Seventy-six years of D-76

A class of fine grain developers best represented by the Eastman D-76 developer is discussed, with its chemistry, variants and techniques.

1. Origin of D-76

The original D-76 formula was invented by Capstaff in 1926, described in a booklet published by Kodak announcing a new duplicating film (Eastman Kodak 1927). The formula is also decribed in Dundon et al. (1930).

D-76 was innovative because of a few factors. It was probably the first developer to be formulated based on scientific data. Specifically, combination of 2g/L of metol and 5g/L of hydroquinone was found to maximize synergistic "superadditivity" effect near the pH of 8.5, with the films of the time. D-76 was also innovative because it used borate buffer to maintain the pH of about 8.5, which is considerably lower than a typical developer from pre-1930 era. One practical consequence was that D-76 provided rather consistent development with clean working solution with excellent granularity and overall image quality. This lead to scientific studies of development at conditions similar to D-76. Some of the well known studies are Carlton and Crabtree (1929), Moyse and White (1929).

2. Variants of D-76

Since D-76 was such a good developer, which gained popularity among motion picture industry, some manufacturers adopted D-76 in their formulary. For example, Ilford ID-11 is identical to D-76, both in published form. However, as more data became available to photographic chemists, variants of D-76 appeared with fine tuning for each user's applications. Such modifications can be categorized in a few types.

1. Addition of potassium bromide in a range from 0.3g/L to 1.0g/L.
2. Amount of sodium sulfite reduced to about 70g/L to 80g/L.
3. Adjustment of borate compounds to improve the buffering property.
4. Slight reduction of developing agents, or change of metol/hydroquinone ratio.
5. Replacing or omitting metol, hydroquinone or both.

The motivation behind (1) is quite simple. D-76 gives rather high fog density when used fresh, and this often leads to lower shadow contrast. It was emprically known by many darkroom workers that previously used D-76 gave lower fog and improved shadow contrast, and one classic technique was to mix previously used D-76 and fresh D-76 at a certain ratio. Slight addition of potassium bromide provides the same benefit, but with consistency. Examples are Kodak D-96, AGFA 17, Fujifilm FD-122, Konica SD-20, and Konica SD-28.

Study by W. E. Lee showed that the rate of solution of silver iodobromide (solvent effect) in sodium sulfite solution is maximum at a concentration of about 0.6M or 75g/L, and this maximum became more pronounced as some potassium bromide was present. Some addition of bromide initially somewhat reduced the solvent effect, but when certain amount of bromide is added, solvent effect was actually increased compared to the case without bromide. Therefore, some formulae adopted this sulfite concentration. Example is Kodak D-96 and AGFA 17.

Original D-76 used 2.0g/L of borax as the buffering agent. This provides reasonable buffering effect if the fresh solution is to be used undiluted, but the pH drops slightly due to development and aerial oxidation. On the other hand, a development reaction product of hydroquinone reacts with sulfite, which causes the pH to increase. This is known to increase contrast of the developer as it is stored unused before use. Therefore, the pH and thereby the activity level of D-76 can disadvantageously decrease or increase depending on the factors of development and storage.

Efforts were made to solve this pH fluctuation problem by increasing the buffering capacity of the solution. It can be done by simply increasing the amount of borax, as in Kodak D-96. However, a better approach is seen in Kodak D-76d and Konica SD-28, in which they used 8g/L of borax and 8g/L of boric acid.¹

The design objective of developer formulation is very often to maximize the development action with given amount of chemical, and this is apparent in D-76 formulation. However, if this objective is slightly relaxed, the ratio can be modified to achieve certain image quality. Examples are Kodak D-96 (1.5g/L metol, 1.5g/L hydroquinone) and Fujifilm FD-122 (2.5g/L metol, 2.5g/L hydroquinone). Also, hydroquinone in D-76 is not absolutely necessary, and in some cases it is omitted. Examples are Kodak D-89, which used 3g/L of metol with no hydroquinone. Grant Haist suggested omission of hydroquinone in D-76, and this is recently advocated by Anchell and Troop (1998) as a standard practice. This version is customary called D-76H though it has never been an official name.

