Making Negatives From Slides?
Answer to How to Make Color Negatives from Color Transparencies
The following is my response to a question posed to me on my youtube channel “Timi Hall Film Photography.” It is information not commonly known anymore. I am posting it here for posterity in case anyone wants to delve into color film photography.
Hello Timi, any advice for making inter-negative's from slide to colour negative for printing?
·
This will take a long time to explain... I'll have to work on a suitable reply and answer in a series of postings. Please be patient.
Making a good color internegative from a color transparency is one of the most difficult things to do well in any color photolab. Technicians who were good at this were rare. You see most internegatives made by photolabs were in fact badly balanced and a good color print can never be made from a bad color negative. The problem was that looking at a color negative could not determine whether it was good or not. You simply had to print from it onto good color paper in good color chemistry before finding out just how good the internegative was. The use of a correctly made gray scale and a densitometer eliminated this costly step, but alas not too many knew how to utilize those two items. Nowadays there is no offering of an internegative material. Both Kodak and Fuji has discontinued this film years ago (though I still have some in the freezer). Also given the fact that tungsten based color films may be rare as well, it may be really difficult to make internegatives from daylight balanced color films. And since internegative films were specifically designed to enhance shadow detail using today's conventional films may require "flashing" of the negative to make it work. (Of course, masking the original color transparency is the ideal way to go) Also you should use a pulled process (meaning processing a shorter time than recommended) to process color films that are used as internegative films. This allows for maximum exposure (to get shadow detail) and still not burn out the highlight detail in an image. I'll work on a response and get back to you. It'll be technical and pretty involved.
Having slept on this I have some thoughts to add. I thought about making a video just on this subject but I no longer have a color lab and all my tests and notes from the 1980s are dispersed and I won't be able to gather them all. And I want to give you guys the most complete answer to any question I am asked. In other words I don't want to just tell you what to do or how to proceed. I want the answer to make sense to you so that you can expand on it. Lastly I am not at all sure that the color materials available today are anywhere near to how they were designed back in 1980. I know that in 2014, before I closed my lab, that what I considered a good internegative was still valid. But today with all the changes that have gone on at Kodak I am not as confident that they know what they are doing. Too many good people have left that company (so I am told). So I will try to explain here in this comment area what I think and why I think the way I do and how I tried to solve the problem of balancing color internegatives. It is similar to how I came up with how I finally decided to make black and white negatives, that is, I looked at how photographic paper responds to exposure and then I designed my negatives to be that way... a kind of "back door" engineering.
I have to say off the bat that I don’t recommend balancing color internegatives the way that Kodak taught me. Their way didn’t work too well, and everyone knew it and complained about it. That was why I came up with my own way to do it and eventually put it on a computer and sold it as “P.S. One.” It was the first software program that my brothers and I sold from our new company which we called “Photographic Software.” That was what PS stood for. “One” was our first program and we followed with “Two” and “Three” and then the “QC” and finally the “Analyzer.” But first let me tell you how Kodak recommended you do it (to balance color internegatives). The 21 Silver step tablet that they sold was their gray scale offering for this purpose. It covered a density range of 3.15 and it was used to contact print onto the internegative material. Kodak, in fact, did this themselves and they published “ideal” curves made in this manner. We, as lab technicians, were to contact print the same 21 silver step tablet onto their internegative material and then match their curves. Now the curves are derived from making densitometer readings off of the 21 steps of the processed internegative. You have to make three readings for each step. You read through a Red filter, Green filter and a Blue filter, hence you get 63 readings. You then go onto graph paper and using a red, green and blue ball point pen, you place a dot on the graph paper in the manner of a HD curve. Basically the bottom part of the graph paper indicates the amount of exposure the film received. The darkest part of the 21 silver step tablet blocked most of the light (representing the black of a transparency scene) and so it would produce the minimum amount of processed density in the internegative material. This is indicated on the very left bottom of the graph paper and as more exposure is given it is placed to the right of that point. Now, If the red density measured .21 you would place a red dot on the left side of the graph paper going up .21 in increments. The next plotted "dot" would be the next step of the gray scale and you would proceed to the right (on the bottom) and mark going up the graph paper - using the graph markings to indicate the increments. In this way you would get 21 dots of red, 21 dots of green and 21 dots of blue. You take a French curve and connect these dots to draw your graphs. Obviously Kodak had to provide graph paper so that their