Primary Color Correction: Tonal Correction Tools
This chapter describes the tonal correction tools available in several of the applications and plug-ins for doing color correction. It also describes their respective strengths and weaknesses.
Main Tools for Tonal Corrections
Across the range of products, there are lots of tools. Some of them help to alter tonal range; some are more commonly used to control the “color” of the image, generically meaning that they’d be used to control hue and saturation, though they’d also have some effect on the tonal range as well.
For tonal corrections, almost every application that has color correction abilities has some slider or numerical controls to adjust brightness, contrast, black level (sometimes also called lift, shadows, pedestal, blacks, or setup, as mentioned in Chapter 1), and gamma. Some applications may also include numerous sliders and numerical entry windows for various tweaks to the gamma, including knee, shoulder, softness, and the width or specific range of each of these gamma adjustments.
In addition to the typical sliders controls—which are sometimes controllable by knobs or dials on an external manual interface like those made by Tangent Devices (Figure 2.1), JLCooper, and Avid—some applications also give you tonal control via the manipulation of histograms, which some applications call Levels. Most of these Level controls allow you to adjust the output level, which is a fairly intuitive thing to do, but some also include the ability to adjust the actual input Levels using a histogram, which will work the reverse of the way you would think.
Another very common way in computer applications to adjust tonal ranges is via curves. Curves is a popular tonal correction tool because it offers incredibly precise control and is very intuitive. So you’d figure that it is the tool to use for tonal corrections, but there is a caveat: you can really screw up your images with this tool. Most applications that have curves allow you to place as many as 16 distinct points on the curve to control it. All of these points can do some very funky things to your image, including creating severe posterization or banding. Patrick Palmer, formerly of IRIDAS and now with Adobe, points to these issues as the reason Speedgrade does not have Curves. Curves is a favorite tool of Photoshop users, After Effects users, and those who move back and forth between Photoshop and video applications. Full-time colorists are much less enamored of this tool, though with exposure to it through Apple’s Color and DaVinci Resolve, that may change.
Fig. 2.1 Tangent Devices Elements control panel.
Definition
posterization: Posterization happens when an image breaks down from having continuous tones to having specific regions of tones where each region has a distinct transition to the next (Figure 2.2b). This effect is similar to banding.
banding: Similar to posterization, banding occurs when a continuous tone image breaks down into bands of distinct, individual tones. This effect usually occurs in gradients such as the sky. For example, the image begins as a continuous gradation from light blue to dark blue, then through overcorrection (or a radical change in color space or compression) turns into individually discernable bands of color (Figure 2.3b). Banding is more likely to happen with lower-bitdepth images (8-bit instead of 10-bit images).
Fig. 2.2 (a) Original image (16 mm film transfer to HD courtesy Vanderpool Films). (b) Posterized image.
Fig. 2.3 (a) Original image, DV resolution (Courtesy Randy Riesen). (b) Sky exhibiting banding.
Sliders and Numerical Controls for Tonal Range
Let’s start the exploration of these tonal correction tools with simple sliders and numerical input. There are two test patterns that can help you understand the specifics of what these tools can do as you start to explore.
Load the chip chart image from the DVD tutorial media folder (“grayscale_neutral”) and load a ramp pattern. See if your application has one of these ramps. It will have been created for the specific way that your application treats video black. If not, you can use one of the ramps on the DVD (“ramp from 0-254.psd”).
Fig. 2.4 DSC CamAlign GreyScale Test Pattern, courtesy DSC Labs, Toronto.
Definition
chip chart: A camera setup chart that has several different gray “chips” that range from white to black. The chart that is used throughout this book is DSC Labs CamAlign GrayScale Test Pattern chart (Figure 2.4). It has 11 gray patches or “chips” that have specific reflectance values.
Don’t Use Brightness or Contrast Controls
Of all of the controls available in most applications, the Brightness and Contrast controls are the ones that should be avoided. Why? Because they limit the control you have, compared to other sliders.
Watch what happens on a waveform when you use the brightness slider to adjust the image of the ramp or the chip chart. The entire trace of the waveform moves up and down uniformly. Compare this effect to adjusting the gain. When using gain, there is some movement in the other tonal ranges of the picture, but the largest percentage of adjustment occurs only in the highlights of the picture.
Now grade the chip chart or ramp image using the Contrast control. The Contrast control will compress or expand the entire trace of the waveform equally from each end of the tonal range by raising the blacks by an equal amount as it lowers the whites, or vice versa. Good color correction is all about having control, and Brightness and Contrast rob you of control.
Good color correction is all about having control, and Brightness and Contrast rob you of control.
So what do you do if you want to make the picture brighter or add contrast? Well, “brightness” comes from several things. Bringing the gain up to its highest legal level is a key one. Sometimes, because of some very bright highlights, the picture will still seem too dark once you have raised the gain on those highlights to 100IRE because the middle tones of the picture—containing much of the real information—were not raised much. In that case, the real sense of how bright the picture looks is done in the midtones (more on that in a moment). Sometimes, getting a picture bright enough may even involve increasing the gain so much that some of the brightest parts of the image will clip. That is a judgment call that you need to make in deciding what parts of the image “deserve” to have the most detail. If the highlights are not important, then you can clip them to the point where detail is lost in order to rescue detail from the high midtones and midtones, because clipping out the highest highlights will also raise the levels of the middle highlights and midtones.