In 1940, Ilford chemists Kendal first published photographic use of 1-phenyl-3-pyrazolidone (Phenidone or Phenidone A), which was found to be more potent than metol when used in conjunction with hydroquinone. However, practical use of this developing agent had to wait until 1953, when phenidone was made commercially available through economically viable preparation method. See Mason (1965) for advancement of phenidone during this period. Some of newly introduced Ilford developers were similar to D-76 formulation except phenidone replaced the metol. Only a small fraction of phenidone is needed compared to metol to obtain comparable results. Examples are Ilford ID-68 and Ilford Replenishing Developer (Microphen).

Hydroquinone is a widely used reductant in industries. It is relatively stable as a reductant of the kind, inexpensive and chemically well studied. However, it is not the most desirable developing agent from image quality point of view, for medium-low contrast pictorial photography. Hydroquinone also imposes moderately high environmental damage, and it renders moderate toxicity to microorganisms. Therefore, industry has been motivated to substitute this compound with other reductants. Incidentally, ascorbic acid and its isomers have been known as a developing agent since ca. 1930, but its practical application had been extremely limited, because it is very challenging to formulate a developer that is stable enough for practical use outside research laboratories.

High volume processing industries such as biomedical imaging (electron microscopy, X-ray, etc.) and graphics art processing (killed by digital technilogy before 21st century) were two of the first targeted markets for ascorbate based developers. These applications usually favor high contrast image of hydroquinone, but ascorbic acid tends to provide lower contrast, especially in absence of other contrast-increasing techniques. Therefore, ascorbate developers are often inferior or comparable at best to hydroquinone for those applications. Ascorbic acid has been used in a photographic print developer, AGFA Neutol Plus.

One recent example of application of ascorbate to photographic developers is Eastman Kodak's XTOL developer, where sodium isoascorbate is used with a phenidone derivative Dimezone. Ascorbate based developers are also adopted in motion picture processing industry (Kodak D-96A) although it appears that this formula is no longer recommended by Kodak. Although not much is disclosed about Fujidol E, this developer also contains sodium isoascorbate. Several U.S. and Japanese patents are granted for photographic developers that use ascorbates in lieu of hydroquinone. Many are aimed at high contrast medical/graphic imaging, and many suffer from the problem that the solution is unstable. A separate article on ascorbate developers is found on this site.

3. Two bath versions

In a similar style and coverage of Carlton and Crabtree paper, Crabtree et al. (1933) investigated a range of two-bath developers that are otherwise similar to D-76. They were pursuing an alternative to replenished D-76 that could deliver consistent result. Among a range of formulae they tested, they found that most satisfactory result was obtained from developers of the following type: the first bath contained all the developing agents and sodium sulfite; the second bath contained alkaline agent and some sulfite. This is the type of two-bath developers that are discussed in the recent book by Anchell and Troop (1998).

Crabtree et al. (1933) also tested two other types of developers. One was such that: both baths contained developing agents. Another was that: both baths were of identical composition, analogous to Kodak's recommendation of two-bath fixation procedure. Apparently, the idea of using multiple stages of developing baths was known at that time, and they cited von Joanovich's 1907 papers, Dundon et al.'s 1930 paper, and Knapp's 1932 paper.

The reason Crabtree et al.'s (1933) discussion preferred the class of developers where only first bath contained developing agents was that (1) this type of developers provided almost constant contrast over a wide range of time of development, and (2) the exhaustion of the chemical was very slow, comparable to one-bath D-76 with top-up replenishment technique. They tested this class of developers closely with Eastman motion picture supersensitive panchromatic negative film. The battery of developers they tested included what are called "divided D-76" or "D-23 with alkaline afterbath" in today's literature. These developers yielded contrast ranging from 0.55 to 0.7, depending on the solution and time of processing. The pH of the first bath was in a range from 7.3 to 7.8, and that of the second bath from 8.4 to 12.0. Many developers produced little contrast increase after 2 minutes in the second bath, especially with high pH in the second bath. With more active second bath, slight increase of speed was observed, but they noted that "the increase was not great enough to be noticeable in camera exposures."

