Ultraviolet Light Sources for Printing with the Alternative Processes

by Sandy King

This article is presented in two parts. Part One is a general introduction to the types of light sources that are used for printing with the alternative processes. In this section we discuss the general operating parameters of some of the various light sources commonly used in alternative printing. Part Two compares the two types of light systems that are practical to acquire and use, weighing such competing concerns as cost, ease of operation, efficiency, and image quality: HID lamps (metal halide and mercury vapor), and a bank of fluorescent tubes.





Most of the alternative processes are much less sensitive than modern silver gelatin papers and must be printed by contact. These processes require a light source that emits much of its radiation in the Ultraviolet A (320-400nm) range. Some processes also have considerable sensitivity to light in the violet and blue range up to about 450nm, but sensitivity falls off to near zero in the green at about 520nm. Most of the alternative processes are also highly sensitive to Ultraviolet B (280-320nm) but sources of this type should be used with care because UVB causes skin cancer and cataracts. It should also be noted that approximately 95% of Ultraviolet B is absorbed by the ordinary plate glass typically found in contact printing frames and vacuum easels so in practice most UVB radiation is lost to the processes unless one invests in the special glasses that allow this radiation to pass.

There are many sources of UV light: the sun, sun lamps, self-ballasting mercury vapor lamps that screw into a normal incandescent light fixture, ballasted discharge lamps such as mercury vapor and metal halide, fluorescent tube banks, and plate-burners manufactured for the graphic arts. Any of these light sources may give good results with most of the alternative processes. The following information provides you with the general characteristics of each of these sources, hopefully helping you to select the best light for your circumstances.

The Sun

The sun is one of the strongest sources of ultraviolet light but it varies greatly in intensity according to season, time of day, atmospheric conditions, and geographic location. In some areas of the country the sun can be a very reliable light source. Its output is especially consistent on clear days between 10 a.m. and 2 p.m. but in many locations printing is not possible from late fall to early spring because of the shortness of the days and because the sun is so low in the sky. On the other hand in some locations the sun is available for 6-8 hours for 320 or more days of the year.

It is possible to print in direct sun or in open shade. It should be noted that with most processes images of greater contrast will result from printing in the shade than with direct sun. Direct sun is of course much faster, by around 2-4 stops.

Sun lamps

Sun lamps are manufactured for tanning the human body. They produce quite a bit of heat and therefore must be placed at least 18-24" from the printing frame. At that distance exposures tend to be rather long in comparison with other units. Being something of a point source they give better results than diffuse light sources with reversed negatives. Sun lamps must be allowed to warm up for a couple of minutes to reach full output, and once the lamp has been turned off it must be allowed to cool before turning it on again. Sun lamps produce high levels of Ultraviolet B radiation and the operator should be shielded from the light as much as possible.

Self-ballasted mercury vapor lamps

Self-ballasting mercury vapor lamps are bulbs that screw into an ordinary incandescent fixture. They are not widely available in sizes of over 175-watts and because of their limited power must be placed very close to the exposing plane, limiting their use to prints of no more than about 5X7" in size.

Ballasted HID lamps (mercury vapor and metal halide)

These lamps are commonly available in sizes from 175 watts up to 1000 watts. Such lights have a discontinuous spectrum but much of the radiation they produce is in the useful Ultraviolet A range.

The ordinary 175-watt street lamps can be used to make prints up to 5X7" with the bulb at 12-15 inches from the printing frame, and prints up to 8X10" with the bulb at about two feet from the frame. This light must be allowed to warm up for at least five minutes prior to the printing session. The unit is slow and produces a lot of heat but it is one of the least expensive light sources available.

Large discharge lamps of 1000 watts offer one of the most attractive options for alternative printing and will be described in more detail later in this article.

Bank of fluorescent tubes

A bank of BL, BLB, Actinic, AQUA or Super Actinic tubes is an excellent source of ultraviolet light and may be the best choice for many printers. This type of light source is reasonably fast, provides a large area of even illumination, is quite easy to operate, produces constant output almost immediately on being turned on (especially when using electronic ballast), and may be switched on and off with no delay. A fan should be used to cool the tubes because if they get hotter than about 105°F the light output decreases significantly. Fluorescent tubes do not require any appreciable warm-up time, and they may be restarted immediately. Electric rather than magnetic ballast is to be preferred because it gives a more constant output, runs cooler and generates about 10% more lumens per watt. Four types of fluorescent tubes are in common use: 1) Black Light (BL) tubes, which peak at around 350 nm, 2) Black Light Blue (BLB), a shielded light similar in output to the BL, 3) AQUA, and 4) the Super Actinic (SA) aquarium tube that peaks at around 420 nm. SA tubes produce images of slightly greater contrast, and many reports suggest they are faster than BL and BLB units in platinum printing. AQUA, another aquarium tube, is reported to also work well with platinum.

