Spectral Sensitivity and Colour Response

A silver bromide crystal cannot see a red apple. Nothing you do to the exposure will make it: a red photon does not carry enough energy to lift an electron across the gap, so however many arrive, nothing happens. For the first thirty-four years of photography that was simply what photography was, and every portrait, every landscape and every copy of a painting made before 1873 carries the consequences in its tones.

Then a dye put there for an entirely different purpose changed it, and the change was so complete that the modern eye reads an unsensitised photograph as strange without being able to say why.

Intrinsic sensitivity, and the shape of a colour-blind photograph

The threshold argument is the one from the silver halides, in the compact form Hamamatsu’s detector note gives:

λ = 1240 / E

Threshold wavelength from threshold energy

with λ in nanometres and E in electronvolts. A blue photon at 450 nm carries 2.76 eV; a green one at 550 nm, 2.25 eV; a red one at 650 nm, 1.91 eV. Somewhere between them is the crystal’s threshold, and below it the photon is not absorbed in a way that frees an electron.

This course quotes no band gap for any silver halide, for the reasons that page sets out. What it has are statements from people who measured where sensitivity stops. Vogel’s own textbook gives the ranking directly: chloride of silver is most sensitive to violet; bromide of silver is also sensitive to green; iodide of silver only to violet and indigo; mixtures of iodide and bromide only to blue and green. ILFORD’s contrast-control sheet says the same thing about modern materials in one sentence: all chloro-bromide black-and-white emulsions are blue sensitive with a slight sensitivity to green light.

Eder’s summary of the consequence is blunt: red, yellow and deep green acted little or not at all, so the daguerreotype, the dry collodion process and gelatine silver bromide alike were colour blind.

📷Why old photographs have white skies and black lips

Two visible signatures, and both follow from one graph.

Blue sky prints white. A blue sky is, to a blue-sensitive plate, the brightest thing in the landscape by a very large margin — brighter, relative to everything else, than it looks to your eye. Expose for the land and the sky goes far past the shoulder of the curve and records as an undifferen- tiated maximum. On the print that is paper-base white, and any cloud in it disappears, because a white cloud and a blue sky are almost the same to a plate that only counts blue. Nineteenth-century landscape photographers routinely printed a second negative of a sky into the blank; it was not artistic licence but repair work.

Reds print black. A red object reflects the light the plate cannot use, so it records as though it were in shadow. In portraits, lips, cheeks and any redness in the skin come out dark; freckles and blemishes are exaggerated; a red dress reads as black. ILFORD’s Ortho Plus sheet still says so of a current film: its lack of red sensitivity means reds appear much darker than normal, and offers this as a usable effect rather than a defect.

And the practical grind of it. With the whole of the spectrum’s red and much of its green wasted, exposures were long, and they stayed long until dye sensitisation roughly doubled the usable light.

Vogel, 1873, and a dye that was in the plate for another reason

The discovery is the most famous accident in photographic chemistry, and the detail that makes it intelligible is usually left out.

In the early 1870s, collodion dry plates suffered badly from halation, and makers were fighting it by adding coloured matter to the collodion — the same problem, and one of the same answers, as the previous page. Stuart Wortley made such a plate commercially in England, containing rubber, gallic acid, uranium nitrate and a yellowish-red dye, corallin, put there to stop actinic light penetrating the film and reflecting back off the glass. As Eder records, such plates did show little halation.

They also, Hermann Wilhelm Vogel noticed in 1873 while photographing the solar spectrum with a small spectrograph from the Berlin Academy, showed a greatly increased sensitivity to the green of the spectrum, which was unknown until then. The insight was not the observation but what he made of it: that this was a specific action, an increase in sensitivity caused by the admixed dye.

Then he tested the idea properly, and the test is what turns an accident into a discovery. Corallin absorbs yellow and green, and it sensitised silver bromide collodion for yellow and green. Green aniline dyes sensitised into the red. The sensitised region tracks the dye’s own absorption band — which is the whole mechanism, stated as an empirical rule sixty-five years before anyone could explain it in terms of energy levels.