graphs would fit our graphs, and they did provide this paper. By placing Kodak’s ideal curves on a light table you can now place your curves on top and compare them. You are not to “turn” the sheets to make them fit. Instead you can slide the sheets left and right and up and down to get a better fit. Depending on how much you slide to the left or right you would then know how much to adjust your filter pack in order to get a better fit. Moving the sheet to the right by ½ stop (.15 points) would basically equate with one step of the 21 silver step tablet. The red curve is adjusted by overall exposure. So in order to move the red curve to the right by ½ stop you’ll need to increase the overall exposure by ½ stop. That means opening up your f stop by ½ stop or by increasing your exposure by 41%. If you want to move your blue curve to the right by ½ stop you need to subtract 15 points of yellow from your color head. Makes sense? It’s a bit of work and it’s well worth it if this produces a good color interneg. You can’t beat a good color neg, believe me. All the dodging and burning in the world done on a print from a bad neg will not produce a better color print than one made from a really good color negative. Problem is, it didn’t work all that well. So after all this work of getting the curves to fit as best they can why didn’t it work? Mainly you couldn’t get your curves to look like Kodak’s. If you did then you could make an interneg from an Ektachrome original and get a very good print. But even when matching Kodak’s curves exactly, that exposure used on a Kodachrome original wouldn’t produce a good neg. And Kodachromes were the bulk of what we got from customers back in the day. What was the difference? When you measure a gray patch on an Ektachrome original you’ll get a reading of Red = .90, Green = .90 and Blue = .90. A great “looking” gray patch on a Kodachrome original (to the human eye) would result in a reading of Red = 1.15, Green = .85 and Blue = .92 (or something like that, like I said I don’t have my data from those days anymore). But what this means is that although we see gray the internegative is receiving an exposure that is basically cyan/blue colorwise. In fact this is also an issue when making color internegatives from Fujichromes as compared to Ektachromes because Fujichrome cannot hold a decent gray scale from white to black. This is a fact that I knew because I ran my own tests again and again over many years. That’s why I always said that Kodak made the best materials. Now their fade characteristics took a long time to improve but that’s another matter.
Okay, here we go… first, you have to make a good color transparency gray scale. In order to do this you need to make a series of 2" x 7" black and white prints that are gray in color. When I did this back in the 1980s I simply made black and white prints onto Ilford fiber based paper. Today you can do this with inkjet paper but the prints must be neutral in color. Since ink jet prints are made with CMYK inks the neutrality of the gray scale will not be as good as a true black and white print. (It takes a bit of cyan, magenta, and yellow to balance the overall color of an inkjet print. Simply using black ink will not do it. Also neutral looking inkjet prints are subject to being off color (not quite gray) when lit by different color balanced lights. Black and white prints don't have this problem. So what you do is make a series of prints from pure white to deep black. Make as many prints as you can then you can mount them side by side going from white to black. Do not make the density of the prints linear in progression. That is.when measuring the reflectance value of these print on a densitometer (that's another thing, you have to have both a reflectance and transmission densitometer) the lightest print should read .06 and the next step should be .09 and then .12 and so on. Do not make them .15 apart as most commercial gray scales are made. Basically you need to make a gray scale that has a large "bump" in it. The raised bump will be compressed down to be more linear once you photograph it and the steps will be more easily discerned. The gray scale should have at least 21 steps and cover a reflectance density (off of a densitometer) from .06 for the white to 2.40 for the black. You can achieve this black on a black and white fiber based print if you selenium tone the prints ... it'll also help in giving your gray scale achival quality. My reflectance gray scale that I made back in 1982 had 22 steps. Once you have this gray scale (steps mounted adjacent to each other going from white to black) then you can photograph the gray scale with the type of transparency film you will be making internegatives from. That is if you will be working from Fujichromes make a Fujichrome gray scale. If you will also be working with Ektachrome film make one on Ektachrome. You will need to test your exposures and probably place CC filters in front of the lens during the exposure to color balance the gray scale to be gray. The "middle" gray should read Red" = .90, Green" = .90, and Blue" = .90. I placed a MacBeth Color Checker above the gray scale for my gray scale transparency (I shot 4” x 5” film). The MacBeth checker has its own gray scale made up of 6 patches. The 4th patch (counting from white to black, white being the first patch) is the middle gray and should read the .90 density. Photographing with 4" x 5" film will allow you to make the individual steps to be large enough (at least 1/8" thick) to be measured on a densitometer. If you are using 35mm film shoot 3 exposures in succession to make up the whole gray scale. Basically shoot 1/3 of the gray scale with each exposure. Cutting the strip of film with 3 exposures together will produce a single gray scale then, and the steps will be large enough to be placed on a densitometer and be read.