Plate-burners, or graphic arts printers

The light source of plate-burners is usually a mercury vapor or metal halide discharge lamp, but some units may be equipped with carbon-arc or pulsed-xenon lamps. All of these sources take several minutes to reach maximum output and should be used with a light integrator, an instrument that measures the total amount of light available for exposure.

Plate-burners, which usually come with integral vacuum frames and light integrators, make excellent UV printer for alternative printing. Unfortunately such units are very expensive if purchased new and are in fairly high demand by alternative printers.


Any of the light sources described above might prove to be ideal for a specific set of circumstances. However, when the decision is based on a variety of factors, including cost to acquire (or build), speed, ease of use, and the ability to cover evenly an area large enough to make prints up to 16X20" in size, the decision will probably come down to one of two choices: a bank of 8-12 BL, BLB or Super Actinic tubes, or a 1000 watt HID lamp.

A. Bank of BL, BLB, Aqua or Super Actinic fluorescent tubes

First, here is a little information on the nomenclature of fluorescent tubes. They are usually designated by a series of letters and numbers. For example, the GE Blacklight Blue tube carries the designation F20T12.BLB.

  • The F stands for fluorescent.
  • The next number, 20, indicates the wattage of this particular tube.
  • The T number indicates the diameter of the tube in eighths of an inch. A T 12 tube, for example, is 12/8", or 1 1/2" in diameter.
  • The letters describe the lighting characteristics of the tube. In the above case, BLB designates a Black Light Blue tube.
  • Tubes sometime carry further indicators, such as IS (Instant Start), RS (Rapid Start), or even letters showing kind of usage, for example R for reptiles, A for Aquarium, etc.

Ballast for fluorescent tubes is available in many configurations but a specific ballast unit must be rated for at least the total wattage of the tubes it is meant to power. There are two main types of ballast, magnetic and electronic. The latter has many advantages, not the least of which is the fact that they generate between 10% (T12 tubes) and 30% (T8 tubes) more lumens per watt, which can of course result in faster printing times.

Some fluorescent tubes are available in standard wattage, high output (HO) and very high output (VHO). Standard wattage tubes, be they BL, BLB, AQUA or Super Actinic, vary in wattage according to size. For example, a regular 24" tube is rated at 20 watts, while the 48" tube of the same type is rated at 40 watts. The power of a 24" HO tube is 40 watts, while the 24" VHO is rated at 75 watts. When using tubes of the same type and wattage there is no difference in printing characteristics between a 24" and 48" tube. HO and VHO tubes put out more lumens and, all things being equal should provide faster printing times than standard wattage tubes of the same type. BL and BLB tubes are not widely available in HO and VHO configurations but SA tubes are. Standard wattage BL, BLB, AQUA and Super Actinic tubes of the same size and wattage are interchangeable in UV printing units. HO and VHO tubes, on the other hand, require special ballast, and VHO tubes may also require special end-caps.



The cost of a factory built UV printer capable of making prints up to 16X20" in size is quite high and many alternative printers opt to construct their own exposing units. Most any reasonably handy person can assemble a UV bank of fluorescent tubes with a few hours of work. Good plans are available in several books, including Luis Nadeau's History and Practice of Platinum Printing, and The New Platinum Print by Carl Weese and Richard Sullivan. See also Issue No. 6 of Post-Factory Photography for some design considerations about building your own UV bank.

The major cost of such a project will be the tubes and ballast, which together should account for between 80-90% of the entire project. Although the total cost can vary greatly, depending on choice of tubes and ballast, one should be able buy all of the materials necessary for a 10-12 tube unit that will provide even coverage for prints up to 16X20" in size for around $250.