🕰It was not believed

Vogel exhibited his first spectrum photographs on colour-sensitised collodio-bromide plates at the Berlin Society for the Promotion of Photography on 17 October 1873, and published the same year, with the fuller account following in Poggendorff’s Annalen in 1874.

Monckhoven, in Ghent, repeated the experiments and got nothing. So did Carey Lea, working with coloured glass plates. Vogel spent years defending the result in print against Monckhoven, Lea and Spiller. Eder records the eventual explanation of the failures, and it is a nice lesson in experimental design: Monckhoven had better apparatus, spectrographs of greater dispersion, and therefore weaker spectra — and the sensitisers of the day were feeble enough that the effect showed plainly only in Vogel’s small direct-vision instrument used in strong sunlight. The first support came from the physicist Edmond Becquerel in 1874, who found extra sensitivity bands in chlorophyll-dyed silver bromide collodion, and that ended the controversy in Vogel’s favour.

The dyes then arrived quickly. Erythrosine became the standard orthochromatic sensitiser — Eder records Lowy and Plener in Vienna producing erythrosin-sensitised gelatine bromide emulsions from 1884 on his own directions. Cyanine proved the best sensitiser for orange-red. Vogel’s own azaline, a 1884 mixture of quinoline red with cyanine, sensitised green, yellow and orange together. By 1911 the Encyclopaedia Britannica could list pinaverdol, pinachrom and pinacyanol as commercial dyes and note that pinacyanol confers on a silver bromide plate as high a degree of sensitiveness for red as erythrosin does for yellow. Modern films still use dyes of the cyanine class; Kodak’s own emulsion patent specifies one.

⁇Who made the first panchromatic plates?

Eder credits Vogel, with the azaline plates of 1884, and calls him “the creator of the first panchromatic plate” in so many words. He also records azaline’s shortcomings — feebly sensitive, not very stable, needing strong filters to subdue the blue — and says it could not compete in orthochromatic work with erythrosin or with the isocyanines of Miethe and König.

The claim you will meet far more often is that Wratten and Wainwright made the first commercial panchromatic plates in 1906. This course has found no source in its corpus that states that, and it therefore does not repeat it. Wratten and Wainwright certainly existed and made plates: Hurter and Driffield used their material and name it repeatedly. The 1906 panchromatic claim is left unverified here rather than asserted.

What is not in dispute is that panchromatic materials were in ordinary commercial use by 1911, when the Britannica describes them as now largely manufactured and distinguishes them from the older “isochromatic” plates most sensitive to yellow and green.

How a sensitising dye works

The dye is not dissolved in the gelatin. It is adsorbed onto the surface of the crystal, held there in contact with the lattice, and that contact is the whole point.

The sequence has three steps.

The dye absorbs a photon that the crystal cannot use — a green or red one, below the crystal’s own threshold. The dye can absorb it because its own energy levels are much closer together than the crystal’s band gap.

The excited dye hands an electron to the crystal. The electron goes into the conduction band, where it is indistinguishable from one freed by an ultraviolet photon absorbed by the crystal itself.

Everything after that is the ordinary Gurney–Mott sequence of the latent image page: the electron is trapped, an interstitial silver ion arrives, a silver atom forms, and the cluster grows. The dye has not changed what the crystal does with an electron; it has changed where the electrons can come from.

Where a sensitising dye sits, in energy

1. 1Valence band, full — the crystal's own electrons
2. 2Conduction band, empty — an electron here can travel and be trapped
3. 3The only photon the crystal can use — blue or ultraviolet: above the gap
4. 4The adsorbed dye, ground and excited states — levels close together, so a green or red photon is enough
5. 5Electron transfer into the conduction band — this single step is spectral sensitisation

Energy increases upwards and the axis deliberately carries no numbers: this course has verified no band gap for a silver halide and no level for a sensitising dye. What the drawing claims is only the ordering — the dye's excited state above the conduction band edge — and that ordering is what Kodak's patent is specifying when it selects a cyanine dye by its half-wave potentials.

⚗How you can tell the model is right from a patent

The electron-transfer picture is not something a photographer has to take on trust. It is visible in how the industry specifies its dyes.

