Negative LAD and 16% neutral gray
Posted 13 May 2011 - 02:00 PM
There's something about Laboratory Aim Densities I cannot figure out by myself, I hope someone will be kind enough to explain the obvious.
I'm studying how LADs are used in the film development process. In an educational brochure, Kodak shows negative film stock characteristic curves where the LADs (0.8 red, 1.2 green, 1.6 blue) clearly don't share the same LogH.
I have done the same projection with different negative film stocks, the results are the same: LADs never share the same LogH. Well, how do these densities can represent a 16 % reflectance neutral gray patch if they don't align vertically at the same LogH ? Of course, the printer can be set up so that the negative LADs are printed on the positive film stock as a neutral gray, but then no "real" gray card filmed on the stage will be printed as neutral gray on the positive film stock, colors will be shifted. Won't they ?...
Reference: Kodak's brochure H-61
Posted 14 May 2011 - 12:46 PM
The printer trims are continuously adjusted to print the standard LAD negative to get as close as possible to these aim densities. Everything else (Colormaster analyser etc) is adjusted to show neutral picture at 25-25-25 printing values. For each 0.025 variation in density on the negative, you will need to adjust one printer point up or down. Labs aim to be within one printer point of aims. The major variation in this is the printing lamp.
There will be a punch tape, disk, USB stick or just a file with the printing lights for scene-to-scene corrections for the whole negative to be printed, but the basic setting of the printer starts with the LAD printed at (about) 25-25-25 (adjusted according to the actual LAD densities).
A sensitometer is not needed in this day-to-day control of printers, just a supply of LAD negative and a good densitometer.
Posted 14 May 2011 - 01:22 PM
On one side you have the densitometer which has an incandescent bulb as light source. Some are equipped with LED. Tungsten filament glow lamps emit a lot of infrared, red, and yellow light.
On the other side we deal with a standardized projection light of 5600 Kelvin colour temperature. Blue light is always of the weakest intensity but absorbed by the yellow layer. Red light prevailes again, being absorbed by the cyan layer. That is why cyan (blue-green) is densest. Green light will be absorbed mostly by the magenta colouring, lying middle.
The matter becomes more complicated if a lab has to produce projection positives for carbon-arc light, I am referring to high-intensity arcs. Those are hard lights with ultraviolet, violet, and blue in abundance. The temperatures within the arc plasma attain 10,000 Kelvin. You see, it’s all about the light.
Posted 15 May 2011 - 08:21 AM
I have programmed a Jones diagram which I hope will show better than words what I have understood.
The Jones diagram itself is based on explanations and sensitometric curves from "La Sensitométrie" by Jean-Louis Fournier. Top-left curves are the characteristic curves of a KODAK VISION Color Print Film 2393 along with the three current densities read on the vertical axis. The bottom-right curves are the characteristic curves of a KODAK VISION2 500T Color Negative Film 5218 along with the three current densities read on the vertical axis.
The bottom-left segments represent the printer trimmer settings. The printer trimmer setting values are displayed under the diagram, at the bottom of the screen copy. Adding one to the values slides the segments by 0.025 LogH to the right, subtracting one slides them by 0.025 LogH to the left, changing the shape of the top-right curves, the sensitometric curves of the final image.
Figure 1. If I have got it right, the laboratory always starts by reporting the negative LAD 0.80-1.20-1.60 on the positive LAD 1.09-1.06-1.03 with trimmers set to 25-25-25, independently from the actual negative curves. These settings are represented on the screen copy below. The three LogH corresponding to the three negative LAD densities are of no use here.
Figure 2. At a certain LogH, the green negative density can be kept around 1.20 with a positive density 1.05 almost equal to the positive green LAD 1.06. Obviously, at the same LogH, the blue and red negative densities are different from the negative LAD, no wonder that the densities 1.49-1.05-1.25 read on the positive film are different from the positive LAD 1.09-1.06-1.03 (except for the green).
Figure 3. Now to simplify matters, suppose that the recorded scene contains a 16% gray card, illuminated with tungsten lights at the exact same temperature than the negative film stock balance color (3200 K), and exposed at the same LogH as in figure 2. In order to get this 16% gray printed as close to the positive LAD as possible, so it looks to the audience like a 16% gray under a projection light (5600 K), the laboratory has to change the printer trimmer settings, as on the following screen copy. The positive densities 1.10-1.05-0.99 we obtain are close to the positive LAD 1.09-1.06-1.03.
Now a 16% gray filmed on the stage should be seen by the audience as a 16% gray.
Is this procedure roughly the way laboratories work or is there something I have misunderstood ?
Posted 15 May 2011 - 11:57 AM
So that means that a perfectly exposed gray scale on 7219 for example will print at xx-30-xx where the green value is most important and the red and blue 'fall' in place to get a neutral picture. I consider a camera negative acceptable with green values between 25 and 35. Higher values than this will mean a loss of highlight latitude and lower means a loss of shadow detail and more grain.
Besides keeping the printers and analysers calibrated, LAD is also very useful for end-to-end control of the intermediate chain. In the digital DI chain a grey with Cineon Value of 445-445-445 is used to calibrate the grading system and the film recorder and also to check the recorded negative, similar to a film LAD.