Should you decide to construct a UV exposing unit with fluorescent tubes one of the most important design decisions is whether to build the unit around two-tube holders with integral ballast, or to use external ballast and individual Bipin holders. If you choose the two-tube holders the tubes will be separated by about 3/4" whereas the use of the individual Bipin holders and external ballast allows one to space the tubes as close together as 1/4". Close spacing of the tubes allows you to use the tubes at about 2" from the exposing plane and still get even lighting. UV banks built with two-tube holders typically have a tube spacing of around 3/4" and should be used at about 4" from the exposing plane for even illumination. In theory the closer spacing, and use of the tubes at a closer distance to the exposing plane, should result in faster printing times. However, in practice the actual difference in printing speed between units with 1/4" and 3/4" spacing is small when used at distances of between 2-4" from the exposing plane.

For construction plans using the individual Bipin holders follow the instructions on the Edwards engineering site.

Construction of UV Fluorescent Bank Based on Two-Tube Holders

This type of unit is much simpler to build as one basically just has to wire together the individual holders, attach them to some kind of support, and connect them to a power source. The disadvantage is that the holders add some unnecessary weight, and perhaps expense, to the construction project. The following instructions provide an outlined guide to building a UV printer with the two-tube holders.

Step 1 — Decide what sized UV bank you need and buy the necessary parts. For a unit capable of printing up to 16X20 inches you will need a minimum of 5-6 two-tube holders and 10-12 tubes.

Step; 2 — Assemble the holders side to side on a piece of plywood oversize plywood, 3/8" to 1/2" thick, leaving about 1 1/2" of free space on all sides. This piece of plywood will be the top of the UV box.

Step 3 — Remove all of the knockout disks from the sides of the holders so that you can run wiring between them.

Step 4 — The holders will have mounting holes at each end. With all of the holders arranged together on the plywood mark the location of these holes.

Step 5 — Remove the holders and cut the plywood piece to size, making sure to leave free a space of about 1 1/2" on all sides. Also drill holes for 10-32 or 10-24 bolts at the mounting marks.

Step 6 — Bolt all of the holders to the plywood top, countersinking the top of the bolts into the outside of the plywood to leave a flat surface.

Step 7 — Wire the holders together and connect to a grounded power cord. To wire, first connect a green wire to all of the holders and run to a common point. Holders usually have designated points for this where you will find a green screw but the ground can just as well be established to any metal part of the holders, including the mounting nuts. Next, connect all of the black wires to a common point, and finally, do the same for all of the white wires. To complete the wiring, connect these three wires to wires of the corresponding color on the extension cord.

Step 8 — Slip the Bipin connection onto the end of the holders. This will not require wiring as the connections come prewired to the ballast. With many two-tube holders there will be uneven spacing between the tubes if they are assembled side by side. This can be rectified by installing half of the Bipin connections on each end away from the indicated spot, creating a common separation for all of the holders. This operation can be very easy or fairly complicated depending on the particular holder in use.

Step 9 — Slip the end caps onto the ends of the tube holders and secure the white metal reflector plates over the tube holders with the small locking butterfly that is provided. Proper installation of the reflector plates is essential for proper operation of the tubes. If the reflector plates are omitted or incorrectly installed the tubes may not turn on.

Step 10 — Now cut the sides of the box, using any good quality 1X8 wood. You will need about 10 linear feet of wood for a unit based on 24" tubes. Cut a 4" diameter hole in the center of one of the end pieces, and in the other end drill about 6-8 holes of about 1 1/2" diameter. The large hole is for a fan, and the small holes are for effective airflow with the fan. The side of the box that is to face forward should be about 2" narrower than the back; this will allow you to slide the contact-printing frame under the tubes.

Step 11 — Assemble the top of the unit to the sides using 1 1/4" wood screws, countersinking the screw heads. Then assemble the sides to each other, using 2" wood screws, countersinking here as well.

Step 12 — Install tubes in the unit and connect to a power source. The tubes should come on immediately and all at the same time. If they do not come on at all the wiring was done incorrectly. If they come on but with hesitancy and with a few tubes off the problem is most likely with the grounding.

Step 13 — Once the wiring has been checked remove the tubes and finish the unit by sanding and applying a coat of varnish or paint.

Tubes Commonly Used in UV Printing

As noted earlier, four types of fluorescent tubes are in common use in UV exposing units, BLB (Blacklight Blue), BL (Blacklight), AQUA and SA (Super Actinic). Some additional information about the characteristics of each type is given below. Also included is information about the Philips Actinic 05 light, which, though not widely available in the US, is also a good light for alternative printing.

