As soon as you start testing film, you quickly find out that using conventional small tank development is very wasteful. I currently use a Patterson tank with a MOD54 insert that allows me to process up to six sheets at a time. Unfortunately, this requires 1 L of chemistry for each of the three baths for the films to be completely covered. If I am to use this, I will end using an entire bottles of developer and fixer on just one film test. With this in mind I started looking for alternative methods for processing single sheets at lower chemistry volumes. Without going in the pros and cons of all available methods here, it suffices to say that I ended up at rotary processing. In contrast to the alternatives I actually had all the required hardware already and I could still use smaller amounts of chemistry. I dug up a motorized roller from my pile of photography equipment and combined it with a Durst development tank. In comparison to the well known Patterson tank this one has no funnel inside and its light proofing is hidden in the cap. This makes it a better candidate for rotary processing, as there is no part where the chemistry can collect. The old Philips roller needed a bit of attention, but after some tender, love and care it works well enough. It only has one speed and does not automatically reverse direction. The single speed does not not seem to be a problem, and the auto-reverse is something I can either mod into it later, or just flip the tank by hand.

As I do not have any special mounts or film holders to hold the film in place in this tank, I have to rely on the so-called “taco-method”, in which the film is inserted into the tank with its back flush against the tank wall. However, the diameter of this tank is too big for this, so that a small gap exists between the film and the tank wall. I think this to be an advantage for it allows chemistry to come underneath the film to wash off (most of) the antihalation layer. I tested this with an old sheet of T-Max 100 and tap water. With 200 ml of water most of the anti-halation layer came off. The rest can be washed off after fixing.

Procedure and (un)even development

To make sure I can rely on the method to give even development across the entire sheet, I took some test photos of a B.I.G. 18% gray card. As the film I will be testing will be Ilford FP4+ I sacrificed three sheets of this stock for this. The developer of choice was Ilford DD-X at 1:4 dilution. At 22 degree Celcius, I reduced the development time to 7 minutes and 20 seconds. Each bath contained 200 ml of solution.

Taking from the tube development procedure suggested by Davis [1], my procedure was as follows:

  1. In the darkroom, put the film at approximately 1 cm from the edge of the tank. Close the lid.
  2. Pour in the developer, shake vigorously for 10 seconds and then put it on the roller.
  3. Invert the tank every 30 seconds to reverse the direction of rotation.
  4. Take off the tank in the last 10 seconds and pour out the developer.
  5. Repeat steps 2 to 4 for the stop bath and the fixer.

As you can see in the image below (Figure 1), this resulted in uneven development of the sheet.

Figure 1: Results of first attempt at rotary processing. Click for larger version.

The Durst tank has a smaller opening than the Patterson tanks and does not allow for a high pouring rate. As I cannot know where the film sits in the tank once I put the lid on, developer may stream over the film when I am filling the tank. This is one of the possible causes for the uneven development. With some help of the great “#believe in film” community on Twitter, I tried a 5 minute prewash with water before pouring in the developer (Thanks Craig!)[2]. I also got rid of the shaking before putting it on the roller. The new procedure was now:

  1. In the darkroom, put the film at approximately 1 cm from the edge of the tank. Close the lid.
  2. Pour in water at room temperature (22 C) and put the film on the roller.
  3. Flip the tank every 30 seconds.
  4. Pour out the water after 5 minutes.
  5. Pour in the developer and then put the tank back on the roller.
  6. Invert the tank every 30 seconds to reverse the direction of rotation.
  7. Take off the tank in the last 10 seconds and pour out the developer.
  8. Repeat steps 5 to 7 for the stop bath and the fixer.

The video below shows the tank rotating on the roller. I flip it every 30 seconds to reverse the direction of rotation.

As you can see in Figure 2 the prewash seems to have rid me off the uneven development. Funny enough this is exactly against Ilford’s recommendations that favour no presoak/prewash in rotation processing, because that would yield uneven development. To not just disregard Ilford’s advice completely, I did a third test in which I put the developer in the tank before putting the film in. As soon as you then rotate the tank onto its side to put it on the already moving roller, the film will be covered by developer.