Hope this all makes sense. Now obviously you will need to test the film. Buy enough film of the same emulsion number so that you can make corrections with exposure and filter adjustments in order to hit your aims. You also have to use a good color lab that has consistent processing as you will be relying on them to maintain this consistency so when you make a 5 yellow adjustment (for example) the new processed film will reflect this change. When you get the middle gray to be .90 .90 .90 (RGB) then measure the dark areas of the gray scale and also measure the high values. Hopefully the RGB readings will be fairly neutral all across the gray scale. In my experience only Kodak's Ektachrome film had this capability. Fuji and Agfa was never gray all across the scale. Kodachome looked gray but when measured it showed that the densitometer determined that the middle gray patch was very cyan in color. That was a separate challenge. More on that later. When the middle gray was made .90 in density the overall density range of the whole gray scale had a density range of about 3.75. The lightest patch measured about.10 and the blackest patch measured about 3.85. Kodachrome was different. The middle gray of Kodachrome was 1.15 and the black went beyond 4.00. But Kodachrome can no longer be purchased or processed so you won't be making any Kodachrome gray scales. Now assuming you've done all this then you can begin to expose your color negative film through this gray scale. The exposure is made in contact, therefore the size of the gray scale on the internegative is the same as on the original and the measurements made on a densitometer pose no problem (the steps are large enough to fit the aperture of the densitometer). The best way to make internegatives is to use a color enlarger with dichroic filters so that you can adjust the color balance of the exposing light source. Color enlargers use a tungsten light source so use a film made for exposure with tungsten lights. Kodak made VPL film (stands for Vericolor Professional Long exposure). It was made for portrait photography in a studio setting with tungsten lights. I'm not sure if color film (sheet filn) like this is still made, but you will not be able to use daylight film with a color head fitted with tungsten lights. If you have to use daylight film then you'll have to fit the color enlarger with an electronic flash unit and give successive flash exposures to the film. The dichroic filters are still required. Color film has three main layers of emulsion that need to be exposed and processed equally. If they are not properly balanced then you cannot get a decent color print from that negative. It would be the same as if you took daylight film into a kitchen at night and shot some portraits of people. The skin tone would be off as well as the color of their hair, not to mention the background going all red. Trying to balance a C print like this would result in blue/purplish hair when the skin tone approached a correct hue. This is because one layer of the film (the blue sensitive layer) was underexposed due to the lack of any blue light in the kitchen. So using the right film for your enlarger is paramount. You need some kind of enlarging light level meter in order to maintain a consistent light level of the easel. That should be obvious. Then when you settle on an exposure of say, 15 seconds, then as you make negatives that are slightly different in size you can still use 15 seconds by adjusting your aperture to get the same light level.