Greater contrast usually comes from setting nice rich shadows using the setup control and then getting as much of a range as possible between the darkest portions of the picture and the brightest portions of the picture. The contrast control rarely works for this purpose because if your blacks are muddy but your highlights are almost at their maximum level, for example, then you can only slightly increase contrast before the highlights can’t go any higher without clipping. If you want lots of contrast, you can ignore all of the warnings about clipping either end of the signal and really stretch out the tonal range. You will have better results doing an extreme contrast move with a system that does its color correction computations in a higher color space. DaVinci Resolve works in 32-bit floating point, as do Apple Color and Color Finesse. Avid works at 10-bit. All of them have to convert these color corrections back to the original bit depth of the source footage or the highest output bit depth, which is often 8-bit, 10-bit, or 12-bit.
Knowing all of that, if you find the need to raise the entire tonal range of the picture equally, or to compress or expand the contrast (entire tonal range) of your picture equally, then feel free to use Brightness and Contrast controls. An example of this could be in a nonlinear editor: if you want to reduce the overall contrast of an image quickly so that a title or text “pops” over the background more, then lowering contrast would be an ideal tool.
Although you might do your color correction with only one of the following products, I hope you’ll check out the description of how the corrections are done in each of the products. I can’t give descriptions of all of the products out there, so I’ll show you specific GUI tools from a certain representative sample of products. The way your particular product works may not be included in the list, but the basic operating principles will probably be similar to one of the other products mentioned here. Some applications or plug-ins have very “proprietary” tools that I’ll sometimes mention and show. Other times, the basic tools operate nearly identically across all color correction products. In that case, it’s hardly worth describing the same process over and over, so I’ll select an application that is representative of them all. Interfaces change from release to release and products come and go, so I’ll try to just give an overview here.
As I’ve mentioned, there are a lot of NLEs, standalone color correctors, plug-ins, and compositors out there with color correction tools. Covering them all—and knowing them all thoroughly—is impossible, but I will show you some of the main ways that you can affect your tonal corrections in each of several applications.
Just before I finished writing this edition of the book, Adobe purchased IRIDAS and its color correction application, SpeedGrade. Because of the timing of the acquisition, I won’t be discussing Adobe’s color correction application in this book, but the principles of operating it will be the same as using the tools in many of the other applications included in the following chapters.
Let’s start with the tools in DaVinci Resolve. You should really read about all of the different applications, because certain tips and material may be covered only when I discuss that specific application, even though that information may pertain to more than one application. I also believe that colorists might become more like editors, who need to know multiple applications in order to stay viable and employed, Obviously, that doesn’t apply to everyone, but many colorists will be able to make a good living by diversifying the products with which they make their living, so knowing the capabilities and limitations of each will be a valuable career enhancer.
Definition
bit depth: The number of bits (the smallest data amount, basically on or off binary information) used to describe a color. There’s a little confusion about bit depth numbers. Sometimes what is referred to as 8-bit is the same as 24-bit because they are saying that 8 bits per color channel multiply to 24 bits (8 × 3); 8 bits of color depth gives you 256 shades of gray. Then you multiply those 256 shades of gray times the three color channels (256 × 256 × 256) to show how many colors you can describe in that color space (16.7 million colors). And 10-bit video has 1024 shades of gray instead of 256. Obviously, 10-bit video is going to be preferable for color correction. The bit depth computations here generate numbers that are on the theoretical limits. You are really limited in the actual number of levels of tones and colors by your recording and display devices and the color spaces that they represent.
Definition
NLE: Nonlinear editor, such as Avid and Final Cut Pro.
In DaVinci Resolve, primary color correction is done in the Color screen (Figure 2.5); screens are accessed from the buttons in the middle along the bottom of the UI.
Inside the Color screen interface, at the bottom left corner, is an area of three tabs labeled Primary, 3-Way Color, and RGB Mixer. For tonal range controls, let’s first look at the Luma Lift, Luma Gamma, and Luma Gain sliders. They’re the vertical sliders to the left of the RGB sliders for Lift, Gamma, and Gain. These controls are best accessed from a hardware color correction control surface, like the Avid Artist series (Figure 2.6) or, of course, DaVinci’s own controllers (Figure 2.7), but you can definitely just grab each slider in the middle (the default position) and drag it up or down.
A very interesting feature is the LUM MIX control, which allows you to mix the tonal range corrections made with the Luma sliders back in with the original tonal ranges of the image. This mix allows for some very interesting effects. I haven’t found in other color correction systems, though it could exist or might have been added since I wrote this.
Fig. 2.5 DaVinci Resolve’s Primary controls are in the Color screen, accessed through the buttons at the bottom center of the screen.
Fig. 2.6 The Avid MC Color control surface works with several color correction applications, including DaVinci Resolve.
Fig. 2.7 DaVinci Resolve’s Primary Tab in the Color screen.
The 3-Way Color tab (Figure 2.8) is also part of the Primary controls. This tab will look familiar to almost anyone who’s spent any time in Final Cut Pro, Avid, Color, or any number of other color correction interfaces. Although the wheels can’t help you with your tonal corrections, you can use the dials below the wheels to adjust lift, gamma, and highlights. The 3-Way Color tab also has a LUM MIX control to blend corrections.
Fig. 2.8 Resolve 3-Way Color tab, added in version 8.01. The tonal controls are the dials directly below each color wheel.
Fig. 2.9 Resolve’s Custom Curve controls allow you a more graphical, intuitive representation of your tonal range control.
Finally, in the next control panel to the right from the Primary panel, in the Curves/Custom tabs, are similar color curve controls to those featured in Photoshop, Avid, and Color (Figure 2.9). The three colored curves give you control over the balance of your image, but the gray curve to the left is the Luminance curve, which can be used to adjust tonal values. In addition to the traditional points that can be manipulated on the curves, there is also a small triangle to the right top of each curve that can be used to compress the tonal scale and even invert it to a negative image by dragging vertically on the triangle. The other unique control device is the small bar directly below the curve with a triangle to the right side. These controls are the LUM MIX controls, allowing you to blend the correction back with some of the original image.