Black Light (BL) tubes emit most of their radiation at between 350-370nm but also quite a bit in the deep violet and blue-violet. They are available in a wide range of sizes, including both 24" (20 watt) and 48" (40 watt). HO (high output) BL tubes are also available in some sizes.

Black Light Blue (BLB) — The BLB is filtered with a tube made of dark violet. The color is rather dim to the eye but the filter is almost entirely transparent to UV light. BLB tubes emit very little radiation above the deep violet at 405nm, and virtually none above 435nm as all longer wavelengths are highly blocked. BLB tubes have become something of Sylvania BLB - click to enlargea novelty tube for parties and decoration and are now available at a number of home supply stores, including Lowes and Home Depot.

Since the blue filter blocks some useful exposing light BLB tubes should in theory print slightly slower than BL tubes. My own testing with several different brands of tubes indicates that the actual difference in printing speed between BL and BLB tubes is very small. As a general rule, however, experience will show that BL tubes of the same manufacturer will print slightly faster and with just a tad more contrast than BLBs.

Actinic — The Philips Actinic 05 tube emits radiation from 300nm to 460nm, with the peak at 365nm, close to ideal for many of the alternative processes. The printing characteristics of the Actinic tube are quite similar, both in speed and contrast, to the BL tube. This tube appears to be more widely available in Europe than in the US. In this country it is usually a special order item and the cost is relatively high in comparison to other tubes.

Super Actinic — This tube, known as the Super Actinic 03, emits mostly violet and violet-blue light between 380nm to 480nm, peaking at around 420 nm. The Super Actinic tube is available in standard wattage, HO (high output), or VHO (very high output) versions, and in a wide range of sizes. The SA tube prints slower than the other two tubes, even in HO and VHO, with all of the processes I tested, but it gives the highest contrast. Many experienced printers who use the SA tube claim that it is faster than the BL for printing platinum. On the other hand I have received reports from equally experienced printers who have made direct comparisons of the printing speed of the BL and SA tubes and claim the opposite. The differences noted may result from the differences in sensitivity of light of different wavelength to the variations in humidity of dry, sensitized paper that occur across the country. The Super Actinic light is one of the most popular bulbs for aquarium applications and is widely available through dealers in aquarium materials.

AQUA — This is another aquarium tube, made by Voltarc, very similar in spectral output and printing characteristics to the Super Actinic.

Best Choice among Tubes

What, then, is our ideal UV exposing bank of fluorescent tubes? If the decision is based primarily on economic consideration, the least expensive choice is a unit with standard wattage BL or BLB tubes and magnetic ballast. Electronic ballast would add to the cost. The same unit with Actinic, AQUA or Super Actinic tubes would be more expensive. Finally there would be an extra cost for a unit with HO or VHO tubes, due not only to the extra cost of the tubes but also to the fact that heavier ballast is required than with standard wattage tubes.

Based on printing speed the answer to the above question is more problematic. In looking through the archives of the alt-photo-process list and through other published materials one finds very little in the way of data derived from actual comparison testing and much of the information that is available is contradictory. Many variables affect exposure times: distance from the bulbs to the printing stage, temperature and age of the bulbs, wattage, type and power of ballast, peak wavelength output, to say nothing of the variables that are found in post-exposure processing.

Assuming that one will be printing with several different processes, including gum, carbon, cyanotype, and Van Dyke, I would recommend either the BL or BLB tubes over the Super Actinic and AQUA. With all of these processes the regular power BL tubes gave the greatest printing speed in my tests, followed just behind by the BLB, while the HO and VHO Super Actinic were significantly slower. In terms of contrast, the Super Actinic gave the greatest contrast, the BL slightly less, and the BLB the least of the three tubes tested.

However, if one will be printing exclusively with traditional kallitype, Pt/Pd, or Van Dyke the Super Actinic (or AQUA) may be a better choice.