Now, how do you determine the correct exposure when making an internegative? Let’s say you made a contact exposure with 40 yellow and 40 magenta with a light level of 2.00 at 10 seconds. Assuming you have processed this test in C-41, you will need to measure each step of the gray scale on that test negative with a densitometer. You need to make a red reading, a green reading, and a blue reading. You need to write these down. Assuming you have 22 steps that will result in 66 readings. Those measurements need to be placed on graph paper and plotted. It is helpful to use a red, green, and blue ball point pen. That way you can plot each curve with the appropriate color on separate sheets and by placing them on a light table you can see if the curves "fit" together. Now here's the rub ... If you simply take color paper and expose it to light (balanced for gray) and then process it you would get a perfect gray print. You don't even need anything in the enlarger to do this. Simply adjust the color head and use the correct amount of exposure to get a gray print that measures Red =: .70, Green = .70 and Blue = .70. BTW - that is a perfect middle gray and although the gray scale middle gray may measure .90 on the transparency it is printed .70 on the print. Now, by giving more suceessive exposures you should get darker and darker prints. By giving successive (less) exposures you should get lighter and lighter prints. And you will, but you will find that the prints do not stay gray. In my experience when the middle gray was perfectly gray the high values went red and the shadows went cyan. So you'll have to adjust your densitometer measurements to reflect this inconsistency of the reproduction of the gray scale. When my brother and I were invited to Rochester to give a demonstration of our internegative balancing software I showed how our program placed the 3 film curves on top of each other to indicate a good balance. The scientists there immediately asked me, "How did you determine your "printing densities?" I answered from Kodak. I didn't tell them anymore than that because it was part of our trade secrets in developing the software. What I didn't tell them is what 1 am telling you now. I used Kodak paper to run the test I indicated above. But I then "forced" the lighter and darker prints to stay gray by changing the filtration in the color head. For some steps I added 4 yellow and for some only 2 yellow. In the shadows I subtracted yellow as well as adjusted the magenta filter. AIl of this was recorded and noted. I also ran the test several times to confirm the characteristics of Kodak paper (which by the way was the best on the market... most consistent and very smooth in the corrections required). In other words the corrections made sense. The paper curves were “off” in a convenient direction and correctable in a consistent manner. I noticed that the red paper curve was distinctly higher in contrast than the green curve and the blue curve too was lower in contrast. The other manufacturers of papers were erratic in their corrections and they weren’t consistent when I tried different emulsions. In other words Kodak was the best and I based my correction curves on their product. So basically you will need to adjust the raw readings taken from the densitometer in order to reflect how the color paper will "see" those curves. This is what is referred to as "printing densities." Funny ... I used Kodak's internegative film and Kodak's C print papers and Kodak's chemicals in order to do all this and they asked me how I did it. Who was I? I studied music and was not anything like a scientist. 1 just tried what made sense to me and it worked. I remember doing this by hand back in 1981 or so. I had all these graph papers around me and my oldest brother asked me, "What are you doing?" 1 answered, "I am balancing a darkroom." And he said, "You should do that on a computer." "You can do that?" I said. Shortly after that I bought an Apple 2e and later graduated to an IBM personal computer. When my 2nd brother finally put corrections into the software (recommending a corrected re-exposure) we began to market the program as "P.S. One." Gotta take a break... more to follow.