Kodak’s double-jet emulsion patent claims grains “spectrally sensitized with a cyanine dye having an anodic half-wave potential less than 1.0 volt and a cathodic half-wave potential less than 0.8 volt”. Half-wave potentials are electrochemical measurements of how readily a molecule gives up or takes an electron. If a sensitising dye worked by, say, absorbing light and passing on the energy, its electrode potentials would be beside the point. Specifying a dye by its potentials only makes sense if what it has to do is donate an electron to the crystal at the right level.

The same patent supplies the other half of the argument, from the failure side: it notes that on conventionally made grains the sensitising dye is prone to desensitise the grains in the blue region of the spectrum. A dye that sits on the surface can take electrons as well as give them, and a badly matched one steals blue sensitivity while it lends green. Sensitisation is not free.

!J-aggregates: what the term means and what this course cannot source

You will meet the term J-aggregate in any serious account of sensitisation, and the covers list for this page asks for it at school level, so here it is with its provenance marked.

The standard account, which this course states as the account the literature gives: when a cyanine dye adsorbs onto a silver halide surface at sufficient concentration, the molecules do not sit independently but pack side by side into an ordered array. That array behaves as a single co-operative absorber rather than as a crowd of separate molecules, and its absorption band is narrower and shifted towards longer wavelengths than the same dye’s band in dilute solution. The shift is useful — it extends sensitivity further into the red than the free dye would reach — and the sharpening is useful, because it lets a maker place a sensitivity peak precisely.

This course has read no source describing J-aggregates, in the corpus it holds. The paragraph above is offered as the received account and not as a verified one, and if you need the structure, the spectral shift or the aggregation conditions, go to the photographic-science literature rather than to this page.

The material classes, and the chart each maker publishes

Five classes of black-and-white material, defined by how far into the spectrum they respond. The diagram below draws each class as what it adds to the one before, so that a panchromatic film is the first three bands together and an extended-red film is the first four.

What each class of material adds to the one before

• Every material: blue and UV (320–500 nm) — undyed, blue-sensitive; ILFORD: blue sensitive with a slight sensitivity to green
• Orthochromatic adds green (500–600 nm) — handled under a deep red safelight, ILFORD 906
• Panchromatic adds red (600–680 nm) — total darkness, on every maker's instruction
• Extended red adds this (680–740 nm) — ILFORD SFX 200: extended red sensitivity up to 740 nm
• True infrared (740–780 nm) — no current sheet in this corpus, so no upper limit is quoted

Read the bands cumulatively from the left: each shows what one class gains over the class before it, so a panchromatic film responds across the first three bands together. Only the 740 nm figure is published; the other boundaries are this course's rendering of verbal statements and safelight recommendations, because the makers publish sensitivity as a picture rather than as numbers. 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.

Every current data sheet in this course’s corpus presents spectral sensitivity the same way: as a wedge spectrogram, headed, in ILFORD’s case, “Wedge spectrogram to tungsten light (2850 K)”. It is made by exposing the material through a spectrograph behind a graduated grey wedge, so that the resulting shape shows, for every wavelength, how far down the wedge the material still registered. A tall region means high sensitivity there; where the shape drops to the baseline, the material is blind.

✕Three things a wedge spectrogram will not tell you

It is not calibrated in absolute sensitivity. The height is relative, and comparing the heights on two different sheets tells you very little.

It is measured to a stated illuminant, and the illuminant matters. ILFORD’s spectrograms are made to tungsten light at 2850 K, which is far redder than daylight. The same film exposed in daylight receives a different mix of wavelengths, which is exactly why the filter factors in the next section come in two columns.

It does not end where the shape ends. Kodak’s warning is the general one and this course keeps repeating it: the colour sensitivity of most emulsions does not end abruptly at a particular wavelength — most emulsions are somewhat sensitive to colours outside the intended range. A spectrogram has a printing threshold, and beyond it there is response you cannot see on the chart.

Variable-contrast paper, and why its curve has two humps

A graded paper has one emulsion and one response. A variable-contrast paper has, in ILFORD’s own description of Multigrade, a mixture of three separate emulsions, each of them a basic blue-sensitive emulsion to which is added a different amount of green sensitising dye. All three have the same inherent contrast and the same speed to blue light; they differ only in how sensitive they are to green.