For more details on how to use curves, see the sidebar on page 54 “How Curves Work” as well as two other curve-related tutorials and tips.
Color Primary In Room
There are three tools to use for the most intuitive tonal corrections. The first is the colored sliders at the top of the screen to the right of each of the three color wheels. The first, rainbow-colored vertical bar in each group adjusts hue. The middle bars adjust the chroma or saturation level of the image. And the black/white bars to the right of each group adjust the level of the black point, midtone (gamma) distribution or white point in the shadow, and midtone or highlight tonal ranges, respectively (Figure 2.10).
Fig. 2.10 Color’s Primary In Room Apple discontinued Apple Color when it launched Final Cut Pro X in 2011, but many colorists are still using it, and the concepts discussed in this section apply to other color correction applications. This book was originally made possible when Color’s predecessor, FinalTouch, allowed me to bring a professional color correction system around the country to visit many of the colorists featured in this book.
Definition
hue offset wheels: A circular user interface patterned like a color wheel or vectorscope that allows for the control of both hue and saturation. Hue values are indicated around the perimeter of the circle and saturation is indicated by the respective distance from the center of the circle. With some color correctors with manual user interfaces, these hue offset wheels can be controlled by multiple trackballs. Each wheel controls a different tonal range.
gamma: Gamma has several definitions, but the primary one that is used by colorists is to describe the midtones or midrange tones of a picture. Gamma can also refer to the curve or steepness of the transition from black to white. These are similar definitions in a way, because by altering the gamma—or midrange—of a picture, the curve or transition from shadow to highlight is also affected. Directors of Photography usually refer to gamma to mean the response curve from black to white instead of meaning the midtones, specifically.
Fig. 2.11 The Tangent Devices Wave control surface is compatible with Apple Color and several other color correction applications.
Each group—consisting of a circular color wheel (sometimes referred to as hue offset wheels) and three vertical bars—controls one of the three tonal ranges. The far left group is for shadows. The middle group is for midtones or gamma. And the right group is for highlights.
If you have an Avid Artist Color controller, a JLCooper MCS Spectrum controller or Eclipse-CX, or a Tangent Devices Element, Wave (Figure 2.11), CP100, or CP200-BK controller, the dials and trackballs on the controller are linked to the various controls on the GUI. For most controllers, each trackball controls the corresponding Hue Offset wheel on the GUI. The rings around each trackball control the black point, gamma distribution, and white point (from left to right). And the dials at the top of the device control saturation. On all of these devices, the dials, buttons, and knobs are customizable and serve several functions, depending on the way it is customized by the user or by “the factory.”
If you are using Color with just a mouse or a trackball, you can bring the shadows or blacks up or down by dragging the small, horizontal cyan line that is at the bottom of the black/white bar on the left. There is a similar small, horizontal cyan line across the middle of the middle black/white bar. Drag this up or down to control the midtones of your image. To control highlights, drag the cyan line at the top of the righthand black/white bar up or down.
The second way to control tonal corrections is to use the entry windows for master lift, master gamma, and master gain controls along the right side of the screen, near the top. You can type in numbers, but that is hardly intuitive. The best way to control these sliders with a mouse is to use the scroll wheel. Hover the mouse over the numeric lift, gamma, or gain number, click down on the scroll wheel, and drag the mouse left and right to adjust in gross increments, or scroll the wheel itself up and down for fine increments.
The default tab in this location is the Basic tab, but you can also make corrections with the Advanced tab. We’ll get into that a little more in Chapter 4, as the Advanced tab is more for making color corrections than tonal corrections.
Fig. 2.12 Because we’re discussing only tonal corrections in this chapter, tonal corrections would be made using the Luma Curve, which is the bottom right control of the four curves.
Much of the use of these specific tools was covered in the previous chapter when we corrected the overexposed interview footage of Brian.
The third way to control tonal corrections is to use the Curves controls (Figure 2.12), which are located in the middle of the Primary In room.
Although the previous two methods of controlling tone in the Primary In room of Color allow control over three tonal ranges, the Luma Curve allows you to alter the black and white points by moving the bottom and top corner of the curve. But it also allows multiple custom selected points in between those points that can be adjusted in any direction.
For more on the use of curves, see the Curves control section later in this chapter.
The Avid Symphony and Color Finesse HSL Control Tab
There are actually a number of subtabs in this area. You can affect the master HSL levels, which include controls for the master highlights, midtones and shadows, or you can gain more discrete control by adjusting the same controls inside the specific tonal ranges. This means you can control the highlights, midtones, and shadows of the highlights; the highlights, midtones, and shadows of the midtones; and the highlights, midtones, and shadows of the shadows. How are there highlights of shadows? There aren’t, really. But instead of simply giving you three tonal ranges, the extra tabs let you restrict your corrections to nine discrete tonal ranges: the high highs, mid highs, low highs, high mids, mid mids, low mids, high shadows, mid shadows, and low shadows. IRIDAS Speedgrade (now owned by Adobe) has these same types of tonal range controls.
Definition
HSL: Hue, saturation, and luminance. Sometimes referred to as HSB, or hue, saturation, and brightness.
Fig. 2.13 HSL controls in Avid Symphony. For basic adjustments to tonal ranges using sliders or numerical values, Symphony and Color Finesse have several options, depending on how much control you want to exercise. The quick and dirty place to start is the HSL Controls tab of the color correction mode.