By plotting the printing curves I was able to see that the red curve was steeper than the green and blue curves. I figured out that I should lower my internegative red curve in order to make the red printing curve fit better. Also I made a series of gray scale prints where I intentionally went “off” from normal gray and I could visually see that when the upper portion of the curve was blue it could pass for “white.” Once this area of the gray scale went yellow, though, it would not pass for white, and so this was to be avoided at all costs. This was especially important when trying to balance internegatives from Kodachrome originals. So, I surmised that I should intentionally raise the contrast of the red curve (on the interneg). Also, at the same time, I should intentionally lower the contrast of the blue curve. The green curve I left alone. How should I do this? The simplest way was to multiply all the red readings by a factor of 1.15. This would make the red curve look steeper (thereby forcing me to lower the red curve in order to make it “fit” the other curves. I also multiplied all the blue readings by a factor of .90, thereby making the blue curve look lower in contrast than it really is. In this case it would force me to increase the blue contrast to make it fit the other 2 curves. Does this make sense? I also made a series of balanced internegatives that had different density ranges. Basically the density range was the darkest step (producing the lightest) minus the lightest step (that produced the blackest step on the gray scale). Along with the series of balanced internegs I also made sample internegs of different slides that had a variety of images. I then made prints of everything and looked at them with a discerning eye. I happen not to be color blind and so I was able to do this myself, although I did get my brother’s opinion on the matter. He, Pat, was also an excellent darkroom technician (who later became my software programmer). We both came up with an ideal density range of 1.05 when measuring only the gray scale. Now keep in mind that a true highlight and true deep black (that would exist in a typical transparency) does not exist on the gray scale itself. Highlights are specular and are too narrow and small to be measured on a densitometer. In my video I physically cut a hole in the transparency to simulate a highlight that can be measured. I also adhere black tape to a dark area of the gray scale to simulate deep black. I didn’t think of doing this back in 1984 when I was designing “P.S. One.” Another thing we did was we took an average gamma reading of the straight line portion of the curve and we came up with an ideal of .450. So our software would look at both calculations and make a recommendation by accommodating the two ideals. I don’t recall how I told Pat to program this, but suffice to say that if the density range went too high in order to achieve an average gamma of .450 then the software would accept a lower gamma in order to achieve a density range of 1.12 (or something like that). It been 34 years since I’ve come up with the parameters for our recommendations in the software and my brain is not what it used to be. By the way the average gamma is calculated by taking the amount of exposure (1/2 of a stop… .15) and dividing the amount of density range that amount of exposure achieved. So for example if the plotted density achieved a range of .07 you would divide that by .15 and you’d get a gamma of .47. There were a series of these (I think 5 or 4 adjacent steps). We added them together and divided the total by the number of steps in order to get the average. By achieving the ideal red curve that had a density range of 1.05 and an average gamma of .450 then it was a simple matter of matching the green and blue curves to the red. In this way one can “see” the how the gray scale would print without having to actually print it. It also would predict whether or not the shadow would remain neutral or have a color cast (green shadows were to be avoided at all costs). Our software customers could return interneg film to Kodak when they didn’t get a good “fit.” Our software actually would state something to that fact and would recommend trying another emulsion. In order to gain more contrast for the green curve (to match the red) you would subtract magenta filtration in the enlarger to make a new interneg. Likewise if you wanted less contrast you would add magenta. To increase the blue curve contrast you would then subtract yellow filtration. One important note: Although I recommend an overall density range of 1.65 for black and white negatives color negatives should not have that high a density range. The normal contrast that color paper has is at least a grade 3 and not anywhere near the softness that black and white paper has. Telling you all this really brings me back in time…
As you can see your question has kept me pretty busy the last 2 days. It also brought back a lot of memories that I had forgotten about. I think this is the last installment. BTW - Thanks for the question. The curves drawn from the data will not give us accurate gamma calculations nor will it result in a smooth characteristic curve. This is because the transparency gray scale (on either Kodachrome or Ektachrome, or etc.) is not linear. We shot our own gray scale, don’t forget, in order to make a transparency that would accurately represent what we would be working from when making color internegs. But that gray scale, as useful as it is, is not linear. So we have to find our new set of domains to be placed at the bottom of the graph paper. This was one of the main concerns on Kodak’s part. They asked me how I came to create smooth curves from transparency gray scales. I didn’t tell them because this was another of our “trade secrets.” I already revealed how I did this in my video lesson on the Zone System. If you watched that you will recall that I marked the bottom of the graph paper in 1/3 of a stop increments. This was because