That single fact produces the whole system.

Expose with blue light and all three emulsions respond, and respond at the same speed. They add together, so density climbs steeply over a narrow range of exposure: high contrast.

Expose with green light and only the emulsions carrying enough green dye respond, and they respond at very different speeds. Their curves are staggered along the exposure axis, so the sum climbs gradually over a wide range: low contrast.

Mix the two in any proportion and you get anything in between. ILFORD’s filters do exactly that: a magenta filter absorbs green and transmits blue, so it hardens; a yellow filter absorbs blue and transmits green, so it softens. Their filter set runs 00 to 5 in half grades, and — a detail that saves a great deal of paper — the exposure time is the same for filters 00 to 3½ and doubles for 4 and above.

Variable-contrast paper: two responses, and where a safelight has to sit

• Blue response (360–500 nm) — all three emulsions, same speed: their curves add, so blue light gives hard contrast
• Green response (500–580 nm) — only the dyed emulsions, at staggered speeds: green light gives soft contrast
• Paper safelight (575–700 nm) — FOMA: 575 nm and higher — yellow, yellow-green, amber or orange

The two sensitivity bands are drawn to show the principle: ILFORD publish the spectral sensitivity of MULTIGRADE papers as a chart without a wavelength scale, and these edges are this course's rendering of their written description of three emulsions carrying different amounts of green sensitising dye. The 575 nm safelight edge is FOMA's published figure. Notice how little gap there is between the green response and the safelight, and that a magenta filter removes the green band while a yellow filter removes the blue. 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.

Two practical consequences worth carrying into the darkroom. A safelight for variable-contrast paper must clear both humps, not just the blue one, which is why paper safelights are orange or light brown rather than yellow-green — and why, on the diagram above, FOMA’s published safelight boundary sits only a little way beyond where the green response ends. And the colour of your enlarger’s light source changes the contrast: ILFORD warn that cold-cathode and pulsed-xenon sources not designed for variable-contrast papers may give a reduced contrast range, and that a condenser enlarger gives about an extra grade of contrast compared with a diffuser.

Safelights follow from sensitivity

A safelight is not a tradition. It is an argument, and the argument is entirely on this page: the lamp emits where the material does not respond. ILFORD define the word with exactly that restraint — for a particular photographic material, a safelight is the illumination that does not cause a significant visible change to it during use — and add the qualification that makes it honest: the word “safe” is relative, since in most cases a sensitised material will eventually be affected by its own safelight.

Read down a manufacturer’s table and the whole of this page reappears as a column of products.

Material	ILFORD filter	Kodak filter	Colour
Blue-sensitive papers, including Multigrade	SL1 or 902	OC	orange / light brown / light amber
Fast blue-sensitive line and technical films	904	1	dark brown / red
Orthochromatic materials	906	1A or 2	dark red
Very slow panchromatic materials	907	3	dark green
All panchromatic materials	908 (with extreme care)	—	very dark green
Panchromatic papers	—	10 or 13	dark amber / amber
Panchromatic and infrared films	none	none	total darkness

The ordering is the ordering of the first spectrum diagram, read from the far end: the further into the red a material sees, the further into the red a safelight would have to sit, until there is no room left and the answer is darkness.

The two makers do not agree at the far end, and it is worth seeing where. ILFORD list a 908 very dark green filter for all panchromatic materials, adding that although it is designed for the maximum possible efficiency it must be used with extreme care, and that fast panchromatic materials must not be exposed to direct light from it for any appreciable length of time. Kodak simply put panchromatic black-and-white films on their total-darkness list. Both positions are defensible — a very dark green filter does transmit where a panchromatic emulsion is least sensitive — and the practical reading is that a safelight for panchromatic film is a specialist tool with almost no margin, not a working light. Kodak’s total-darkness list is short and absolute — panchromatic black-and-white films and plates, colour camera films, duplicating and internegative films, colour reversal and colour negative papers, and high-speed infrared films.