Perhaps one of the most frequently mentioned property of two bath developers by some darkroom workers today is that the two bath method gives compensating effect. Crabtree et al. compared one-bath D-76 and a two-bath variant vis-a-vis and noted that "the strips treated in the two-bath developer (formula 13) showed a suggestion of a shoulder at the upper end of the curve while those treated in the D-76 developer were perfectly straight. However, [...], it would require more than eight times the normal exposure before this effect could be detected in a picture negative. For this reason, the slight loss in latitude is of very little practical importance." Regardless of viewing compensating effect as a positive or negative effect, Crabtree et al.'s result concludes that the difference is practically negligible.

One thing to note in interpreting their study is that they used Eastman motion picture supersensitive panchromatic negative film, which is quite different from today's medium speed films. With today's films, two-bath method probably imposes more limitation on control than providing useful consistency, unless very low contrast is desired.

There are a couple of interesting observations in Crabtree et al. (1933). They tried to minimize development that takes place in the first bath by adding a few agents in the first bath. Sodium sulfate reduces swelling of the gelatin support, thus reduces the amount of developer held by the gelatin matrix. However, addition of sulfate limited the solubility of metol, and metol eventually precipitated out as sulfate was increased. Naturally, they tried other route to this goal, and found that glycerin and sucrose (cane sugar) retarded development in the first bath. The mechanism in which the development was retarded is not described.

4. Products

Eastman Kodak D-76. This is said to be a modified version similar to D-76d. Comes in one package dry powder.

Ilford ID-11. This is said to be almost identical to the published form of ID-11. Comes in two packages dry powder.

Phenidone varieties: Ilford Microphen, Ilford Ilfotec DD-X.

Ascorbate varieties: Kodak XTOL, Fujifilm Fujidol E.

5. Formulary

Because there are many published formulae that are identical or almost identical to D-76, similarity is often more emphasized than individual modifications. In this article, I included formulae that made one distinct step away from D-76 but share the same principle to be considered members of D-76 family.

KODAK Developer D-76     D-76 (developer)   D-76R (replenisher)   metol   2.0g   3.0g   hydroquinone   5.0g   7.5g   sodium sulfite (anhydrous)   100g   100g   borax (decahydrate)   2.0g   20.0g   water to make   1 liter   1 liter               target pH (at 25°C , ± 0.05)   8.50 (unpublished)   unknown   specific gravity (at 25°C , ± 0.003)   unknown   1.091

Ilford ID-11 in published form is identical to D-76.

D-76R was originally presented as top up replenisher for D-76, at a rate of 22.2ml to 29.6 ml per 80 sq. in. of film processed. This original replenishing method is found in Kodak publication J-78.

A developer whose formula is identical to D-76R was used especially when rapid processing or push processing was desired. There is also D-76F, whose formula is identical to D-76 except the borax is increased to 20g/L. D-76F takes about a half of the developing time necessary with D-76. It is also said that using 20g/L of sodium metaborate (Kodalk) in place of borax would requier half the developing time of the borax version, and thus 1/4 the developing time of original D-76 processing time. However, these approximate guide are quoted from old literatures. If you desire to use these modifications, testing with your intended material is almost necessary for successful results.

DK-76 is a version of D-76 where the borax is substituted with sodium metaborate or "Kodalk."

D-76c is a low contrast developer intended for metalographic and spectrographic plates. D-76c contains 0.25g/L potassium bromide and 10mg/L of potassium iodide in addition to straight D-76. One reason for the addition of iodide is said to suppress abrasion marks.

Kodak D-76b     D-76b   metol   2.75g   hydroquinone   2.75g   sodium sulfite (anhydrous)   100g   borax (decahydrate)   2.5g   water to make   1 liter   D-76b is a low contrast developer intended for motion picture variable density sound track processing, where softer contrast than D-76 was desired.

Kodak D-103     D-103   metol   2.0g   hydroquinone   5.0g   sodium sulfite (anhydrous)   100g   borax (decahydrate)   1.0g   boric acid (crystalline)   15.0g   potassium bromide   0.125g   water to make   1 liter   D-103 is a developer intended for small scale development of variable density sound negatives. According to Richard Knoppow, this developer was probably meant for making test negatives for evaluation purposes in the sound recording department. See Appendix for details of technical requirements of sound track processing.