If you have Symphony or Color Finesse, you can get a better sense for how all these crazy “shadows of the highlights” and “highlights of the shadows” controls work by putting up the grayscale “chip chart” in the color correction mode and watching your waveform monitor as you adjust each of the controls.
There are a couple of controls on this tab that you should not use, even though they may seem to be the most obvious ones to start with. The Hue and Saturation controls are not really for doing tonal corrections, so I’ll discuss those later. And as I already mentioned at the start of this section, Brightness and Contrast really limit your control over specific tonal ranges, as they affect all areas of the picture equally.
Histograms or Levels
I’m not much of a fan of the Levels controls that are available in many NLEs. But if you are comfortable with viewing the levels in histograms, this could be a powerful way for you to have intuitive control over the same basic gain, midrange, and shadow control that was available in the HSL Controls tab (Figure 2.14).
If you are unfamiliar with the concept of histograms, then you may want to read the sidebar that details how to read them.
Fig. 2.14 The Symphony Levels control allows you to adjust levels while viewing histograms.
A histogram is a very simple graph. Horizontally across the x axis (redundant, I know, but I’m trying to be clear) indicates the image’s tonal range with black to the left and white to the right. The vertical axis, or y axis, shows the number of pixels at each tonal value. Most histograms of a well-exposed image will look somewhat like a bell curve, with some pixels at absolute black and some at absolute white, and then most of the pixels will be in the midrange of the picture. There are some beautiful images that may not look like a bell curve, but if you have a histogram that looks like a bell curve, it should be a pretty well-exposed image.
When analyzing an image on a histogram, though, the bell curve shape is less important than identifying the danger signs of a histogram: sharp peaks or “cliffs” at either end of the histogram. These indicate that clipping is occurring. So you can use this “cliff” effect to help you judge when your correction starts to clip.
Look at the following example, which uses the “Steve dark at the park” clip from the DVD.
In the Master tab, at default, the numbers under each histogram are the same (16, 128, and 235); see Figure 2.15. This indicates that black (16 in NTSC levels for a 7.5IRE setup) on the source is mapped to black on the output side.
Fig. 2.15
Now, on the input side, slowly move the black triangle to the right, releasing when the source black side reads 25 (Figure 2.16).
Fig. 2.16
Now the slope on the output side has a nice clean incline, from the black triangle to the peak, but if you slowly move the input slider up from 25 through 30 and 40, you will see that the output side starts to have a significant spike of pixels above the black triangle.
Fig. 2.17
That’s because all of the pixels to the left of the black triangle on the input side have “piled up” at pure black on the output side (Figure 2.17). The more technical description of what’s happening is that the levels are being remapped; all of the pixels in the source footage between 0 and 40 have now been remapped to 0. And all of the pixels between 40 and 255 have been remapped between 0 and 255. So all of the levels of tonality that are in the slope to the left of the input black triangle have been compressed, or clipped down to black on the output side. This means you have lost all the detail in the darkest pixels, because where there used to be subtle differences between these very dark pixels, now all of them are at 0. Usually this “crushing” of detail is a bad thing, but it can also be useful to create a punchy, crushed black look. As long as you don’t mind losing the detail in the deepest black areas, you’re fine. But if you want to preserve that detail, then stop before the big spike gets too big.
Fig. 2.18
Here is the final position of the Levels (Figure 2.18), including a midtone correction to lift the face out of the shadows. I wouldn’t call this a final correction, but that’s as far as we’ll take it in Levels. Notice that there is a small spike to the left of the output histogram. I chose to let some of the blacks lose detail, but not a lot. I assume that most of those pixels are in the black leg at the bottom of the frame and possibly in my (yes, that’s me—or it was, back in about 2005) black hair.
Another way to gain an understanding of histograms is to look at the grayscale chart (Figure 2.19) as a histogram (Figure 2.20). There is a video version of this file called “grayscale_neutral” on the DVD.
Fig. 2.19
Fig. 2.20
If you look at the Luma tab, you can see the five small double spikes to the right of the big middle spike and the five small double spikes to the left of the middle spike. These five left spikes are the five darker chips. The five spikes to the right are the five brighter chips. The big middle spike is the middle gray chip plus the surrounding gray background. That’s why the middle spike is so much bigger. In this case, the spike does not indicate clipping but simply a large area that is supposed to have the same color to it. Remember that the vertical axis of the graph doesn’t indicate brightness, but the number of pixels at a given brightness level. So, because there is a greater area of middle gray on the chart, the histogram shows many more pixels at this middle tonal range. Having spikes that do not indicate clipping is very rare in most “natural” video images. Usually, there is some kind of gradation or imperfection to the color.
There is also a larger bump to the far left side, indicating the larger patches of black on the sides and in the middle of the chart and possibly even some of the black text on the chart.
Forms and Functions of Histograms
In some programs, the histogram is not limited to displaying the master histogram. For example, in many applications, you can also display histograms of the Composite level (saturation and luminance together), Luma, and individual levels for each of the red, green, and blue channels.
… the danger signs of a histogram: sharp peaks or “cliffs” at either end …
These individual histograms can be useful for spotting problems with the individual channels. As I mentioned in the sidebar, sharp spikes at either end of the histogram indicate clipping (lots of pixels jammed into the same tight tonal range). You can look at the individual color channels’ histograms and see whether a specific color channel is more clipped than another, or if there is an entire tonal range that is weak in a certain color channel. I’ve actually seen footage from cameras with severe technical problems where there was absolutely no information in a particular color channel.