HID Lamps

HID printing units can be adapted from commonly available commercial units. Lamps of 1000 watts put out a lot of light and give very fast printing times, from 2-4 times as fast as fluorescent tubes. There have also been claims that the prints made with such units are sharper and of greater contrast than those made with fluorescent tubes. In my testing of printing units I found that HID lamps print with more contrast than fluorescent tubes. However, the differences were not great and with most processes there are controls that would allow us to equalize contrast. As for sharpness, when making small prints where it was possible to maintain very good contact between the exposing negative and printing paper during printing I could see no difference in sharpness, regardless of whether the exposure was made in a contact printing frame or with a vacuum easel. On the other hand, when making large prints in a contact-printing frame I did find a small difference in sharpness between prints made with the metal halide light and those made with fluorescent tubes, with the advantage to the metal halide unit. My theory is that contacting printing frames in large size are not capable of maintaining tight contact between the negative and printing paper. This lack of close contact leads to a greater scattering of the light (and loss of sharpness) with a diffuse printing source — where many of the lights rays pass through the negative at very low angles — than with a semi-collimated unit — where the rays pass through the negative at relatively high angles.

My interest in assembling a HID light was sparked by an exchange of messages on one of the alternative printing lists about the printing characteristics of point source and diffuse printing lights in printing with the alternative processes. Based on my understanding of the various aspects of the issues involved I determined that a HID unit of 1000-watt power would likely print faster than my bank of BL tubes, an important consideration given the fact that many of my stained negatives required exposures of 30-60 minutes. Most importantly, however, the HID light, being semi-collimated, would allow printing from reversed negatives without loss of apparent sharpness, something not possible with diffuse light sources such as fluorescent banks.

To set up a HID unit you will need what is known as a Luminaire, a complete lighting unit consisting of a lamp, ballast, and reflector. For a standard 1000-watt Luminaire the ballast will be ANSI specification M47 for metal halide lamps, H36 for mercury vapor lamps, with a Mogul E39 base for the lamp. The reflector for this unit will typically be 23" in diameter. The size lamp required for this application is designated BT56, which is 56-eights of an inch in diameter. Large electrical supply houses should have these units in stock, or you can order direct from Grainger.

The unit I chose was a MHSS 1000-MT Lumark Fixture with a 23" round reflector (see Grainger Item No. 7v197 in the 2002 catalog for equivalent), complete with ballast, and a MH1000 w/v/5k Venture Lamp. I ordered both the Lumark fixture and lamp through a large electrical supply store in Greenville, SC. Total cost for the fixture with ballast and lamp from the local supplier was around $230.

Metal Halide and Mercury Vapor lamps are available with Kelvin ratings from 3200-6200K. In general lamps with a higher Kelvin rating will work better for our purposes since they also tend to radiate more energy in the UV and violet range. The 5200K Venture lamp puts out about 85,000 lumens, with a Kelvin rating of 5200K. Although this is a broad spectrum lamp that radiates considerable energy outside of the useful UV and violet zone it proved in comparison printing to be as fast an the Olec L1252, a lamp designed for graphic art printing. Based on an analysis of the SPD chart I suspect that the Mercury Vapor H36GV-1000 would also perform very well for alternative printing, perhaps even better than the Venture 5200K lamp.

The fixture was very easy to assemble and set up. The ballast supply voltage, which is known as a Quad Tap, has several wires coming out of the unit: one green for ground, a common (white), and four black wires, one each for 120 VAC, 208 VAC, 240 VAC and 277 VAC. Connect the green, white, and black wire of the correct voltage to the corresponding wires of an extension cord rated for the amperage for your outlet. WARNING: Don't attempt this if you don't know basic wiring.

The top of the fixture has a 3/4" female pipe thread for attaching or hanging the fixture. You will need to suspend the unit about 20 inches from the print surface or build a box to set it on. The 20 inches is measured from the light pod in the bulb to the plane of the negative to be exposed. I hung the fixture from a 2X4 in my ceiling with a 5/16" X 4" eye screw. Total weight was about 30 lbs., and the reflector has a diameter of 23".

At twenty inches the lamp will provide enough illumination to expose a print up to about 11X14" in size with very little falloff at the corners. If you want to make larger prints you will either have to increase the distance from the lamp to the exposure plane to 30-40", or equalize exposure by suspending a round solid disk of about 6" in diameter about 10" directly under the center of the light. I set up my unit at 20" from the exposing plane with a swinging center filter that allows me to use the light either with or without the filter. When used without the filter (I call this Mode A) printing times are very fast but there will be about a 1/2 stop in light falloff at the corners of a 16X20" print. When used with the filter (Mode B) the printing speed is slower by about 1.5 stops but the light is absolutely even over an area of approximately 23X29". The attached photograph of the actual installation should clarify the operation of the center filter. The center filter should either be painted white or covered with a piece of aluminum foil as this will increase light at the exposing plane by about 15%.