Kodak D-76d     D-76d (developer)   "optional replenisher"   metol   2.0g   2.2g)   hydroquinone   5.0g   5.4g   sodium sulfite (anhydrous)   100g   100g   borax (decahydrate)   8.0g   10.0g   boric acid (crystalline)   8.0g   6.7g   water to make   1 liter   1 liter               target pH (at 25°C , ± 0.05)   8.50   unpublished   specific gravity (at 25°C , ± 0.003)   1.078   1.080 (unpublished)   D-76d is a buffered version of D-76 which provides almost identical result when used undiluted with slow to medium speed films. Slight deviation may be observed if used diluted or with higher speed films. Packaged D-76 products are considered to be much closer to this composition than original D-76 of 1927. The "optional replenisher" is a new replenishing method suggested by Kodak, as recommended in their publication O-3. They instruct to make the "optional replenisher" by mixing 5 parts of D-76 and 1 part of D-76R, both packaged versions sold in Kodak label. Based on MSDS information, it is assumed that packaged D-76 is practically identical to D-76d, and D-76R is unchanged from its original form. Based on this assumption, the composition of "optional replenisher" was obtained as shown above. To best of my knowledge, no direct formula for the "optional replenisher" was published in any official Kodak publication. This "optional replenisher" is used at a rate of 70 ml per 80 sq. in. of film. This replenishing method is less economical than original use of D-76R, but it provides excellent image quality with consistency, even in small scale processing. For more details, consult Kodak publication O-3.

Konica SD-28 hydroquinone   5.0g   metol   2.0g   borax   8.0g   boric acid   8.0g   potassium bromide   0.4g   sodium sulfite   100g   water to make   1 liter       target pH (at 25°C , ± 0.05)   8.50 (unpublished)   This is identical to D-76d except it adds 0.4g of potassium bromide. This addition would slow down the development by 10 to 20%, but the fog level and the shadow contrast are improved.

Fujifilm FD-122 hydroquinone   2.5g   metol   2.5g   sodium metaborate   2.0g   potassium bromide   0.5g   sodium sulfite   100g   water to make   1 liter   This developer is quite similar to D-76b, the variable density soundtrack developer. However, FD-122 was intended for sill images of pictorial quality.

Konica SD-20 hydroquinone   3.0g   metol   1.5g   sodium metaborate   2.0g   potassium bromide   0.5g   sodium sulfite   100g   water to make   1 liter

AGFA-Ansco 17     AGFA 17 (developer)   17A (replenisher)   metol   1.5g   2.2g   sodium sulfite (anhydrous)   80g   80g   hydroquinone   3.0g   4.5g   borax (decahydrate)   3.0g   18.0g   potassium bromide   0.5g   ---   water to make   1 liter   1 liter   This is AGFA's D-76 counterpart for still photography, but it is similar to Kodak's D-96, the standard developer in motion picture industry.

Kodak D-96 (Negative) (from Kodak Publication H-24 Module 15)     D-96 (developer)   D-96R (replenisher)   metol   1.5g   2.0g   sodium sulfite (anhydrous)   75g   80g   hydroquinone   1.5g   2.0g   borax (decahydrate)   4.5g   5.0g   potassium bromide   0.4g   ---   water to make   1 liter   1 liter       target pH (at 25°C , ± 0.05)   8.62   8.72   specific gravity (at 25 °C, ± 0.003)   1.069   1.074   Suggested replenishment rate varies. For Eastman Fine Grain Duplicating Panchromatic Negative Film 5234/7234 in 35mm format, the rate of 1,250 ml per 100 feet is specified.

Developer D-96A     D-96A2   metol   1.5g   sodium sulfite (anhydrous)   75g   L-ascorbic acid   2.0g   borax (decahydrate)   4.5g   potassium bromide   0.4g   calgon   1.0g   water to make   1 liter   This developer works at comparable development times to D-96, but at a lower pH. Granularity and sharpness is somewhat enhanced.

This formula does not seem to be widely promoted by Kodak. My speculation is that users had troubles similar to defective development observed with XTOL, and Kodak no longer lists this formula. Discussions on improving stability of ascorbate developer is found in a separate article.

Kodak D-89     D-89   metol   3.0g   sodium sulfite (anhydrous)   100g   borax (decahydrate)   5.0g   pinacryptol green 1:500   5 ml   water to make   1 liter   This is a very simple and effective developer, if pinacryptol green is substituted with 0.5g of potassium bromide.