Figure 2: Results of the second attempt at rotary processing including a 5 minute prewash. Click for larger version.

As a last test I took more inspiration from Davis [1] and added the developer in the tank before putting the film in.

  1. Pour in the mixed and measured quantity of developer. Prevent if from splashing and hitting the side wall. The inside wall needs to be dry.
  2. In the darkroom, put the film at approximately 1 cm from the edge of the tank. Close the lid. Keep the tank upright.
  3. Rotate the tank onto its side and rotate it quickly along its axis so that the entire film surface wets with developer. Then put it on the already spinning roller.
  4. Flip the tank every 30 seconds.
  5. Take off the tank in the last 10 seconds and pour out the developer.
  6. Repeat for the stop bath and the fixer.

The results to the previous method are indistinguishable and the negative shows no sign of uneven development. This method may be even more fault tolerant than the presoak method, as it does not suffer from the pour-in problem that I encountered in the first test.

Figure 3: Results of third attempt at rotary processing. In this attempt the chemistry was already in the tank when the film was inserted. Click for larger version.

Used volume of developer

In conventional small tank development and in stand development the entire film needs to be covered by chemistry the entire developing time. In rotary processing, however, the continuous agitation allows a much smaller quantity of chemistry to be used that only covers part of the film at any given moment. The concentration of the developing agents and other chemicals needed in developing, set a lower limit to the quantity of solution that needs to be used. In the case of Ilford DD-X a minimal amount of 63 ml of stock solution is required to develop 80 sq in of film. A sheet of 4×5″ film will thus require one fourth of that (15.8 ml). Because the recommended dilution for DD-X is 1:4, this means we need a minimum of 78.8 ml of solution from a chemical point of view.

The coverage of the film in rotary processing depends on the amount of fluid and the size of the tank. The Durst tank has a radius of 4.5 cm and a length of 20 cm. Using these numbers we can find the fluid level in the tank and the corresponding volume of fluid. The results are depicted in Figure 4. For the mathematics scroll all the way to the end of the article and read the Appendix.

Figure 4: Fluid level and corresponding liquid volume for the Durst tank.

I find it easiest to put the film in the tank with its long edge aligned with the curvature of the tank wall. From Figure 5 below, we find that this requires a half angle of 1.41 rad (see Appendix for explanation) for fluid to be able to cover the film completely. This corresponds to a fluid level of 3.8 cm or a volume of 506 ml. At the moment I only used 200 ml of liquid that gives me a fluid level of 1.9 cm. At most this can cover 68 % of the film surface. The results above shows that this is more than sufficient.

Figure 5: Half angle as function of the fluid level for the Durst tank. The half angle required to cover a 4×5″ sheet is indicated in with a horizontal line.

During developing the concentration of developing agent slowly changes and with it the rate of development changes. Locally it may exhaust and bring the development to a halt. In stand development this is used to remove the need for agitation completely. In the highlight areas the developer will quickly exhaust and stop development, while the shadows are given plenty of time to catch up. In this case we only need to control the developing time. However, this is typical cause for unwanted effects such as uneven development, bromide drag or unwanted adjacency effects. To mitigate these effects agitation is introduced. By stirring the chemicals either by rolling, shaking or spinning the tank, the chemistry is redistributed and forced to mix and effectively always have a supply of fresh developer available. For semi-stand development this can be one tank inversion at the beginning and one mid-way the entire developing time of perhaps an hour. In conventional small tank developing agitation is much more frequent at 5 seconds per 30 seconds or 5 seconds per 60 seconds. We can define an agitation ratio as the ratio of time in which the chemistry is refreshed to the time the chemistry sits still at the surface. If we assume that the liquid does not move once agitation has stopped, this ratio is in the order of 1/6  or 1/12 for typical small tank development. For rotary processing the liquid volume sets the agitation ratio. The results for the Durst tank are shown in Figure 6 below.

Figure 6: Agitation fraction as function of liquid volume for the Durst tank.

If we compare the agitation ratio for rotary processing and small tank development we can conclude that we would need a very small amount of liquid to have a comparable agitation ratio. As a volume of 35 ml the agitation ratio already exceeds that of small tank development. Note here that the agitation ratio is independent of the rotation rate of the tank.