The light is very bright and intense but it is not dangerous . However, it should be shielded from the operator during use and you should not look directly at it.

Once you screw in the bulb and turn on the power the lamp will need about 3-5 minutes of warm-up time to reach full brightness. And, after shut down you will have to wait a few minutes for the lamp to cool before restarting. However, don't worry if you accidentally turn it back on without allowing it to cool and nothing happens. The lamp is still ok and will come back on in a few minutes when it has had a chance to cool.

Unless you have a light integrator the best way to use the lamp is to leave it on for the duration of the printing session, planning the beginning and end of exposure with some kind of manual timer. The down side of this kind of use is waste of power and heat.

Testing of Printing Sources

Philips BL/BLB - click to enlargeIn preparing data for the article I ran a series of tests with four different kind of exposing lights with the following processes: carbon, cyanotype, gum, traditional kallitype and Van Dyke. The lights were:

  1. URI Super Actinic 24" VHO, 75 watts
  2. 24" Phillips BL (Black Light), 20 watts
  3. 24" Sylvania BL
  4. 24" GE BLB (Black Light Blue), 20 watts
  5. Venture 5200K 1000 watt Metal Halide Lamp

The tests with all of the fluorescent tubes were made in the same four-tube UV exposing unit, powered by an Icecap 430 electronic ballast unit rated to handle up to four 24" VHO tubes.

Test Conditions

1. The exposing negative was a Stouffer TP 45 step wedge. A step wedge is piece of transparent film that consists of 21 graduated densities, beginning at B+F with Step 1 and increasing by approximately log 0.15 per step to a maximum of log 3.00 or slightly higher. Since a density difference of log 0.30 corresponds to one full stop, each of the steps of the step wedge represents a 1/2-stop difference. My personal preference is the 4X5" Stouffer TP 4X5 guide for both film and paper testing.

2. Consistent work patterns were used. All tests of same process were on paper sensitized and developed together. All tests were repeated at least twice to verify consistency. The same paper, Arches Aquarelle, was used for all tests with cyanotype, traditional kallitype and Van Dyke.

3. All tests with the fluorescent tubes were made in a four-tube bank powered by an Icecap 430 electronic designed for VHO tubes. The tubes were allowed to warm up for two minutes before beginning the exposure and exposures were made with the tubes at exactly 4" from the exposing plane. It is important to note that the use of the Icecap 430 ballast resulted in printing speeds of about one full stop faster than when the same tubes were tested in units powered with magnetic ballast designed for standard wattage tubes.

4. All tests with the metal halide lamp were with the printing frame at 20" from the light, with the use of a center filter that evens out the light over a 23X29" area. The lamp was allowed to warm up for 5 minutes before beginning the exposure.

5. All tests were carried out with room conditions at about 70F and 55% RH.

Visible Characteristics of the Lights

There was a huge difference in visible light between the Super Actinic, BL and BLB tubes. The Super Actinic puts out a very bright, bluish/violet visible light, the BL a bluish light, less intense than the VHO Super Actinic, and the BLB a very pale blue light. The 1000-watt metal halide lamp emits an intense light, bluish/white in appearance.

Analysis of Results

The step wedge prints were analyzed to determine speed and contrast of each of the combinations of light and process. Speed was determined by the first step wedge that provided maximum density, contrast by the total number of steps between maximum density and paper white.

Exposure and Contrast Data from Tests

Van Dyke (5-minute exposure)
Light SourceDmaxDminScaleExposure Scale
Sylvania BL42017 steps2.55
GE BLB41815 steps2.25
URI SA31816 steps2.40
Metal Halide31715 steps2.25


Cyanotype (20-minute exposure)
Light SourceDmaxDminScaleExposure Scale
Sylvania BL41512 steps1.80
GE BLB41411 steps1.65
URI SA4129 steps1.35
Metal Halide4129 steps1.35


Carbon (5-minute exposure)
Light SourceDmaxDminScaleExposure Scale
Sylvania BL51511 steps1.65
GE BLB51511 steps1.65
URI SA5139 steps1.35
Metal Halide5139 steps1.35