the spot meter that I was using only gave me readouts n 1/3 increments. So if an area showed 10 and 2/3 of a reading then the negative that resulted from an exposure of that area would be marked on the graph paper at that particular position. Here I take a slight departure from the subject at hand… I believe that this is the reason why Ansel did not make his film tests in the way that I did. You simply can’t place card boards and garbage cans that will reflect exactly 1 stop apart from one another. Instead he shot 10 separate sheets of film that were exposed exactly one stop apart from one another, thereby ensuring that he could plot the data exactly .30 apart from one another and get an accurate curve shape. The problem, as I see it, is that that is not how we photograph. We don’t make 10 exposures in order to get a final negative. We shoot one with 10 (or more) reflectance values in it. The way I chose to test film and development (and shared with you) incorporated the fact that flare always occurs during the exposure. Flare is the non-image exposing light that affects all exposed film. It is the light bouncing around between the lens’ elements and inside the camera that basically fogs the film. It highly affects the shadow and black areas of the image because those areas get very little image exposure, but the high values and highlights get a lot more exposure so the flare is almost non-existent (percentage wise) in those areas. What this results in is a lower density range because the flare raises the readings of the dark areas of the image. It will result in a longer time of development of the negative to make up for this lowered density range due to flare. Ansel Adams and Fred Archer came up with their technique of photographing individual zones so that they could accurately make equal stops of different exposures. Zone I, then, receives a fast shutter speed and small F stop in order to place that sheet of film in Zone I. Zone 0 received no exposure at all because it represented no exposure and prints maximum black. BUT in actual practice a “real” exposure of a negative would be near the exposure used to produce Zone V. That would be 4 stops more exposure than what was used to produce Zone I. In other words, an exposure of 1/250 of a second at F 22 is a lot less than 1/125 of a second at F 8, and the 2nd example would let in a lot more light to “bounce” around inside the camera and lens. Now back to the subject…. How do we linearize the data by adjusting the baseline that will produce the curves on our graph? That is the question, and that is what I wouldn’t tell Kodak when they asked me their question. What you need to do is to graph the transparency gray scale. Taking measurements (in this case just the red readings will do) of each step you write these down and proceed to place the “dots” of data on graph paper. You can place the data equal distance apart; the resulting plot will have “bumps” and curves in it because the steps are not linear. It won’t matter. Also the curve will be reversed, that is, the highest numeric data will be plotted on the left and not the right. Next what you do is to take a ruler and mark a line following the straightest portion of your data points. It will be a straight diagonal line. Then you look for each data point and mark a horizontal line from it to that straight line produced by the ruler. Finally you mark a straight line going down to the bottom of the graph paper. Does this make sense? This is where a video would help a lot, because I am describing something that could be easily demonstrated. But in this manner you will get new “domain” locations at the bottom of the graph paper that will indicate where to plot your data coming from that transparency gray scale. Doing this you will get both accurate curves as well as gamma calculations. Note that every transparency gray scale will have different domain locations based on how the steps are aligned from each other. This is why our software had a “Configuration” program hidden from the user and my brother had to configure each individual program with each transparency gray scale that we provided with each program. I also made many series of tests with our software deviating from a good balance to many bad balances. We went too light, too dark, too yellow, too blue, too magenta, too green, too red and too cyan. From this data we produced a table. The software would compare the user’s film balance result with our table and make recommendations on how next to proceed. When they were nearly balanced the software would say, “The balance is close enough to produce work, but test once more to fine tune to the final balance. Or something to that effect. At the end of the next test it should say, “You are in balance.” One note about Kodachrome. Basically Kodachrome was impossible to balance perfectly. When asked about making prints from Kodachrome originals Kodak’s official comment was that Kodachrome was always designed to be a final product to be viewed. It was never intended to be used to make prints from. Nevertheless everyone was shooting Kodachrome to make prints. It was (and may always be) the sharpest and finest grain of all color films. Okay, so what I did was I made the best balance I could with Kodachrome following the procedure outlined above. Then I made intentional deviations surrounding that balance… more yellow, more red, more cyan, more blue, etc. Then I made prints from all those negatives. I found the best deviation was the one where the blue high values were exaggerated. The problem was that the software would not recommend this balance because the blue curve was way too “off” from the red curve. So I instructed my brother to move the domains of the blue curve only (in only one section) so that the software would recommend this balance. It was pretty clever, I think, on my part and it resulted in the best internegatives from Kodachrome originals that I had ever produced. Kodak scientists simply said that my balance was “wrong.”