While the film rotates through the pool of liquid in the bottom of the tank, it is only wet by the bulk fluid for a fraction of the time. The larger part of the time only a thin film of liquid sits on the film surface and chemistry gets the chance to exhaust locally. Although the agitation ratio is independent of the rotation rate, the latter sets the interval at which this film is refreshed. The results for the Durst tank are shown in Figure 7 below.

Figure 7: Refresh interval for the Durst tank as function of fluid volume and rotation rate.

The refresh interval for small tank processing is between 25 and 55 seconds. For the greater majority of the rotary processing systems this is significantly shorter and in the order of a few seconds. If you have the patience and want to count with me, you will find that the tank makes 15 complete revolutions in 32 seconds, or a rotation rate of 0.47 revolutions per second. At this rate and with 200 ml of liquid in the tank, the refresh interval is only 1.6 seconds. In small tank development diffusion can replenish exhausted developer and a longer interval can be sustained, while in rotary processing the thin film op developer does not allow for such efficient diffusion and only holds a small amount of developer to begin with. A shorter interval is therefore also required to keep the development going.

The increased agitation fraction and refresh interval both lead to the same conclusion: agitation is much more frequent and takes up a much greater fraction of the time. This more vigorous agitation causes a higher average developing rate that leads because the chemistry gets almost no opportunity to exhaust. This means that highlights will continue to develop and the shadows get no time to catch up. This results in a higher contrast and typically denser negatives. That is why for rotary processing the developing time is normally reduced by 10 – 30% in comparison to small tank development.


Rotary processing offers a good alternative for processing single sheets and can be done with simple tanks and a motorized roller if temperature control is not a concern. For colour processing you will still need a Jobo-type processor or BTZS tubes that allow rolling in a tempered water bath to keep the processing temperatures constants and above room temperature. To attain even development either a prewash is required or the developer has to be poured into the tank before the film is put in the tank.

With a good developing method for single sheets available, I can now start testing FP4+ with DD-X for my zone system adjustments. Hopefully the results will be online in the coming weeks. Stay tuned, or sign up for notifications!


[1] P. Davis, “Tube Development of Sheet Film,” in Beyond the Zone System, 4th ed., Burlington, MA: Focal Press, 1999, pp. 76–82.

[2] C. Pindell, Response to my question on Twitter, 29 Sept. 2017. Available at:

Appendix: Fluid height in a horizontal tube

To find the fluid height in a cylindrical tube that is placed horizontally, simple geometry will suffice. The required parameters are indicated schematically in Figure A1 below. By calculating the area of the circle segment A_1 and the triangle A_2, we can find the segment A_{1/2}.

Figure A1: Schematic of the tank cross-section.
 From geometry we know that
A_1 = \frac{\theta_{1/2}}{2\pi} \pi R^2 = \frac{\theta_{1/2}}{2}R^2
and A_2 = \frac{1}{2}xy = \frac{1}{2}R \cos \theta_{1/2} R \sin\theta_{1/2} = \frac{1}{2}R^2\cos\theta_{1/2}\sin\theta_{1/2}.
Using the double angle identity \frac{1}{2}\sin\left(2\alpha\right) = \sin\alpha\cos\alpha this can be rewritten to read:
A_2 = \frac{1}{4}R^2\sin\left(2\theta_{1/2}\right).
From these relations then follows that
A_{1/2} = \frac{R^2}{2}\left(\theta_{1/2} - \frac{1}{2}\sin\left(2\theta_{1/2}\right)\right).
To introduce the fluid level h, we then use
x = R - h
and \cos\theta_{1/2} = \frac{x}{R} = 1 - \frac{h}{R},
to write A_{1/2} in its final form:
A_{1/2} = \frac{R^2}{2}\left( \cos^{-1}\left(1 - \frac{h}{R}\right) - \frac{1}{2}\sin\left(2\cos^{-1}\left(1 - \frac{h}{R}\right)\right)\right).
The total volume of liquid is equal to V_\mathrm{liquid} = 2 A_{1/2} L where L is the length of the tank.
Back to Top