⚠The conditions are part of the recommendation

A safelight filter on its own is not a safelight recommendation. Both makers attach three further conditions, and dropping any of them voids the rest.

Wattage. Kodak’s table specifies 15 W for most direct applications and 7½ W for the amber filters used with panchromatic papers. ILFORD specify a 15 W bulb for the SL1 and 902.

Distance. No direct illumination closer than 1.2 m (4 ft) from the working surface, in both makers’ tables.

Time. ILFORD’s general recommendation for their black-and-white papers is that they should be safe for up to 4 minutes under an SL1 or 902 at 15 W and not less than 1.2 m. Kodak define safe time formally: any exposure time less than or equal to one-half the time required for a safelight to produce a detectable change, which builds in a factor of two.

And the filter ages. ILFORD note that a filter in use for several hours a day fades, transmitting more and more of what it is supposed to absorb, and advise changing it annually and recording the installation date. Kodak suggest every three months at 8 to 12 hours a day.

None of this is a substitute for testing, which is the next paragraph.

The proof is a fog test, and this course treats it as an acceptance test the room passes or fails. The method is Kodak’s and ILFORD’s, and the two agree on its shape: expose a sheet to a light grey tone, give parts of it increasing times under the safelight, develop it in total darkness alongside an unexposed control, and read the result. Kodak use 0, 1, 3 and 7 minutes; ILFORD use 0, 1, 2 and 4.

Two details from those procedures matter more than the times. First, the test must include a strip that was given an enlarger exposure before the safelight exposure, because paper is more sensitive to safelight fogging after it has been exposed than before — the latensification effect of the reciprocity page. Second, Kodak state the failure mode plainly: poor safelight conditions can produce a loss in photographic quality before actual fogging is visible. You do not lose black borders; you lose highlights and contrast, and you blame your negatives for months.

Station 4 of this part’s darkroom session is that test, run on your own room.

✦Why an orange LED is not automatically a safelight

Because the colour you see is not the spectrum the lamp emits.

Kodak put it exactly: the apparent colour of a safelight filter is only a partial indication of its transmission characteristics, and coloured bulbs or improvised safelights may appear to be the right colour but may actually emit light, or other radiant energy, that will fog a photographic emulsion. ILFORD say the same of filters. A cheap coloured LED, or a white LED behind an orange gel, can have a substantial blue component that your dark-adapted eye simply does not weight heavily, because the eye at low light levels is not a spectrometer.

There is a second reason, and it is the more interesting one. Ware records that a print-out silver image absorbs visible light itself and passes the energy on to the halide around it — the Becquerel effect, which Becquerel discovered in 1841 and which Abney later used to photograph the red and infrared before dye sensitisation existed. Ware’s conservation conclusion is that early chloride-fixed prints may even be at risk by prolonged exposure under yellow or red photographic safelights. A material that has already been partly exposed is not the same material you tested with.

So: the fog test, on your lamp, in your room, at your working distance. Not the colour, and not the label on the box.

Filters, and the tonal control they buy

If the material responds across the spectrum, you can choose which part of it to use. That is what a contrast filter over the lens does: it removes some wavelengths before they reach the film, so subjects that reflect those wavelengths record darker.

Every filter costs light, and the cost is the filter factor — the number you multiply the exposure by. ILFORD publish factors on the sheet for each film, and two of their tables together teach more than either alone.

Filter (Wratten equivalent)	Factor on SFX 200	Stops
8 yellow	2	1
12 deep yellow	2.3	1⅓
21 orange	2.4	1⅓
25 red	2.8	1½
29 deep red	3	1⅔
89B very deep red	16	4

🧮Why the same filter costs different amounts on different films and in different light

Now the second table, for the orthochromatic ILFORD Ortho Plus, with two columns:

Filter	Factor in daylight	Factor in tungsten
104 Alpha (yellow)	2.5	1
109 Delta (deep yellow)	5.5	3
304 Tricolour blue	3	5
404 Tricolour green	8	4.5

Three things fall out of it, and they are all consequences of this page.

