This article describes a method of measuring relative film density for a cost of about $25, assuming that you already have a light box. It should help those who have no means of measuring film density. Film density is discussed here in the context of film hypersensitisation. My experiences with film hypersensitisation have hitherto been very limited, having been confined to only two types of colour film. Nonetheless this method has been tested quite well, and although inexpensive and somewhat crude, appears to be remarkably effective. After a little practice you should expect to achieve an accuracy within 0.02. Film density is important because it is used as a means of determining the correct level of hypersensitisation. For most types of colour negative film, Lumicon recommend that the density of hypersensitised film should be 0.1 over base - meaning that the transmission density of hypersensitised film should read 0.1 higher than unhypersensitised film of the same type, batch, and processed at the same time.
Limitations
This method has been tested only with two types of colour negative film. It should work well with virtually any type of colour negative material, however it is doubtful that it will work with slide or BW material without some modification (BW material should be hypersensitised to a higher density level, which is outside the range that has been tested with this method).
This method is only suitable for measuring relative density. There are two instances of this:
- The density difference between two strips of unknown density (where the absolute density of neither strip is known).
- The density difference between a known control and an unknown sample (where the absolute density of one strip is known).Without modification, this method is probably unreliable as a means of measuring absolute density .
Alternatives
Kodak Photographic Step Tablet #2. It seems that many start off using this method, which consists of visually comparing the
film density against a strip of stepped density wedges between 0.05 and 3.05. Regrettably I found this method completely ineffective for this application. There are two problems:
- The density difference between each step is 0.15. Since we are looking for a density difference of only 0.1 this means that the step granularity is far too coarse.
- The density comparison is entirely visual and subjective. In some cases I even had difficulty determining which of two adjacent steps my film was closest to. That's a potential error of 0.3, and in practice the accuracy I achieved was hardly better than this.Purchase a densitometer. This is of course the ideal solution, but a new densitometer is a costly item, though it is possible that a used model may be obtained at reasonable cost.
Build a densitometer. Page 118 of AMOACP gives details.
Use a Photolab. It is possible that a nearby professional photolab has a densitometer and would be willing to test odd film strips for you, but this may prove inconvenient on an ongoing basis, even if their generosity holds good after the first few strips! However using a photolab to measure an initial few strips (to be used as controls) may prove very useful with the technique described in this procedure (more on this below).
The Materials
Only two items are required:
Photographic Light Meter. The type of light meter may play an important part here. The temptation might be to opt for an expensive spot meter with digital readout, however such a meter has not been tested and may even be counter productive for this application. The meter that has been tested is at the complete opposite end of the price scale. It is a very cheap Russian made light meter that sells for GBP 18.00 in the U.K. I would expect it to sell for something in the region of $25 in the US. It turns out that several cosmetic variants of this light meter exist, but they are all made in Russia and they all appear to have the same components and the same functionality. The one I use is called a 'Leningrad 8', marketed by a company called Helios, and was purchased from a high street photoshop. It will probably sell under different brand names according to the market but if it's a Russian made light meter selling in this price range then it will probably be the same.
Most importantly, the meter movement is excellent - very smooth, very accurate (absolutely repeatable) and very sensitive to tiny fluctuations in light level. It has a 35mm x 15mm light reading aperture which is ideal for this purpose. It has three measurement ranges: 1 - 5, 5 - 12, 12 - 18. For this application I found it best to use the middle range: 5 - 12. To ensure that typical film density readings would fall in the centre portion of this range (most sensitive part) I cut out a rectangle from a Cokin x4 neutral density filter and glued it over the light aperture. This way, all film density readings fall in the range between 8 and 10.
Light Box. You may already have a light box, but if not I recommend paying something extra for a high quality model that has bright, even, flicker-free illumination from fluorescent light tubes. I use an A4 size model sold for professional use (Kenro model KL 024 - a U.K. company)
The Method
The method is basically very simple. It consists of laying film strips on the light box and positioning the light meter directly on top of the film strips to take light readings. With the light meter described above the light reading may be taken while the light meter is in position on the film strip. Since this method will only measure relative density you are always taking comparative readings between two separate film strips. There are three important points when taking light readings:
1) It is important to take light readings from the same part of the light box.
2) Since the light box is most unlikely to have a regulated power supply, it is important to take comparative readings immediately after one another.
3) You should take three comparative readings per measurement to even out the effects of any voltage fluctuation between readings.Record the light meter readings from the two film strips. How these readings are interpreted depends on two methods of relative measurement:
#1 - Two Film Strips of Unknown Density
This method would be used to determine the density difference between a hypered film strip and an unhypered control.Subtract the lower reading (hypered film) from the higher reading (unhypered film). Multiply the difference by 0.25. The answer is the density difference between the two films. For example, an unhypered control might read 9.2, and a hypered film might read 8.8. The difference of 0.4 multiplied by 0.25 equals 0.1, which is the density difference. This method is only accurate if the density difference is close to 0.1 - the further away from this, the more the accuracy degenerates. Hence, it is easy to determine if a film has been correctly hypered or not, though not the exact degree of any over or under hypering. This is where method #2 may come to the rescue.
#2 - Two Film Strips Where the Density of One Strip is Known
This method would be used to determine the density of an unknown strip by direct comparison with a known strip. The unknown strip may be either hypered or unhypered.
A reading is taken from the unknown strip and successive known strips are read until a match is found. Obviously, once a match is found you then have the density of the unknown strip. Where an exact match cannot be found, then for very small variations, a meter scale difference of 0.05 corresponds approximately to a density difference of 0.01. Note that a value as small as 0.05 cannot be read accurately from the scale (only estimated), so this method loses some accuracy where an exact match cannot be found.
Generally I use method #1 followed by method #2 as a double check. In other words, I first measure the density difference between an unhypered control and a hypered strip according to method #1, then I try to obtain the absolute density of each of these strips according to method #2. Subtracting the results obtained in #2 should give the same result as obtained in #1.
Calibrated Strips
Fundamental to the success of method #2 is obtaining a collection of strips of differing densities that have been measured with a densitometer. As mentioned above, you should be able to find a professional photolab that has a densitometer and is willing to measure a few strips for you. If you already have a working relationship with the lab then they will probably do it free of charge - it only takes a few seconds to measure film with a densitometer. You should aim at having about five hypered controls and about three unhypered controls. You need more hypered controls because these will show more variance according to how they have been hypered, whereas the unhypered controls should only vary according to the quality of the C-41 process machine and how well it has been maintained. Choose a selection of hypered strips that show a range of slightly varying densities with the middle of the range equal to your best estimate of the desired density target. At first this may prove somewhat tricky and you may need to get a second batch measured. Method #1 should help you pin down the correct density target before you have any controls.
Accuracy
I conducted a test in which I measured the density of 9 strips of varying density using this method. I then had these strips measured by a densitometer. The average error I obtained was 0.0178.
Here is a simple histogram of the errors:
x
x
x x x
x x x x
0.00 0.01 0.02 0.03This test was conducted in the early days when I was still experimenting with this technique and I had only a few control strips of widely varying density (there were only two matches between the controls and the strips under test). You will notice that all of the errors are on the positive side of the baseline. Since then I have refined the technique, corrected the baseline, and have far more control strips. Currently I believe that all errors are easily within 0.02. I will have this verified shortly and will update this article accordingly.
Acknowledgement
Thanks to Jerry Lodriguss for his tremendous help with film hypersensitising, density measurement, and for reviewing this article.
|
|
|
|