The trick with Symphony’s Levels controls is that there are both input and output Level controls. These are tied together. The input controls seem counterintuitive; as you move the shadow triangle to the left, the level of blacks goes up, and moving it to the right makes black levels go down. The reason for this is that when you move the black level to the right on the input side, you are telling Symphony to map all of the levels to the left of the triangle to a lower level.
However, if you move the same black triangle on the output side to the right, you are telling Symphony that all of the levels of black that are at or to the left of the input black triangle should now be mapped to a higher level.
A way to understand this is to look at the curve in between the two histograms as you move the input, then the output black triangle. On the input side, moving the black triangle to the left or right moves the black point on the curve go left or right. But moving the black curve on the output side moves the black point on the curve up and down.
To best study how this works and get it “under your fingers” is to call up a black-to-white ramp or chip chart and simply slide the faders around, watching the waveform carefully as well as looking at the result on the video monitor.
Curves Tab
Curves in Apple’s Color, Avid, Resolve, and Synthetic Aperture’s Color Finesse are similar to the Curves controls of several other products, like Adobe’s After Effects and Photoshop. They allow pinpoint control over specific tonal ranges. This power of control is a wonderful thing, but like any kind of power, it can be abused.
Definition
under your fingers: This is a term that musicians use to describe practicing something enough times that the movements become second nature. Basically, it means developing muscle memory, in which your body knows what to do without your brain consciously thinking about it.
Fig. 2.21 The Symphony Curves controls allow very precise control over specific tonal ranges.
DaVinci Resolve, Color, Avid, and Color Finesse give you a curve for each color channel and a fourth curve for the Master (or Luma). Looking at the default position of curves makes you wonder why they are called curves—they’re straight!
That’s because at their default position, the curves are simple graphs that indicate that the input level or source level, which is indicated along the horizontal axis of the graph and is mapped perfectly to the output level, which is indicated along the vertical axis of the graph.
Imagine if you were to draw numbers from 1 to 100 along the horizontal axis and then do the same along the vertical axis. As you trace vertically up from the 50 level on the source side, the diagonal line intersects perfectly with the horizontal 50 level on the output side. Similarly, 0 horizontally matches up with 0 vertically and 100 horizontally matches up with 100 vertically, giving you a perfect diagonal graph.
But if you “curve” that graph—by adding or selecting a point along the diagonal line and moving it—you remap the input or source levels to new output levels. For example, to lower the gamma of an image, create and drag a point in the middle of the curve and pull it down a little. Note the numerical values at the bottom of the curve. If you pull the center of the graph—128—down to 120, that will mean that the pixels that were originally at a brightness value of 128 are now mapped lower, to 120. That also means that all of the other values that were between 128 and 0 are slightly compressed into a smaller tonal range between 120 and 0 and all of the values between 128 and 255 have been lowered but into an expanded range between 120 and 255. So each time you place and move a point on the curve, you are not only remapping the tonal values of that point but also compressing and expanding the tonal ranges on either side of the point on the curve.
Fig. 2.22 The Curves tab in Avid Media Composer 6. Avid color codes the curves so that it’s easier to see what will happen when you move a curve. For example, pulling the green curve down or to the right will make the image more magenta.
A valuable thing to understand about using curves is that the steeper the angle in a curve, the greater the contrast of the image in that range. Although few professional colorists use curves, they utilize the concept of expanding tonal range where it is needed and collapsing it where the eye does not need the information as much.
This concept is the basis for a great little tip about how to quickly make images look better quickly using curves.
S Curve Tip
In the master curve, place a point—by clicking on the diagonal line of the curve—about a quarter to a third the way up from the bottom of the curve and another about a quarter or a third the way down from the top. Now drag the top point slightly upwards and the bottom point slightly downwards. This creates a shallow “S” curve. The curve makes the blacks rich and the whites brighter (possibly clipping detail in each of these areas, depending on how much the points are moved) and then spreading the tonal values out over a wider range across the middle of the picture.
Fig. 2.23 This is what the S curve looks like in Color’s Luma curve in the Primary In room.
A valuable thing to understand about using curves is that the steeper the angle in a curve, the greater the contrast of the image in that range.
Fig. 2.24 a) The video image as it was shot. A fairly washed out, low-con image. b) The video image with the simple S curve applied. The blacks are richer, the highlights pop more, and the contrast in the midtones is improved.
Of the colorists who participated in sessions for this book, the biggest proponent of using curves as the primary place for doing color correction is Los Angeles–based Avid Symphony Nitris editor, Terry Curren of Alpha Dogs in Burbank. Although his primary job is as an editor, Terry’s corrections stood up very well against those done by the full-time colorists, and he made his corrections very quickly and efficiently using curves. Before Apple bought FinalTouch from Silicon Color, Terry’s biggest complaint about the product was that it lacked curves, which are integral to his approach to color correction. Under Apple’s first release of Color, curves were added.
We will get into using individual color channel curves to fix color cast issues in Chapter 4, page 131, but the master curve can also be easily and intuitively used to do either quick or very complex tonal adjustments to a picture.
Curves Tutorial
In the previous chapter, we corrected the “brian_interview_overexposed.” Let’s try this same correction again using the curves in Color Finesse HD+.
Launch Color Finesse. Start a new project or open a previous one and import the shot “brian_interview_overexposed.” In the upper left corner, call up the LUMA WFM, or Luminance Waveform. I recommend doing all of these corrections looking at an external waveform/vectorscope, such as the Tektronix WM7000 used throughout the rest of the book, but I will explain these corrections using the built-in scopes in Color Finesse.
Below the built-in scopes, in the lower left corner, are tabs that allow you to select tools with which to do your corrections (Figure 2.25). Select Curves. The Curves tool presents you with four graphs. As I mentioned earlier, there are Curves tools in many different applications. If you prefer to follow along in Color’s Primary In room or in the Avid Curves tool or one of the other applications, the process will be very similar.