A yellow filter costs nothing at all on ortho film under tungsten light. Factor 1. A yellow filter works by removing blue; tungsten light at 2850 K contains very little blue to remove; so there is nothing to pay. In daylight, which is blue-rich, the same filter costs 2.5×.

The blue filter reverses the pattern, costing 3× in daylight and 5× in tungsten, for the same reason read the other way: under tungsten there is less blue for it to pass.

A green filter costs more in daylight than in tungsten — 8× against 4.5× — because Ortho Plus is most sensitive to blue, and in blue-rich daylight a green filter is throwing away the very light the film uses best.

The general rule. A filter factor is not a property of the filter. It is a property of filter, light source and the material’s own spectral sensitivity together, which is why the number is printed on the film’s sheet and not on the filter. Compare the deep red on SFX 200 at 3× with the same class of filter on an ordinary panchromatic film, where it costs considerably more: SFX’s extended red sensitivity means less of what the filter passes is wasted.

One warning both ILFORD sheets carry, and it costs people negatives: through-the-lens metering can under-expose by as much as 1½ stops with deep red and orange filters, because the meter’s own spectral response is not the film’s. Meter without the filter and apply the factor.

What each filter does to a picture, on a panchromatic film:

Yellow is the mild, general-purpose one. It darkens blue sky enough for white cloud to separate from it, and it corrects the tendency of a panchromatic film to render blue skies lighter than the eye remembers. Foliage and skin are barely touched.

Orange takes the same effect further: a stronger sky, more visible cloud, and haze in a distant landscape begins to clear, because haze is scattered short-wavelength light.

Red is dramatic. Blue sky goes to near-black, cloud stands out white, haze largely disappears, and green foliage darkens. On an extended-red film the foliage goes the other way: ILFORD’s SFX sheet promises that with a very deep red filter skies can be rendered almost black, and most green vegetation almost white, because leaves reflect strongly beyond 700 nm where SFX still responds and the eye does not.

Green lightens foliage and grass, and darkens skin — which makes it useful for a landscape with pale grass and unflattering for most portraits, since it emphasises exactly the redness in a face that a portraitist usually wants suppressed.

And for skin generally: a panchromatic film with no filter, or with a light yellow one, is what renders a face the way the eye expects. Orthochromatic material, blind to red, darkens lips and any flush or blemish; that is why nineteenth-century portrait sitters look severe, and why Ortho Plus is offered today as an effect.

Desensitising, and what comes next

One last twist, because it completes the logic. If a dye can lend sensitivity by handing electrons to the crystal, another can remove it — and doing so without destroying the latent image already there would let you develop by a light bright enough to work in.

That is exactly what happened. Eder records that in 1901 Lüppo-Cramer found that certain developers of the paramidophenol class, and ferrous oxalate, greatly reduce the sensitivity of unexposed silver bromide without destroying its capability of developing the latent image. Nineteenth-century photochemists had assumed the two must go together; they do not. In 1920 the same worker found far better desensitisers among the safranine dyes, and published a method for developing highly sensitive and colour-sensitive plates in ordinary yellow light without fog — which, for anyone who had been developing panchromatic plates by touch, was a considerable improvement in life.

The hand-off from here is to Part V, where you make an emulsion of your own and have to decide what it should be able to see. The choice is not abstract: an undyed chloride emulsion is blue-sensitive, easy to handle under a safelight and useless for anything but contact printing from a negative; adding a green sensitiser turns it orthochromatic and takes away most of your working light. Everything on this page becomes a decision on the coating bench.