Gum (5-minute exposure)
Light SourceDmaxDminScaleExposure Scale
Sylvania BL5128 steps1.20
GE BLB5128 steps1.20
URI SA4118 steps1.20
Metal Halide4107 steps1.0


Kallitype (10-minSodium acetate developer, 75g per 1000ml)
Light SourceDmaxDminScaleExposure Scale
Sylvania BL41613 steps1.95
GE BLB41613 steps1.95
URI SA41916 steps2.40
Metal Halide51612 steps1.80
  • Dmax — First maximum density of the step wedge.
  • Dmin — First density on the step wedge above paper white.
  • Scale — Number of visible steps from Dmax to Dmin.
  • Exposure Scale — Effective exposure scale of the printing medium.
  • Metal Halide — As noted earlier, the tests with fluorescent tubes were made in a four-tube bank powered by an Icecap 430 VHO ballast unit that gave faster printing by about one full stop than when the same tubes were used in fixtures powered by magnetic ballast rated for tubes of standard wattage. When this factor is taken into account the metal halide unit is faster by about one full stop than typical UV banks.

Platinum Tests

Dick Arentz and I collaborated in making tests on Platinum/Palladium similar to the original ones I did with gum, vandyke, kallitype, cyanotype and carbon. Dick coated two sheets of 11X14 paper using his Pt/Pd Mixture #7, then cut the paper into 4X5" pieces and overnighted it to me by FedEx on Saturday. I ran the tests the next day, developing in 25% Potassium Oxalate and clearing for 10 minutes in EDTA. All of the tests have at least two steps of maximum black and all cleared well.

Tests included several different 24" fluorescent tubes, all powered with an Icecap 430 ballast. Since this ballast is rated VHO and is electronic its power could not be a limiting factor in light output. The same step wedge was used for all tests, and all exposures were made for ten minutes in the same contact printing frame placed at 4" from the light source. The tubes reported in this article are:

  • URI Super Actinic VHO tube, 75 watts
  • GE BLB, 20 watts
  • Philips BL, 20 watts

I also tested my HID Mercury Vapor lamp, at 20" from the pod of the lamp to the printing frame, with a center filter. Exposures were also for 10 minutes, after the lamp reached full output.

Results are as follows:

Platinum/Palladium (10-minute exposure)
Light SourceDmaxDminScaleExposure Scale
Philips BL71812 steps1.80
GE BLB61813 steps1.95
URI SA71711 steps1.65
HID-MV51511 steps1.65

General Remarks about Results.

1. By inspection and simple analysis of the step wedge prints it becomes quickly apparent that in general there is a correlation between image contrast and the wavelength at which the specific light source emits the greater percentage of its radiation, though there are some anomalies in the test data. Light sources that emit most of their radiation at long wavelengths of 400nm and above give images of higher contrast than sources that emit a greater percentage of their energy at wavelengths below 400nm. If you plot a characteristic curve of the step wedges you will find quite simply that long wavelength light gives a greater gamma (or CI) than light of short wavelength. For example, the tonal scale of cyanotype varies from 9-11 steps with different light sources.

The apparent reason for this phenomenon is that long wavelength light penetrates deeper into the emulsion, which causes high contrast, while light of short wavelength is diffused and scattered on the surface. I had long suspected this to be true for carbon printing based on observations from printing with different light sources and was not at all surprised that actual testing with a step wedge verified the premise. However, the concept also appears to apply to cyanotype, Van Dyke and traditional kallitype, though there may be other explanations.

2. Relative speed of the processes.

a) Cyanaotype is the slowest process as tested.
b) Traditional kallitype is about one stop faster than cyanotype.
c) Carbon, gum and Van Dyke are the fastest processes tested, about two full stops faster than cyanotype.

3. Relative printing speed of the light sources.

a) The 5K metal halide unit, used in Mode A, was the fastest by over two full stops of all lights tested, and in most cases it also gave the most contrast. When used in Mode B with the center filter it was faster by about one stop than fluorescent banks powered by ballast designed for standard wattage tubes.

b) The GE BLB, Phillips BL, and Sylvania BL gave virtually identical results. In a typical installation with standard wattage ball they are about one stop slower than the metal halide light used in Mode B, and of slightly lower contrast.

c) The URI VHO Super Actinic was the slowest light tested, except with kallitype, but gave contrast similar to the metal halide lamp. It is interesting to note that with traditional kallitype, which like Pt/Pd uses ferric oxalate as its sensitizer, the SA light proved to be as fast as the other light sources. This fact appears to be in keeping with the experience of many Pd/Pd users who report that the SA prints faster than the BL.

