As a follow-up to my previous post (#33, above), I ran some film tests over the last few days to evaluate what impact, if any, doing a water pre-soak has on development, and the results really surprised me. This is a pretty long post, so if you prefer not to read the whole thing, the primary takeaway is this: Doing a water pre-soak does inhibit development (i.e., reduces negative density), at least for the combination of film, developer, and process type that I tested (Tmax 100, Adox XT-3, rotary development). I was really surprised at just how significant the effect is. I describe the testing procedure and results in detail below.
To run the test, I exposed six 4x5 sheets of Tmax 100 with a 21-step Stouffer step wedge, developed them using different development routines, and measured their resulting densities with a densitometer. Plotting the results (zone versus visual density) allowed me to visualize how the density ranges compared among the different development routines.
For each exposure, the step wedge was placed on top of the film in a film holder. Following that, a photograph of a flat, uniform surface (e.g., a white wall) was made with the camera and lens under daylight to create a film negative of the step wedge. The step wedge allowed me to record a range of densities in half-stop increments on the film, ranging from what is essentially equivalent to a Zone 0 exposure all the way up to a Zone 10 exposure. I’ve included scans of one of the sheets in Figure 1 below. The non-inverted negative is on the left, while the inverted positive is on the right.
Figure 1: Tmax 100 film exposed through 4x5" Stouffer step wedge. Non-inverted negative on the left; inverted positive on the right. The dashed box indicates where the base plus fog density measurements are taken (a patch of black electrical tape on the step wedge). The Roman numerals correspond to equivalent zones.
I adjusted the brightness and contrast in the scanning software (Flexcolor 4.0.3) so that the tonal difference in each of the steps would be more clear in this composite image. This was done for purely illustrative purposes and had no effect on the test (the evaluative portion of the test was done with a densitometer and the actual negatives).
All six negatives looked similar to the one on the left-hand side of Figure 1 with only minor variations in density. I’ve labelled the steps in the inverted positive with the corresponding equivalent zone numbers as I find this a useful, if potentially inaccurate, way of thinking about tonality. (I’ll note here that I’m glossing over some of the technical details regarding how a step wedge is used that aren’t particularly important for the interpretation of these specific test results; I'm simply trying to record a series of uniform, repeatable densities on the various sheets of film so that I can make some measurements and comparisons with the densitometer.)
I used a Chamonix 45H-1 camera with a Nikkor-M 300mm f/9 lens, shooting at or near f/22 for all exposures. The metering was done with a Sekonic L-558 in spot mode using an ISO of 100, which is box speed for Tmax 100. The photographs were all made during mid-day under sunny conditions. On the first day, I exposed three sheets (Sheets 1-3) within approximately five minutes of each other under virtually identical lighting conditions (I re-metered before each exposure). On the second day, I exposed an additional three sheets (Sheets 4-6) under slightly brighter (+0.3 EV) ambient lighting compared to the previous day. The slightly brighter conditions on Day 2 were accounted for in the exposure settings for Sheets 4-6. Subsequently, all six sheets should have received essentially identical levels of exposure, say, within +/- one-sixth of a stop (i.e., the metering and exposure settings were all read to the nearest one-third stop, so the standard uncertainty in the measurements should be half that).
All sheets were developed in Adox XT-3 (chemically comparable to Kodak Xtol) using a dilution of 1+1 at 20 degrees C. For processing, I used a Jobo CPE-2 with lift, a 2520 tank, and a 2509n reel with the 4x5 retaining panels installed. The tempering bath of my CPE-2 is equipped with a small circulating pump to help with temperature equilibration and I monitor the temperature with an in-calibration digital lab thermometer. All chemical solutions were prepared using distilled water and ensured to be at the development temperature of 20 degrees C before starting development. Although I only developed one sheet at a time, I used Jobo’s recommended minimum volume of 270 mL for all solutions and processing runs.