Recap

• An undyed silver halide sees only blue and ultraviolet, because a longer-wavelength photon does not carry enough energy. Vogel’s own ranking: chloride most sensitive to violet, bromide also to green, iodide only to violet and indigo.
• That is why early photographs have white skies and black reds. Blue sky overwhelms a blue-only material; red objects reflect light it cannot use. ILFORD still sell a film that behaves this way and describe the effect in the same words.
• Vogel, 1873. A dye put into a plate to stop halation turned out to sensitise it, and he saw what it meant. The sensitised region tracks the dye’s own absorption band. It was disbelieved for years, and the failures to reproduce it had an instrumental explanation.
• The mechanism is electron transfer. The adsorbed dye absorbs a photon the crystal cannot, and hands an electron into the conduction band; everything after that is the ordinary latent-image sequence. Kodak specify sensitising dyes by their half-wave potentials, which only makes sense in that model.
• J-aggregates are the received account of how the dye packs on the surface and why its band shifts red and sharpens. This course has read no source for them and marks the paragraph accordingly.
• Five classes: blue-sensitive, orthochromatic, panchromatic, extended-red (SFX 200 to 740 nm) and true infrared. Makers publish sensitivity as a wedge spectrogram to a stated illuminant, not as numbers.
• Variable-contrast paper is three emulsions differing only in how much green dye they carry. Blue light exposes them all at one speed and gives hard contrast; green light exposes them at staggered speeds and gives soft.
• A safelight is an argument about wavelength, with a wattage, a distance and a time attached, and it is proved by a fog test rather than by its colour. Panchromatic and infrared films get no safelight at all.
• A filter factor belongs to the film, the filter and the light together, which is why a yellow filter costs 2.5× on ortho film in daylight and nothing at all under tungsten.

Check your understanding

Question 1. A red apple sits among green leaves in daylight. Describe how it records on blue-sensitive, orthochromatic and panchromatic material.

• Blue-sensitive: apple black, leaves black. Orthochromatic: apple black, leaves mid-grey. Panchromatic: apple and leaves both mid-grey, and separable
• All three render it identically; only the speed differs
• Blue-sensitive: apple white. Orthochromatic: apple grey. Panchromatic: apple black
• Blue-sensitive: apple grey, leaves white. Orthochromatic and panchromatic: both black

Show the answer and why

Answer: Blue-sensitive: apple black, leaves black. Orthochromatic: apple black, leaves mid-grey. Panchromatic: apple and leaves both mid-grey, and separable

Work it out from what each material can absorb. A blue-sensitive emulsion uses neither the red the apple reflects nor the green the leaves reflect, so both subjects record as though in shadow and both print dark, with nothing to tell them apart. An orthochromatic emulsion adds green, so the leaves lift to a mid-tone while the apple stays dark: this is the classic ortho rendering, and ILFORD describe it on the Ortho Plus sheet by saying reds appear much darker than normal. A panchromatic emulsion uses both, so both record according to their brightness and the picture reads as the eye expects. Notice that the separation you want here is a spectral question, not an exposure question: no exposure of a blue-sensitive plate will separate the apple from the leaves.

Question 2. Why does a magenta filter increase the contrast of a variable-contrast paper?

• Because magenta light is more energetic than green light
• Because it absorbs green and passes blue, so all three of the blended emulsions expose at the same speed and their curves add rather than staggering
• Because it reduces the exposure, and less exposure gives more contrast
• Because it fogs the green-sensitised emulsion

Show the answer and why

Answer: Because it absorbs green and passes blue, so all three of the blended emulsions expose at the same speed and their curves add rather than staggering

ILFORD describe Multigrade as a mixture of three emulsions, each a basic blue-sensitive emulsion with a different amount of green sensitising dye. All three have the same speed to blue and the same inherent contrast, so under blue light they respond together and their characteristic curves add up into one steep curve over a narrow exposure range. Under green light only the more heavily dyed emulsions respond, at very different speeds, so the curves are staggered along the exposure axis and their sum is long and shallow. A magenta filter absorbs green and passes blue, and a yellow filter does the reverse. Note also that the useful engineering detail follows from the same design: exposure time is unchanged from filter 00 to 3 and a half, and doubles above it.

Question 3. From a paper sheet you learn only that it is blue-sensitive with slight green sensitivity. You have an orange LED. Can you use it as a safelight?