"D-76H"     "D-76H"   metol   2.5g   sodium sulfite (anhydrous)   100g   borax (decahydrate)   2.0g   water to make   1 liter       target pH (at 25 °C, ± 0.05)   8.53 (unpublished)   This is an unofficial variant of D-76, which omits hydroquinone and increases metol by 0.5g. This is promoted by Anchell and Troop (1998) who credit the idea to Grant Haist. This composition precludes the increased contrast of original D-76 during storage, but it does not prolong the shelf life from original D-76. However, use of this developer does not discharge hydroquinone to waste water. This formula is recommended for one shot use at 1+1 dilution, or repeated development up to 8 counts of 80 sq. in. of film per liter of stock strengh solution through reuse with increased processing time (10 to 15% increase after each roll). Reused solution should be kept for no more than two weeks as decreased activity may be detected.

6. Techniques

D-76 stock solution is a surprizingly good match with conventional films. For example, Plus-X at EI 80 developed in fresh undiluted D-76 for 7 minutes provides excellent image with almost invisible grains even at 13x magnification.

However, D-76 does not pull out excellent sharpness from some of tabular grain films, notably T-MAX 100, and good highlight is often difficult to obtain with T-MAX 400.

Very often, D-76 diluted 1+1 is suggested for one shot use. This is often said to improve sharpness at the expense of slight increase of granularity. However, I think the sharpness increase is very small, and grain increase may not warrant the dilution. It is also more economical to use stock solution, with careful reusing or replenishment.

Use of D-76R or D-76F as a push developer has been propsed in the past, especially with Tri-X film. Such developers are more effective for normal development of tabular grain films in diluted form, particularly if hydroquinone is left out. Also, phenidone developers are found to be effective in increasing the real film speed.

D-76 forms some cloudy substance in the solution as it gets very slowly oxidized. This is mostly metol precipitation promoted by increased sulfate ion in the solution. Sulfate ion arise through slow aerial oxidation of sodium sulfite in alkaline solution. This loss of sulfite may also result in slightly (but visibly) increased granularity.

D-76 may become contrasty as it is used at an elevated temperature. This is because the activity of hydroquinone increases more rapidly than that of metol as the temperature is raised. Therefore, I strongly recommend to process at 20°C, or at least in the range from 18°C to 22°C, but no higher.

7. Practical recommendations

For those who use only packaged developers. Use Kodak D-76 stock strength with D-76R in the modified replenishing method (optional replenisher) recently suggested by Kodak, instead of original instruction for D-76R. This provides the best image quality, consistency and economy out of D-76 with conventional technology films and Delta 400.

Alternatively, XTOL developer stock strenghth, replenished, provides excellent image quality. However, there is a debate as to adapting XTOL in darkroom proecss because of high incidence of failure reported. My recommendation for XTOL users is to use (1) purified water free of calcium, magnesium, iron, copper, manganese, and any other mineral; (2) avoid contact of the solution with any metallic apparatus; (3) avoid direct sunlight from prepared solution.

For those who mix developers from raw chemicals. Many small improvements made to D-76 are optimized for older materials, and they are often for motion picture applications, as seen in examples above. This is my recommendation, which consolidates all the rational updates to D-76, with some fine tuning. Processing time is roughly comparable to D-76 1+1.

DS-8     developer   replenisher   metol   1.5g   1.7g   boric acid (crystalline)   4.0g   3.8g   borax (decahydrate)   4.0g   4.4g   sodium sulfite (anhydrous)   75g   75g   potassium bromide   0.5g   ---   water to make   1 liter   1 liter       target pH (at 25°C, ±0.05)   8.50   8.55   Replenish by bleed method, at 120 ml per 80 sq. in. of film processed.

For XTOL type developer, DS-10 attempted to provide increased robustness to XTOL developer, with excellent fine grain property and sharpness. Formula and detailed information for DS-10 are found in Film developer recommendation.

8. References

Anchell, S. G. and Troop, B. 1998. "The film developing cookbook," Boston:Focal Press.

Carlton, H. C. and Crabtree, J. I. 1929. "Some properties of fine-grain developers for motion picture film," Trans. of the Society of Motion Picture Engineers, XIII, 406ff.