For the standard development time, I referred to the most recent Tmax 100 data sheet (see page 4
here). For 4x5 rotary development in Xtol at a dilution of 1+1, Kodak recommends a time of 9:45 (mm:ss). In general, the processing regime went as follows:
- Pre-warm or pre-soak, 5:00 (dry or wet, depending on test type);
- Develop in XT-3 1+1, from 9:45 to 11:42 (depending on test type);
- Kodak Indicator Stop Bath, 0:45;
- Tap water rinse, 0:30 (to prevent acid carryover into fixer);
- Kodak Rapid Fixer, 4:00 (Part A only, no hardener used);
- Tap water rinse, 2 x 0:30;
- Kodak Hypo Clearing Agent, 2:00 (used to eliminate magenta stain);
- Tap water rinse, 2 x 0:30 (to rinse the pour-in channel in the Jobo lift);
- Running tap water wash, 5:00 (using a Jobo Cascade film washer);
- Kodak Photo-Flo 200, 0:45;
- Hang to dry, approx. 60:00.
(A quick note about Step 7 – Because I’m using Kodak Rapid Fixer, which is based on ammonium thiosulfate, Hypo Clearing Agent (HCA) isn’t really necessary for the purpose of removing residual fixer. It is quite useful, however, for removing the sensitizing dyes that are commonly present in tabular-grain films. I know from experience that these dyes can, if not removed, have a pretty significant effect on densitometer readings, so that’s why I’ve included an HCA step in the developing regime. This step could be eliminated but would need to be compensated for by using much longer wash times.)
Once a sheet of film was dry, I used an X-Rite 811 densitometer in Status M mode to measure the visual density of the various steps recorded on the sheet. I checked the calibration of the X-Rite with a calibration target before each set of measurements to ensure it was reading accurately. In total, 22 density measurements were made for each sheet of film – one measurement to determine the base plus fog (B+F) density, followed by an additional 21 measurements to cover each step in the wedge. The density measurements plotted in the charts below are all “net” density measurements, where net density equals the density measurement of each step minus the B+F density.
Sheet 1 is my control sheet and was developed for 9:45 with no pre-soak. To ensure temperature consistency, I used a five-minute, dry, pre-warm step with the tank rotating in the Jobo to develop Sheet 1. Sheet 2 was similarly developed for 9:45, but utilized a five-minute pre-soak in distilled water while rotating in the Jobo. All other development steps were kept the same. The density curves for Sheets 1 and 2 are plotted in Figure 2 below.
Figure 2: No pre-soak vs. pre-soak density.
The above result really surprised me. Although I expected
some loss of density in the sheet developed with a pre-soak, I didn’t expect it to be this significant. These results do strongly suggest that, at least for this specific film-developer-processor combination, doing a water pre-soak does inhibit development to some degree.
My next step was to develop a third sheet of similarly exposed film using a five-minute pre-soak plus extended development time in an attempt to compensate for the loss in density seen in Figure 2. Sheet 3 was developed for 10:30, which is equivalent to an additional 7.7% of development time over the standard time of 9:45. The density measurements for Sheet 3 are plotted along with those from Sheets 1 and 2 in Figure 3 below.
Figure 3: Evaluating the effect of extended development time (+7.7%) following a water pre-soak.
As can be seen in Figure 3, 7.7% wasn’t enough additional development time to compensate for the loss of density due to the pre-soak. I had initially guessed that 5-10% additional development might be enough, so this result was another surprise to me. There’s also a curious observation to be made in the shape of the curve for Sheet 3. Specifically, there's a steeping of the density curve in the highlights near Zones 9 and 10. My instinct is that by extending the development time to compensate for the loss in density due to the pre-soak, we end up with over-development in the highlights. This seems consistent with what photographers generally assume will be the effect of increased development (i.e., greater impact on highlight density and a corresponding increase in overall contrast).