• Yes, because orange is well clear of blue
• Yes, provided you keep it 1.2 m away
• Not until you have run a fog test, because the apparent colour of a source is only a partial indication of what it emits and a white LED behind an orange gel may carry a substantial blue component
• No, because LEDs are never suitable as safelights

Show the answer and why

Answer: Not until you have run a fog test, because the apparent colour of a source is only a partial indication of what it emits and a white LED behind an orange gel may carry a substantial blue component

Kodak state the principle directly: the apparent colour of a filter is only a partial indication of its transmission characteristics, and improvised safelights may appear to be the right colour while emitting light or other radiant energy that fogs an emulsion. Your dark-adapted eye is a poor spectrometer, and a white LED is not a narrow-band source. The answer is not a rule about LEDs, which can make perfectly good safelights, but a test: expose a sheet to a light grey tone, give parts of it increasing times under the lamp, develop in darkness against an unexposed control, and read the result. Include a strip that received its enlarger exposure before the safelight exposure, because paper fogs more readily after exposure than before.

Question 4. ILFORD give a yellow filter on Ortho Plus a factor of 2.5 in daylight and 1 in tungsten light. Why the difference?

• Tungsten lamps are dimmer, so the correction is smaller
• The film is faster under tungsten light
• A yellow filter works by removing blue, and tungsten light at 2850 K contains very little blue to remove, so there is almost nothing to pay for
• The factor in tungsten is a misprint for 2.5

Show the answer and why

Answer: A yellow filter works by removing blue, and tungsten light at 2850 K contains very little blue to remove, so there is almost nothing to pay for

A filter factor is not a property of the filter alone. It is the product of what the filter removes, what the light source contains, and where the material is sensitive. Daylight is rich in blue and Ortho Plus is most sensitive to blue, so a yellow filter in daylight throws away a large part of the light the film could have used, costing 1.3 stops. Tungsten light at 2850 K is blue-poor, so the same filter removes almost nothing the film was going to use, and the factor falls to unity. The same table shows the pattern reversing for a blue filter, which costs 3 times in daylight and 5 times in tungsten. This is also why the factor is printed on the film sheet rather than on the filter.

Question 5. What did Vogel actually discover in 1873, and what made it a discovery rather than an observation?

• That silver bromide is sensitive to green, which nobody had noticed
• That a dye added to a plate for an unrelated reason had increased its green sensitivity, and that this was a specific action caused by the dye, whose sensitising region matched the dye's own absorption band
• That collodion plates suffer from halation
• That coloured objects photograph in different tones

Show the answer and why

Answer: That a dye added to a plate for an unrelated reason had increased its green sensitivity, and that this was a specific action caused by the dye, whose sensitising region matched the dye's own absorption band

The corallin in Stuart Wortley's commercial dry plates was there to prevent halation, and its effect on green sensitivity was an accident. What Vogel contributed was the interpretation and then the test: he recognised it as sensitisation by the admixed dye rather than as a peculiarity of the plate, and he confirmed it by showing that corallin, which absorbs yellow and green, sensitises for yellow and green, while green aniline dyes sensitise into the red. That rule linking the dye's absorption band to the sensitised region is the empirical form of the mechanism, arrived at sixty-five years before the energy-level account. It was also disbelieved for years, and the failures to reproduce it turned out to be a consequence of better spectrographs giving weaker spectra.

Question 6. Two of the numbers on this page carry the label "published" and the rest of the band edges are drawn. Which pair is published?

• The band gap of silver bromide and the wavelength of a red safelight
• The 740 nm upper limit of SFX 200 is extended red sensitivity, and the 575 nm lower limit of the safelight FOMA specify for its papers
• The 500 nm edge of blue-sensitive paper and the 680 nm edge of panchromatic film
• The peak wavelengths of the two humps of a variable-contrast paper's response

Show the answer and why

Answer: The 740 nm upper limit of SFX 200 is extended red sensitivity, and the 575 nm lower limit of the safelight FOMA specify for its papers

ILFORD state 740 nm for SFX 200 in words on the data sheet, and FOMA state that their papers are routinely processed under indirect illumination of wavelength 575 nm and higher. Every other edge on the spectrum diagram is this course's rendering of a verbal statement or a safelight recommendation, and the caption says so. The reason so few numbers are available is that manufacturers publish spectral sensitivity as a wedge spectrogram, a picture without a calibrated vertical scale, and publish nothing at all for a band gap. That is why the safelight tables are as useful as they are: a safelight recommendation is one of the few places a maker commits to a statement about wavelength.

Check my answersClear