Crabtree, J. I., Parker, H. and Russell, H. D. 1933. "Some properties of two-bath developers for motion picture film," J. of the Society of of Motion Picture Engineers,21, 21--53.

Dundon, M. L., Brown, G. H. and Capstaff, J. G. 1930. J. Soc. Motion Picture and Television Enginners, 14, 389.

Eastman Kodak Company. 1927. "Eastman duplicating film," 2nd ed., Rochester:Eastman Kodak Company.

Eastman Kodak Company. 2002. "Black-and-white tips and techniques for darkroom enthusiasts", Publication O-3, Minor revision in Feb 2002, Rochester:Eastman Kodak Company.

Germain's "Darkroom Formulary and Handbook" published in 1940.

Haist, G. Modern photographic processing, Vol 1 and Vol 2. Wiley.

Henry, R. J. 1988. Controls in black-and-white photography, 2nd ed. Boston: Focal Press.

Mason, L. F. A. 1965. "Role of phenidone in modern processing," J. of Photogr. Sci., 13, 205--209.

Moyse, H. W. and White, D. R. 1929. "Borax developer characteristics" Transactions of the Society of Motion Picture Engineers, XIII, 445ff.

9. Appendix: Motion picture processing

Since much of the original development as well as subsequent research of D-76 type developers was motivated by motion picture industry, it is of interest to note the technical requirements of motion picture processing, with emphasis on the difference from pictorial photography. Richard Knoppow contributed much of the knowledge presented in this section.

The first successful sound motion picture was The Jazz Singer released in 1926, by Warner Brothers. There had been attempts at sound pictures before but none had any box office success.

The Warner Vitaphone system, used for this picture, was a sound-on-disk system using 16" records running at 33-1/3rd RPM. Those familiar with old time broadcast transcriptions will recognize that these were essentially identical.

During the time that this system was being developed several sound-on-film systems were also being researched. Without going into detail the main ones to reach the market were the Fox Film Company's Movietone system and systems by Western Electric (worked out at Bell Laboratories) and RCA. The RCA system, called Photophone, was developed many years before at General Electric's research laboratory.

The sound-on-disk system used by Warners had a number of drawbacks. For one, if the film broke and had to be spliced it was no longer possible to keep the record in synchrony with the picture. Another major problem was the short life of the discs. Warners had decided to make the system as cheap for theaters as possible. As a result the needles for the reproducer were coarsly shaped steel points. Some abrasive was included in the pressing material to grind the needle to shape as the record played. Because the groove velocity becomes larger at the outside of the record making the quality a little better there Vitaphone discs were recorded from the inside out, the reverse of most phonograph records. This was an attempt to compensate for the wear. The records had a life of about one play, but of course were run until they either wouldn't play any more or sounded so bad no one could stand it. In any case, the system did not last long. The last sound-on-disc movie was released in 1931 (the same year the last silent movie was released) but the industry had long before begun to release films with both types of track where disc was used.

There are two types of sound tracks: variable density and variable area, more properly called variable width. Western Electric and Fox-Case Movietone used variable density tracks, RCA used variable area tracks.

It was soon discovered that the contol over development and printing for sound was much more critical than for picture. At first it was thought that the variable area system was independant of sensitometry, the track was either black or clear. In fact, both types of track require very close control if distortion is to be kept low.

One problem with sound tracks is that the release print contrast is pretty much dictated by the picture requirements. It is not quite right for either type of track. One result is the need to develop density negatives to rather low contrast. OTHO, variable area sound tracks require very high contrast. The compromise is to print picture negatives at a gamma of about 2.0 or a little higher. (Picture negative is generally developed to around 0.6.)

While it was first thought that the ideal system for density sound was to have a linear transfer, i.e., an overall gamma of 1.0 on the print, it turned out that sensitometric testing didn't work very well for either system. The bete-noir is image spread due to irradiation in the emulsion. This spreading out of the image causes distortion of a rather unpleasant nature, especially for area tracks. The solution arrived at by both Western Electric and RCA was to use a special test signal, known as a cross-modulation or intermodulation test. The two companies have somewhat different approaches to the test but they both work the same way. A test negative is run though the sound recorder and printed at varous steps of exposure (by controlling lamp voltage). Several tests are made the negatives are developed and sent to the laboratory making the prints. The tests are printed and the prints are played back using an analyser to measure the distortion. At some combination of exposure and development the sound negative will be right to just cancel the opposite going distorion in the print. In the old days this could be done fairly easily but is a problem now because of the very high volume of material going through the processing machines.