Posted 15 May 2011 - 02:02 PM
On average, a value of 30 seems to be a serious leap, close to half a stop, from 25 which happens to be an inferior limit to the green value. Is this partly because most DP would normaly over-expose their 18% gray to time for a less grained / more saturated image, or is there any technical reason that wouldn't show on characteristic curves ?
These days a typical Vision is about 27-30-24 or thereabouts.
I consider a camera negative acceptable with green values between 25 and 35.
Posted 15 May 2011 - 11:48 PM
Posted 16 May 2011 - 03:23 AM
Yes, unfortunately I couldn't get my hands on the Adams' Bible, but I do use "Zone VI Workshop" by Fred Picker who describes the Adams' Zone System quite well I think. Apart from the time development control (mainly used in black and white processing), he explains in details how one has to control the shadows and highlight placement onto the characteristic curves, much more than the mid-range gray itself, in order to obtain a satisfying tone distribution in the final image. As in painting, an artistic approach cannot be summed up by placing a single tone on the curves. But don't get me wrong, I am not trying to obtain actual images. I'm involved in the development of an educational software, and so my purpose is merely to represent a simplified workflow in which a laboratory, getting a piece of negative film exposed following strictly lightmeters recommendations, calibrates a printer with the LAD, then tweaks the trimmer values to obtain a "correctly exposed" positive film. By "correctly exposed", I mean a 16% gray exposed on the negative (wherever it's been chosen to be placed on the curves) looks like a 16% gray to the audience, no less no more. All this is purely theoretical, there is absolutely no craft involved. It's a starting point to let people (and myself) play with, and eventually depart from to gain more specific knowledge about the real craft.
In addition to the mathematical, sensitometric textbooks you are reading, may I suggest an empirical approach in a book called 'The Negative' by Ansel Adams. DoPs "play' with the exposure of the negative to get a certain look. They will talk about a 'thin, normal or thick' negative, a sensitometrically perfect negative is rarely what they have in mind.
In this aspect, everything that you and Simon Wyss has written is a huge help to me, I'm extremely grateful to both of you. I would just need to be sure that the three diagrams I have showed are roughly representative of what I have just described, a very simplified workflow used by laboratories, lacking any kind of artistic considerations. If this approach complies enough with reality, I can start experimenting with the negative latitude, placement of the zones, consider the impact of color shift, grain, saturation, even processing (pushing, flashing, bleach bypass, ...).
Posted 16 May 2011 - 12:45 PM
And here you trespass the line. While you are stating that all this is purely theoretical lab technicians take care building the base for the more or less artistic work of a production.
By "correctly exposed", I mean a 16% gray exposed on the negative (wherever it's been chosen to be placed on the curves) looks like a 16% gray to the audience, no less no more. All this is purely theoretical, there is absolutely no craft involved. It's a starting point to let people (and myself) play with, and eventually depart from to gain more specific knowledge about the real craft.
I would just need to be sure that the three diagrams I have showed are roughly representative of what I have just described, a very simplified workflow used by laboratories, lacking any kind of artistic considerations. If this approach complies enough with reality, I can start experimenting with the negative latitude, placement of the zones, consider the impact of color shift, grain, saturation, even processing (pushing, flashing, bleach bypass, ...).
Let me cite from the Kodak Master Darkroom Dataguide for Black and White of which I have the 3rd Edition from 1964. The verso of the first page is imprinted with an 18 % Gray Target, actually a raster of 60 dots per inch. A reflected-light-type meter before this target, “will (on the average)”, Kodak writes, “give the correct reading for calculating exposure.” I think there is nothing theoretical about that.
Why an approach, why should it complie “enough” with reality? There you are: you’ll never start experimenting with negative latitude without the thing. What one does with an electronic calculator, that is a computer, is a simulation in total abstractness. You will always stay somewhere else, that is in the electric realm. Film is chemical.
You can’t be farther away from it than by saying “even processing”. Please try to follow me: film manufacturers produce films which are to be processed according to their specifications, that is in aqueous solutions of chemicals after given recipes at given temperatures and some degree of agitation. How should I simulate that?
Apart from the other self-deception with the so-called analyzer (introduced 1957) which is again an Imitation of Life, a computer aided simulation of chemophysical imagery is, to me at least, kind of, how shall I say, preposterous. It’s like trying to bake a cake with microwaves. Why not slides, in the dark, plain photochemical pictures? The real craft.
Posted 16 May 2011 - 03:53 PM
Now I do believe that understanding relationships between curves on a computer program is better than nothing, when you cannot get your hands on the "real thing". I just had a question about the use of LAD that I think is now answered, thank to you and Dirk DeJonghe.
I sincerely hope that a debate about the validity of synthetic imagery is not about to start, and I don't want to argue the case for it. That's not my point.
Posted 17 May 2011 - 11:35 AM
I have nothing against someone’s curiosity but I can’t see a point in here. Would you want to participate in a lab technicians’ round some day? Do you find anything out of control with the raw stocks? I’m just blind for juggling games.
There is the scientific, objective, way as initiated by Hurter and Driffield. Yes.
There is the subjective view. Yes and yes. A calm look at the green green grass of home in a cinema. Comparison between two pictures side by side, one filtered, the other one not. That discussion is my passion. What is shown in the shadows of a black-and-white scene.
Posted 17 May 2011 - 03:49 PM