If everything went right, that is, if you got a “perfectly” balanced internegative, then you could print every step of the gray scale perfectly gray. That means that if you printed the darkest step gray it would measure .70, .70, .70 on a reflectance densitometer. You would also achieve that same reading from the lightest step on the gray scale (by giving more exposure). In fact, this is how the photo industry determined the “slope” of any given film. Consider the gray patches on the MacBeth Color Checker. I actually use one to include in my transparency gray scale. It is placed above the 22 step gray scale (refer to the video lesson above). That color checker has 6 gray patches. It was normal procedure to make a negative of such a color checker onto say Vericolor Professional Short (VPS) film. The negative would show all 6 patches and a machine printer (such as those used in mini-labs) would make 5 prints from that same negative. Making a print where the darkest patch would print a middle gray would be called a Normal minus 2 negative. Basically that is what the industry would call an underexposed negative. Printing the adjacent patch to be gray would constitute a Normal minus 1 negative. It is also underexposed but not as badly underexposed. The 3rd patch is printed gray as well and it is a normal negative… it sets the basic balance for the printer treating that particular film type. Then we come to the lighter patches. Printing them with more exposure would make them gray and we move into the territory of normal plus, overexposure. The idea behind this is that the mini-lab (or professional, they had these printers in wedding and portrait labs as well) operator would select the film type and the printer would analyze the negative to be printed and use either normal, N-1, N-2, N-1 or N+2 as a starting point to treat that particular negative and make a final print. There’s no testing here… just straight print and sold to the customer. It was imperative to select the correct film type because each film type had a different “slope.” In the example above we are dealing with a 5 point slope; some printers only had capability for a 3 point slope. Well, in fact when making an underexposed negative gray the printer had to adjust the color balance for that area otherwise not only would that area be lighter but it would take on a color shift. So in order to print a normal minus 1 neg the printer would have to move the magenta adjustment to plus 5 (or something). Then if the printer came across a normal minus 2 neg it may have to move the magenta adjustment to plus 8 or 9. That was how things were managed with automatic printers. Now here’s the thing. If you made a perfectly (or close to balanced internegative then you could make every patch of that MacBeth Color Checker (the gray scale) a middle gray simply by changing the exposure. No color adjustment necessary. This is what we did at Color3Lab. Our internegative balance did just that. And in this way whenever a print needed dodging or burning there was no color shift in the areas dodged or burned. When I worked for my father’s lab it was quite common to adjust the color head before burning in the highlights otherwise there would be a color shift and the burned in area would appear too obvious. A correctly balanced color negative did not pose this problem. Why don’t they design ALL color negatives to be like this? I don’t know. I will say this. Of all the films tests I did before I closed my color lab, the best film in this regard was the natural color film sold by Kodak. It was called “NC” for natural color. By far it was the best. Going for high color saturation at the expense of a neutral gray scale is very distracting. The idea behind a good photograph, in my mind, is getting rid of distractions so that the viewer can see and feel the idea behind the photograph. Strong colors pull the eye towards elements that may not (usually do not) tell the story. Now it’s pretty clear to me that I am still fairly proud of what my brothers and I accomplished back when we designed and produced “P.S.One.” We did something Kodak couldn’t, and that’s saying something. Keep in mind that Kodak has a solid history of making some of the greatest improvements in the field of photography. It’s also pretty clear that our software was not something easily designed. It may have been fairly easy to write, but the design was not easily thought of. I remember at a Photo Marketing Association meeting where there were at least 700 people in the audience. Someone stood up and said, “How come 3 brothers from San Francisco can come up with a program to balance your film when Kodak, with all their resources, can’t come up with something that works?” I was there and when I heard that I began to shrink down into my seat. You see Kodak offered us $100,000 for the rights to see, offer and own our software. Of course they were not going to use our software as we completed it. They were going to look at what we did (our source code) and re-write it to offer to their customers (there were thousands and thousands of photolabs all over the world back in 1986) for free. We were selling our software for $4,200 (we never raised the price and in total we wound up selling about 100 programs throughout the world over the course of about 10 years.) To their credit, in order to pay for my brother and my expense of traveling to Rochester Kodak purchased 3 of our programs. When we