A short time before completing this article Ed Stander of New York and I did some testing of cyanotype, comparing the speed of three different metal halide lamps, his Olec L-1250 and L-1252 and my Venture Lighting 5K commercial unit. The L-1250, which has stroke spikes at 400 and 420nm, and a weaker spike at 365nm, similar to a typical metal halide bulb, was the slowest of the three bulbs tested. The L-1252, which has a much broader spectrum from about 340nm to 420nm, was much faster than the L-1250. Curiously, the Venture Lighting 5200k commercial light printed with about the same speed as the L-1252.



All of the UV light sources used in testing for this article are capable of excellent results with any of the processes. However, there are subtle differences involving both contrast and speed that may make one or another of these lights more attractive for a specific process. For example, considering only the fluorescent tubes there is little doubt in my opinion but that BL or BLB tubes are on the whole a better choice than Super Actinic or AQUA. On the other hand, printers working primarily with traditional kallitype or Pt/Pd may find that the Super Actinic or AQUA tubes give better results for them. The use of metal halide lamps imposes other considerations involving ambient impact.

There are many factors that must be weighed to determine what kind of exposing unit is best for an individual situation because in the end this is a personal choice that relates as much to work flow and environmental impact as to absolute efficiency of operation. You should consult widely before making a choice, and especially with printers working with the process(s) you are interested in using, because the best advice will come from those with experience.

Because ultimately the choice of an UV light is a highly personal one I choose merely to provide some of the advantages and disadvantages of a few systems, without ranking. For the purposes of comparison I am going to assume that the reader is interested in making prints no larger than about 16X20" in size, well within the limits of a typical 1000-watt metal halide light or a bank of 8-12 fluorescent tubes.

1) HID Light

The major advantages of this type of light source are, 1) ease of assembly — it can be purchased essentially ready to use, 2) printing speed — it prints faster in most typical conditions than a bank of tubes, 3) low cost, and 4) quality of image — in some conditions the semi-collimated light unit will provide greater apparent sharpness than the diffuse light of a bank of tubes. The disadvantages of this type of unit are, 1) must be left on for several minutes to reach maximum lamp output, and after the lamp is turned off it is necessary to wait for several minutes before it can be turned back on, 2) puts out quite a lot of heat, and 3) there is considerable waste of energy as much of the light radiated is in the green and red area and of no use to alternative processes.

My own HID unit, described earlier, cost less than $250 to purchase and took less than an hour to wire and hang. The printing speed of the unit varies according to the distance from the pod of the lamp to the exposing plane. However, even when the unit is set up to cover an area of up to 28" in diameter, more than enough to provide even illumination for prints up to 16X20" in size, it is about one stop faster than my bank of BL tubes.

2) Bank of 8-12 24" BL, BLB, AQUA or Super Actinic tubes with magnetic or electronic ballast rated for standard wattage tubes

The cost of materials for such a unit would be about the same as for a 1000-watt HID unit. The advantages of this type of unit are, 1) low energy consumption — most of the radiated light is useful to alternative processes, 2) produces very little heat, 3) tubes reach maximum output very soon after powered up, and they can be turned back on immediately after shut down. The disadvantages are, 1) requires quite a bit of time to assemble, 2) printing times will in most set-ups be slower than with a 1000-watt HID unit, and 3) some loss of apparent sharpness may occur in certain conditions, especially when making large prints in contact-printing frames where it is difficult to maintain tight contact between the negative and sensitized emulsion.

3) UV bank of 8-12 24" HO or VHO tubes with appropriately rated ballast

The advantage of this type of unit is printing speed. The disadvantage is cost, as HO and VHO tubes are much more expensive than standard wattage tubes, and the ballast rated for HO and VHO tubes is very expensive compared to standard wattage ballast. Also, HO and VHO tubes are available in a very limited number of types and sizes.



KITS and UV Exposing Units






Copyright notice: This article and all its illustrations are copyright 2001 by Sandy King. All rights are reserved. No portion of this article may be reproduced by any means whatsoever without the express permission of the author.