In addition to cancelling distortion the recording film must have very fine grain. Grain shows up as background noise. It is also necessary to maximize the resolution of the film and reduce edge effects as much as possible. 35mm film moves at 18 inches per second. Sound tracks generally must record up to about 8kHz, so the resolution must be very good. Edge effects tend to cause distortion and to limit resolution. Films used for sound negatives tended to have thin emulsions but, in the early days, all sorts of tricks were used to try to confine the image near the surface of the film, at least on the negative. One method was to dye the emulsion yellow. Because sound recording film is blue sensitive the dye prevented the image from progressing very far into the emulsion. Another method used for a time was the employment of untra-violet recording lamps. This had only moderate success due to the low level of brightness available from the lamps and other problems.

A number of special developers were devised to meet the ideal conditions for sound negatives of various types. Density negatives had to be recorded at low contrast and area negatives at high contrast. Actually, two kinds of films were used. Density negatives were recorded on special low-contrast film, area negatives were often recorde on release printing film.

The control needed to insure consistent quality was one of the features that made sound unwelcome to the old silent picture cameramen. They were used to being able to have camera negatives processed to their specification and to have the prints developed to make those negatives look best.

The very tight requirements of sound meant that release prints had to be developed all the same so any variation in scene contrast would have to be done by careful exposure and development of the camera original or else the use of a duplicating negative (big loss in quality).

At the time D-76 was anounced (1927) most negative development was done manually on rack-and-tank systems. These consisted in a sort of open barrel shaped rack on which the film was wound. It was dipped into a tank containing the developer and rotated. The system works fairly well for relatively small amounts of film but is far from as controllable as the automatic machines which soon followed the introduction of sound.

The sound track of modern color films are developed by an "applicator" meaning sound and picture are applied different developers. Applicator is a roller device which applies the developer in the form of a viscous paste. The reason for this is that, at least for older types of sound playback heads, silver tracks must be used. The photocells in these machines are very sensitive to infra-red light. Most dyes used in color films are transparent to IR, so dye tracks are very low level and very noisy.

Both Movietone and Vitaphone were victims of the patent monopolies at RCA and AT&T. As soon as it became apparent that sound motion pictures were money makers the two companies insisted on all studios contracting for full licenses to use the recording systems. Since Warners and Fox wanted only vacuum tube and some circuit rights in order to continue using their own equipment, they were forced to adopt either the Western Electric or RCA systems because there was no economic reason not to.

But I am not sure if either would have survived anyway. Certainly sound-on-disc would not have, it was a temporary makeshift to begin with. Fox Movietone used a glow tube, called an "Aeolight" for recording. This used a ceasium arc to provide a variable intensity souce. This is actually a very good method of recording, but the light output of the lamp was too low so recordings had to be made on the toe of the film resulting in rather high noise. I suspect, that if research had continued on the Aeolight this problem would eventually have been solved.

With digital discs, we are back to using double system, sound on disc again. The playback system stores about seven seconds of sound so that a quite large gap in the film can be compensated for. A very cleaver system

I've left out more than I've put in; the history is very extensive. Photogrpahic sound recording is important because it led to a lot of scientific research into emulsions and the nature of the image which have been beneficial in other kinds of photography.

I had the great privilege of knowing Dr. John Frayne, one of the Bell Labs researchers and a pioneer in sound recording. Dr. Frayne also helped develop the Western Electric 3-D cutter head used for stereo recordings. I got a lot of inside stuff from him.

Dr. Frayne died several years ago. He received a lifetime achievment Oscar from the motion picture academy toward the end of his life. I got to hold it for a minute.:-) Beside being a thorobread researcher he was a wonderful and very generous man.

Notes

¹ Some formulae used sodium metaborate, also known as Kodak balanced alkali or Kodalk, in place of borax as the buffering agent. Examples are Kodak DK-76, Fujifilm FD-122, and Konica SD-20. However, sodium metaborate alone is a poorer buffering agent than borax.(back)

² based on Anchell and Troop 1998. Original source unverified(back)