rejected Kodak’s offer they told us they figured it would take them one year’s time and about $100,000 in resources to write their own program. That was how they came up with the $100,000 offer. Two years later they had not released their software. When they finally did release it the software did not work; it didn’t even offer a different balance between Kodachrome and Ektachrome (much less Fujichrome and Agfachrome which our software did). Many labs had been waiting for Kodak’s version of balancing software since it was going to given to them for free. Some labs finally purchased ours after waiting for over 2 years (all the while making bad internegs), and that was why so many in the audience were angry with Kodak. Kodak’s response was that the Hall brothers did a great job and a great service for the photo industry, and they thanked us. I remember my father being angry at us for embarrassing Kodak. I visited many labs throughout the country as well as Canada; I was demonstrating our software. I was told on many occasions that our software was “expensive.” As best as I can remember this is what I said (at least it’s the gist of what I said), “Are you kidding? If you buy our software it is like you are hiring me… forever! And no health plan, no payday, no workman’s comp and no payroll tax. It’s as if you make one payment and I am here at you lab forever. AND your prints will look better and be easier to print!” Last true story. After making 3 test exposures at one lab (I forget the city… maybe in Oregon somewhere) the lab owner told me to stay put. I had already been there for 4 hours. He went away for about an hour (he gave me a sandwich to eat) and then he came back in swearing. He said, “God damn it! I am sending out 100 30” x 40” prints today and I just had my guys make me a new interneg and print. YOUR print has all the detail in the sheer highlights down to the darkest area of the curtains. (It was a print of curtains, I guess) Shit, shit, shit… I have to ship today. Today is the deadline.” And then he cut me a check for the software.
Hello Timi, Incredible information - I've had to print it out to read it clearly. Thank you so much for sharing this. My first and formost thought is the total lack of any interneg material being available. I've made a few tests on Kodak Portra160 with a tungsten to daylight CTO filter under the lens. These made a semi workable neg but certainly not balance. Reading your use of step wedge and densitometer, it's obvious i'm going to have to go through a similar process.... PS One sounds like very useful software! Do you still have the source code?
No kidding! But worth it and of course, incredible gratitude and thanks for taking the time to share. I've read dozens of books, kodak manuals but nothing is a substitute for real person with real experience.
@TaiChiBeMe
Apparently the 100 T Portra Tungsten film is no longer available. I just checked the Kodak site. If you can't find a negative film designed for tungsten light then you'll have to fit an electronic flash head into your mixing chamber of your enlarger to make color internegs. This is what I think. It's not as hard as you might thing to do this. Back in the day many labs fitted their color heads in this manner to make duplicate slides from regular daylight slide film. I knew one photographer who sent his work to a slide stock house and he wanted to send Kodachrome daylight slides for this purpose and he did this. The number of bursts it took coupled with the appropriate F stop was all it took to make very sharp duplicate slides. Kodachrome, by the way, was very stable and faded less than any other color slide film. Of course he used a bellows with a 105 micro lens and in most cases he didn't need to filter the light. In some cases he flashed his slide film before making the final exposure. He just shot a dark (blank) slide through his contraption (whole roll) and then rewound it and started all over again fitting the end of the roll of film back into the same starting slot. My guess is that you will never get a decent balance with any kind of daylight negative film exposed with an enlarger using the tungsten light source. It is simply too critical that the 3 main layers of color film be exposed and developed evenly. Don't forget you are not making an original exposure. The internegative is a "Second" generation exposure so all your ducks have to be in a row. I think the main thing concerning what transparency gray scale to use is to use the type you will be making internegs from. The Provia, if that is the only film available now and the only one you use will work if it is a good gray scale. How does it measure (in terms of neutrality) on a densitometer? You can use either Status M or A; it doesn't really matter. What you need to have is an idea of what the interneg material is "seeing" after light passes through the film. If the Provia E6 IT8 target is overall blue or magenta or whatever... then it will not serve as a good target.
@TaiChiBeMe
Yes, I'm sure my brother still has the source code for PS One, but it was written for a DOS driven computer. Windows was not yet developed by MIcroSoft and Apple was still in its infancy. The only program we wrote in Windows was our QC program which was very popular. I still have an old DOS computer with our old software on it... somewhere in the basement. If you are interested in trying a box of old internegative film email me at my gmail address.