Fig. 2.25 Curves interface for Synthetic Aperture’s Color Finesse.
The leftmost graph, which has a white diagonal line and is called Master, gives you control over the overall signal—the composite of red, green, and blue. The next three graphs to the right give you control over the individual color channels: red, green, and blue, respectively. Each graph has a diagonal colored line indicating the color channel it controls.
How Curves Work
The graphs are interactive. Clicking on a point in the graph and dragging it in a direction alters the relationship between the incoming and outgoing levels. The incoming (source) level is represented along the y axis (up and down) while the outgoing (corrected) signal is represented along the x axis (side to side). So if we want to raise the black level, we click on the point at the bottom left corner and raise it. To lower it, we click on the same point and drag it straight to the right. This remaps the 0 value of the incoming signal, basically saying “all of the values to the left of where I drag this point should be mapped to 0.” If you drag the top right corner down, you remap what was 100 on the source down to where you leave the point. If you drag the top right corner point to the left along the top, you are saying “every source value to the right of this point should be mapped to 100.” In turn, every value along the line to the next point is also remapped.
The best way to see this visually is to load the “ramp_from_0-254” file. I created this image in Photoshop; it has a gentle S shape to it if you look at it with the waveform monitor. Make some adjustments to it in the Curves tool. Watch the Luminance Waveform monitor (LUMA WFM in Color Finesse) while moving the bottom left point of the graph to the right, straight along the bottom of the graph. Notice that the waveform flattens along the bottom right, corresponding to how far you move the point along the Master Graph. Also notice that in the video monitor, the amount of the ramp that is completely black has increased. Now move it straight up along the left edge. The black level in the waveform rises and the video monitor becomes washed out because there’s no longer anything mapped to pure black. In both cases, notice that the most extreme change, visually, is in the blacks or shadows. The gammas move fairly significantly and the highlights don’t change much.
Now do the same for the point at the top right corner of the graph, representing the highlights. Move it down along the right edge, watching as the waveform monitor drops along the right edge from 100 and the video monitor becomes less contrasty because there are no longer any bright whites. Then move the top righthand point to the left along the top edge. Notice that the image in the waveform monitor flattens along the top as all of the values are clipped to 100. Also notice how it changes in your video monitor. The pure white band to the right of the ramped gradation becomes broader and broader as you move the point to the left.
For now, we will concern ourselves only with the first graph, which controls the master level.
We’ll start by determining where our black level (shadows) should be. Looking at the LUMA WFM display, you can see that there is almost no part of the image registering below 30IRE. You can also see that there is significant clipping of the highlights by looking at the tight, flat line along the top of the waveform at 100IRE. That means that we probably will not be able pull any detail out of the sky.
Let’s start with fixing the black levels first. This item is usually the first thing that should be fixed, but it’s doubly important to start with in this instance because that’s where the majority of the problem with this image resides.
So, to lower the black level using curves, you click on the point at the lower left corner and drag it to the right, along the bottom of the graph. If you wanted to raise the black level instead, you’d drag the same point straight up, along the left hand edge of the graph.
TIP
Some applications allow you to lock the axis in which you drag the cursor to either horizontal-only or vertical-only by holding down the Shift key.
Under the graph, you can see a numerical value for your adjustment. When you get to about 40 input, 0 output, some of the darker portions of the waveform display start to crush along the bottom. This is where you start to lose detail in some of the blacks. From here, the amount that you crush that detail is a personal preference. The image still looks washed out to me and the majority of the darker portion of the image is still around 10 or 12IRE.
Lower the blacks even more. Somewhere between 58 and 73 on the input of the graph, I think the black is pulled down low enough and the clipping isn’t too extreme. As you try to focus your adjustment for black somewhere in that range, change from watching the waveform monitor to watching the video monitor. Try to rack “focus” back and forth while looking at some critical areas. In this image, for me, the critical areas are the hair just above his forehead, his eyes, and his skin tone. I don’t want the texture of the hair to be lost by pulling the blacks too low. I also don’t want to pull the blacks down so low that I lose the sparkle in his eyes. There are a lot of reflections in his eyes that give the image life, and if the blacks come down too far, you will lose them. Other colorists that I watched work on this image were less concerned with the eyes and brought the blacks down a lot farther than I did. The other tonal region you need to look at in addition to those two areas is the skin tones in the midtones. You want a nice, rich skin tone that doesn’t look too washed out. There are other things you can do to enhance the skin, so instead of trying to get it perfect right now, use the detail areas of the hair and eyes to determine how low to bring your black levels. I settled on a value of 60 for input and 0 for output. That means that in the source footage, everything (on Color Finesse’s scale from 1–255) that was below 60 is now remapped down to 0. Then the entire range of the source from 60–255 is now spread from 0–255 (Figure 2.26).
Fig. 2.26 Note the difference in the image in the split screen between the source image on the left and the effect that the black correction has on the right.
Now let’s turn our attention to the highlights, which are controlled by the top-right point on the graph. Here’s where that camera focus analogy comes into play. We can see that the highlights are already clipped, so we don’t really need to try to drag the white point to the left along the top. That would just cause the clipping to worsen. But we can try to focus the adjustment down a little to see if we can undo some of the clipping. Pulling the white point down a little (Figure 2.27) basically just lowers the overall level and doesn’t unclip the whites. But the original level was at 110IRE, so I brought my white level down to 231 to bring the image into legal range. (This is just legal for luminance. Chroma information could still be illegal.)
