there is nothing for the stop to throw away and the two geometries converge — which is also why Q is not a constant for an instrument but a function of how much silver is in the patch. Option four names a spectral difference, and nothing about the colour changed between these two readings.

Question 2. A Kodak datasheet labels its characteristic curve "Densitometry: Diffuse visual". You read your own strip of the same film with a green LED and a photodiode behind a 2 mm aperture, with an opal diffuser in contact underneath. Which of your figures may fairly be compared with theirs?

• None of them, because your instrument is not calibrated to a standard
• The shape of the curve and the contrast index, with the caution that your spectral product is not theirs; absolute densities only after a calibrated wedge ties your scale down
• All of them, because your geometry is diffuse and theirs is diffuse
• Only the base plus fog, because it is the one reading taken on unexposed film

Show the answer and why

Answer: The shape of the curve and the contrast index, with the caution that your spectral product is not theirs; absolute densities only after a calibrated wedge ties your scale down

Your geometry is close to theirs in kind: opal in contact, collecting the scattered light. That is what makes the shape comparable, and a contrast index is a ratio of differences along that shape, so the parts of your scale that are merely offset cancel out of it. What is not comparable without further work is an absolute density, because your scale has no anchor until a calibrated wedge gives it one, and because a green LED and a bare silicon photodiode do not make the same spectral product as their instrument. Option one is too austere — a relative measurement whose limits are stated is a real measurement — and option three ignores the colour half of the definition entirely.

Question 3. You develop a film in a pyro staining developer and read the same step at 525 nm and at 450 nm. What do you expect, and what does the manufacturer's datasheet actually establish?

• The 450 nm reading is higher; the datasheet establishes the size of the difference
• The 450 nm reading is higher; the datasheet establishes that the stain is a second density and that printing qualities improve, but the size of the difference is yours to measure
• The 525 nm reading is higher, because green light is scattered more by the stain
• The two agree, because the stain is proportional to the silver and cancels in the ratio

Show the answer and why

Answer: The 450 nm reading is higher; the datasheet establishes that the stain is a second density and that printing qualities improve, but the size of the difference is yours to measure

The direction follows from what "yellow-green" means: a yellow-green material transmits green and absorbs blue, so its density at 450 nm exceeds its density at 525 nm. Bergger describe the tint, state that the density of a pyro negative is the conjunction of the silver density and the coloration, and say the printing qualities are increased — but they give no numbers, and no source in this course's corpus does. That gap is exactly why the course tells you to measure it rather than quoting somebody. Option four is worth refuting on its own: the stain is proportional to development, which is proportional to exposure, so it is largest in the highlights and the difference between the two curves is not a constant offset but a change of slope.

Question 4. Your finished head shows a stray-light problem and you are choosing an aperture. Why does the course not simply specify the 0.5 mm aperture the 361T offers as an option, which would sample the most precisely placed patch?

• Because a smaller aperture is harder to drill accurately
• Because a smaller aperture passes less light, so the signal falls while the stray-light floor does not, and because the sample becomes small enough for grain to add scatter to repeated readings
• Because ISO 5-2 forbids apertures below 1 mm
• Because a 0.5 mm aperture would not fit within a Stouffer step

Show the answer and why

Answer: Because a smaller aperture passes less light, so the signal falls while the stray-light floor does not, and because the sample becomes small enough for grain to add scatter to repeated readings

Two effects push the same way. Signal scales with aperture area, so going from 2 mm to 0.5 mm divides the light by sixteen while any light leaking in around the edges of the head is unchanged — the ratio that sets the usable density range gets sixteen times worse. And at 0.5 mm you are reading a few thousand silver clumps rather than a few hundred thousand, so the granularity of the film itself appears as scatter between nominally identical readings. Option four is arithmetically wrong: 0.5 mm fits inside a 6.0 by 12.7 mm step with room to spare, which is why the option exists on a commercial instrument built to hold its own stray light down. Fine apertures are for reading fine detail, and they are bought with light.

Question 5. A print reads 1.90 with your reflection head when you present it face-on to the port, and 1.72 when it is tilted a few degrees. It is a glossy paper. What has happened, and what does it tell you about a matt print measured the same way?

• The head is drifting; repeat after the warm-up period
• The tilt has let part of the specular reflection into the detector, adding light that carries no image information, so the density reads low; a matt print would show much less change because it has little specular component to admit
• The tilt has increased the path length through the paper, raising the true density
• The reading is correct in both positions; reflection density is defined without reference to angle

Show the answer and why

Answer: The tilt has let part of the specular reflection into the detector, adding light that carries no image information, so the density reads low; a matt print would show much less change because it has little specular component to admit

The whole purpose of the 45/0 arrangement is to send the specular reflection away from the detector, and a tilt of a few degrees is enough to start steering it back in. The extra light is a picture of the lamp rather than of the print, so it looks like transmission that is not there and the density falls. The matt comparison is the diagnostic: a matt surface scatters its surface reflection over a wide angle, so there is no concentrated specular beam to catch, and its reading barely moves. That asymmetry is why the course insists that the sample lies flat against the port and why it declares its reflection scale relative — an angular discipline you have not measured is an uncertainty you cannot state.

Check my answersClear

Further experiments you are equipped for

Measure your own Q. Read one step of a developed wedge with the detector close and wide, then again with a 6 mm tube of blackened card between film and detector to restrict the collection cone. Repeat across the whole wedge and plot Q against diffuse density. You will have measured, for your film and your developer, the quantity this page refused to quote.

Measure the neutrality of your own silver. Read the same step with a green, a blue and a red LED at matched signal levels. A perfectly neutral image gives three equal densities; the departure is your material’s own colour, and it is usually largest at the toe where the silver is finest.

Find your base’s ultraviolet. When the alternative-process cluster gives you a UV source, read a piece of clear, fixed, washed film base in green and again in UV. The difference is the head start your negatives are giving away before any image is present.