Because these were surprising results, I wondered whether my control sheet (Sheet 1) might not be a fluke (e.g., metering error, etc.). To investigate that possibility, I exposed three more sheets under similar conditions the following day. Sheet 4 was developed as per normal (9:45, no pre-soak) to provide a check on the accuracy of my original control sheet. The density measurements for Sheet 4 are plotted with Sheets 1-3 in Figure 4.
Figure 4: Repeatability test on the original control sheet. The density curve for the repeat control test (Sheet 4) closely matches that of the original control test (Sheet 1).
As can be seen above, there’s very little difference in the density curves of the two control sheets. What difference there is can likely be attributed to the +/- one-sixth of a stop uncertainty in the metering and exposure settings. On the whole, I feel Sheet 4 confirmed that my control densities measured in Sheet 1 were accurate and representative of the standard development. Consequently, I’ve chosen to continue using the Sheet 1 density curve for all subsequent comparisons.
Next, I made a second attempt at determining how much additional development is needed to compensate for the loss of density due to the pre-soak. Sheet 5 was developed for 11:42, which is equivalent to an additional 20% of development time over the standard time of 9:45. The density measurements for Sheet 5 are plotted with Sheets 1, 2, and 3 in Figure 5.
Figure 5: Evaluating the effect of extended development time (+20%) following a water pre-soak.
This result compared well with the densities in the original control sheet, particularly in Zones 3-8, which suggests that 20% of additional development time may be a good place to start with this film and developer combination. That said, Sheet 5 exhibits the same rapid increase in highlight density that was previously observed in Sheet 3. This is, again, consistent with the idea that extending the development time has a disproportionately large effect on highlight density. This might make extending the development time problematic -- i.e., you can counteract the effect of the pre-soak on the shadows and midtones by simply extending the development time, but doing so may come at the expense of blown out highlights. That said, it’s debatable to what degree we need to be concerned with density in Zones 9 and 10, since we’re usually trying to engineer an exposure in such a way as to avoid recording overly harsh light.
For my final test, I wondered if the duration of the pre-soak might be having an effect on the development. In developing Sheet 6, I reduced the pre-soak time from 5:00 to 1:00 and developed using the standard time of 9:45. The results are plotted with Sheets 1 and 2 in Figure 6.
Figure 6: Testing the effect of pre-soak duration on density. The shorter pre-soak (Sheet 6) exhibits less inhibition of development compared to the sheet pre-soaked for 5:00 (Sheet 2).
This was another interesting result. It appears that the duration of the pre-soak does correlate with the degree to which the development is inhibited. Going off my previous hypothesis about how pre-soak water in the emulsion dilutes the incoming developer, it makes sense that a longer pre-soak would result in more water uptake in the emulsion and, hence, greater dilution of the incoming developer. I suspect there are lower and upper bounds to this effect, meaning: The inhibition phenomenon probably begins building at the very moment the pre-soak is initiated, and then reaches a maximum level after a few minutes. Presumably, this would be slightly different for every film type, developer, dilution, and agitation scheme.
To summarize:
- Doing a water pre-soak does appear to inhibit development.
- The duration of the water pre-soak is an important factor -- i.e., the longer the pre-soak, the great the inhibition to the development and the lower the density in the developed negatives (up to a certain point).
- The reduction in density arising as a consequence of the pre-soak can be compensated for in the shadows and midtones by extending the development time (e.g., by +20%), but doing say may simultaneously cause the upper highlights (Zones 9-10) to blow out compared to the highlights in the "standard" development that utilizes no pre-soak.
Some closing remarks... Your results may vary. I've tried to conduct these tests in as consistent and precise a manner as possible, but given different conditions and a different operator, who knows? Also, this isn't intended to be a true Zone System-style film speed and development time test; I'm simply trying to record some identical latent images on different sheets of film so we can see how the developed densities compare following different development routines. And most importantly, these tests aren't intended to suggest whether doing a pre-soak is "correct". Many people swear by it, many others don't. My intention here was merely to ask (and attempt to answer) a purely scientific question without regard to issues of aesthetics. Hopefully it was useful to you.