To set the midtones of the image—similar to what you did with the gamma adjustment in Color—click on a point about halfway up the diagonal line. Pick a spot at about 160 input, 120 output (looking at the small numbers below the master curve). The values of this point actually tell you a little about the mathematics of how the signal is being remapped as you make your corrections. If you hadn’t moved the highlights or shadows, all of the points along the diagonal line should match perfectly. The input number and the output number should be identical. But you remapped the shadow values to have a big difference in value (60 input, 0 output); the highlights had a small drop (255 to 231) and the middle of the graph kind of split the difference with a 40-point difference. So you can see that by lowering the shadows, we’ve also already lowered the midtone values somewhat.
Fig. 2.27 Note the split screen showing the source and the correction.
You don’t need to pay too much attention to the waveform monitor as you adjust the gammas. There’s not much that they can tell you. Sometimes, if you adjust the gamma enough, you will create clipping or illegal levels in the shadows if you lower the gamma a lot or in the highlights if you raise the gamma significantly, but as we’re pretty close to the correct level for gamma, all you need to watch is the same detail areas we were monitoring before when we were concerned with setting the shadow (black) level. Watch the eyes and hair for loss of detail and the skin tones for richness as you lift or lower the point in the middle of the line. You can also move the middle point left or right. The move in the blacks richened up a lot of the midtones, and the move in the highlights brought down the overall levels as well. Because of this, even though I usually end up pulling down gammas to richen up the image, with this one I ended up bringing them up a little bit so that the skin tones didn’t look too dark (Figure 2.28).
With curves, you can also make adjustments to very specific tonal ranges. For example, you may be able to soften some of the clipping in the sky areas by adjusting a point near the top of the graph. You won’t be able to bring back any detail into the blown-out sky, but you may be able to create some texture in the areas where the clipped sky rolls off or transitions to another element, like the edges of the trees.
Fig. 2.28
The thing you want to avoid when you use two points on a curve that are fairly close to each other, is “posterizing.”
Try placing a point on the Master Graph less than a quarter of the way from the top. Then pull that point down a little while you watch the areas that transition from the clipped sky. The thing you want to avoid when you use two points on a curve that are fairly close to each other, is “posterizing.” (See definition for posterization at the top of this chapter.) If you adjust the point too radically, it will posterize. As much as you may want to eliminate the clipping in the sky, having posterization is much worse. If you find a good balance before the posterizing occurs, you may be able to create some texture in the areas surrounding the clipped sky. I was able to bring it down by only a very little bit. You can also watch the top of the waveform while doing this and you will see the compressed, clipped area at the top of the waveform start to stretch out (Figure 2.29). Don’t get “target lock” while doing this. You’ll be thinking: “Wow! I’m unclipping all of the detail in the sky.” But you need to see on the picture monitor whether you are really just introducing noise and banding or posterizing. Be like a doctor: “First, do no harm!”
Fig. 2.29 Final Curves adjustment in Color Finesse.
One of the problems that begins to emerge in this image as the tonal range improves is that a lot of noise and artifacting of the video becomes evident. If you have no means of correcting this noise, then you may have to either limit how far you take your correction or determine that the noise is a small price to pay for the improved contrast. Many color correction systems have noise reduction built in to them, and you can use that feature to fix the noisiness that you introduced. The thing with noise reduction is that it must be used with a very gentle hand. Always check the image before and after noise reduction to make sure you haven’t applied too much.
Isolating Tonal Ranges with Curves
Another cool thing about working with curves is that you can isolate specific tonal ranges using points that you don’t move so that other corrections on either side don’t affect them.
Let’s try an example. To best understand exactly what this isolation is doing, open the “grayscale_neutral” file from the DVD and load it into an application with curves, like Color, Avid, Photoshop, or After Effects.
With this example, what we want to see is how we can isolate a specific tonal range so that it is not affected by the corrections that you make in another area. First, let’s see what happens when we don’t isolate the tonal range.
With the “grayscale_neutral” clip loaded, look at the image on the waveform monitor and on the video monitor. The levels are almost correct. The highlights are a little below 100IRE, but the black levels are correct and the gammas are just about right (Figure 2.30).
Now, let’s adjust the shadow area with a point about 25 percent up from the bottom of the master curve. Because the range is 1–255, that that would equal an input value about 63. Hover the cursor over the diagonal line in the master curve while watching the input value just below the master curve graph. When it gets to around 63, click on the line and drag that point down to around 40 while watching how the trace of the waveform monitor reacts. You’ll notice that while the majority of the correction is in the bottom of the waveform monitor (the shadows), the midtones and even the highlights are affected to some extent (Figure 2.31).
In some real-world corrections, you may want to adjust the deep shadows without changing your midtones or highlights. The key to doing this with curves is to add points on the curve that you do not move at all. These points will protect the rest of the curve from moving.
Fig. 2.30 Starting point for the curves adjustment.
Fig. 2.31 First Curves adjustment in the deep shadows to make the black deeper.
Reset the master curve. In Color Finesse, this is done with the reset button at the lower right. In Avid, Alt-click on the activated tab to reset. In many other applications, click on the altered point and hit the delete key on your keyboard. In Color, click on the small diamond in the upper lefthand corner of the curve to reset the entire curve or just drag the point off the graph to remove it. In Color Finesse, right-click a point on the curve and choose Delete Point from the pulldown menu.
This time, let’s place a point on the graph that is just below the halfway point. I set my point at 120. This point will isolate the midtones and highlights from the corrections we make lower on the curve (Figure 2.32).
Now, click on the same point (63) that you adjusted before and drag the point up and down as you watch the waveform monitor and video monitor. You’ll notice that the changes to the midtones and highlights are much less obvious than before. They still move a little bit depending on what application you use because the point is controlling the curve like a Bezier curve, so the point is not really a hard and fast cutoff of the correction. Some applications add much less of a Bezier effect to the curve. Avid should really require only a single point to isolate the correction. If you want to, you can add a second point just above the first isolation point. Try 130 or 140. This will limit the amount that the Bezier curve affects the curve about the higher mark. Now, move the lower (shadow) point again while watching the waveform monitor. Also watch the video monitor; you can see that the shadows are being deepened, while the highlights and most of the midtones don’t move at all.
Fig. 2.32
If you bring the shadows down far enough, you will see that you start to crush the middle chip that is at 0IRE. The crushing or clipping is indicated by the waveform trace starting to flatten out. If you want to protect the deepest blacks from clipping, you could also place a point on the curve that is very low. This will protect the detail in the deep blacks while allowing you to pull the rest of your shadows deeper. I placed my point at 16. Then, as I brought my shadow point (input 67) down to an output of 51, the darkest black did not clip (Figure 2.33).
To practice this some more, continue this same correction without resetting it and try to get the brightest chip on the right side (the chart was lit a little brighter on the right side) to 100IRE and the next three chips to land at 80, 70, and 60.
WARNING
A reminder about moving points that are too close together: There are only so many levels of gray between points, and if you stretch them too far or condense them too closely, your image will fall apart. By “fall apart,” I mean that it will either exhibit a lot of noise or posterization or banding. These faults in the picture will be much more noticeable than any errors that you’re trying to correct, so try to avoid taking the image beyond where it can go.
The key to this exercise is figuring out what point on the master curve corresponds to the luminance value of the chip as it is viewed in the waveform monitor. This is a very valuable skill to have. You should be able to look at a waveform display and figure out approximately what parts of the image on the video monitor correspond to the waveform display.
Fig. 2.33
Luma Range Display
The skill of knowing what parts of the image on a waveform will be affected by a specific tonal correction isn’t important only in using curves. You also have to be able to do this if you’re just using shadows, midtones, and highlights. Color Finesse and Avid Symphony have a nice function that allows you to develop this skill. It’s called Luma Ranges. In Color Finesse, this is one of the main viewing modes at the bottom of the list of tabs in the upper right pane. When you view your source using Luma Ranges, it shows you each of the three tonal ranges as a shade of black and white. So parts of the image that you would control with the shadows control are black. Parts that are considered midtones are displayed in gray, and parts that are considered highlights are white (Figure 2.34).
Import some of the test images from the DVD, or bring in some movies and images of your own and try to guess what the resulting Luma Range display will look like. This is a fairly simple exercise with a well-lit image, but if it’s over or underexposed, your eye will get fooled into spreading the tonal range. For example, on an underexposed image, you will be surprised how little of the image is considered a highlight.
Fig. 2.34 Luma Range GUI in Synthetic Aperture’s Color Finesse 2.
In the image of Brian in Figure 2.34, I used the Luma Range editing tools so that there would be a good distribution of black, gray, and white in the image. The actual source image shows up as mostly gray and white when no Luma Range editing is done.
Luma Range Editing
With Color Finesse and Avid Symphony, there are ways to alter the definitions of the three Luma Ranges: shadows, midtones and highlights. In other words, you can determine what the application considers to be shadow and where it transitions to being a midtone and then where midtones become highlights. For most normal images, you don’t need this kind of control. The definitions make sense and give you the kind of control you want. But for certain shots, you may want to alter the definitions of these tonal ranges so that you can gain greater control over specific parts of the picture. I don’t have specific technical information from the manufacturers, but I would assume that most applications consider the range from 0IRE to about 25IRE to be shadows, from 25IRE to 75IRE to be gammas and from 75IRE to 100IRE to be highlights. Obviously, these numbers (even if they’re right) wouldn’t be hard cutoff points. It would work something like this: if you move the shadow control, the darkest 10 percent would be affected 100 percent by the control, the next 10 percent of the darkest part of the image would be affected 90 percent by the control and so on, until the brightest parts of the picture aren’t affected by the shadow control at all.
Avid Symphony and Color Finesse both have tools to define Luma Ranges. They are essentially the same, though Symphony gives a little bit of added control. At the current time, there is no ability to edit Luma Ranges in the lower-end Avid products.
Figure 2.35 shows the Luma Range control for Color Finesse. It shows a histogram of the image with three curving lines overlaid and two straight lines that are the controls for the Bezier curves of the shadow and highlight curves.
Fig. 2.35
The curving arcs represent the definition of the shadows, midtones, and highlights. The tonal ranges are represented by overlapping curves instead of as strictly defined and delineated areas because if the definitions of the tonal ranges were defined with a sharp cutoff, corrections to individual tonal ranges would cause sharp, visible transition lines in the picture at the point where one tonal range was defined from another.
In Color Finesse, the curves of Luma Range are interactive. The only way to alter the definition of the midtones is by altering the curve for highlights and/or shadows. In Symphony, each curve can be radically altered by setting numerous points on any of the curves. This is a lot of power, but it can really cause bizarre artifacts in the shot because the curves can be set to overlap with each other or cut off with abrupt transitions or even leave entire portions of the picture that are not defined by any tonal range at all!
The value of this control is that it is possible to define very specific portions of the picture. One example of where this ability would be useful would be in a shot with a hot window. If the rest of the shot was very well lit and you tried to use your highlight control to bring down the intensity of the window, it might also bring down other highlights in the rest of the room. If you want to limit your correction to the window alone, you could use Luma Ranges (Figures 2.36 and 2.37) to define the highlights of the picture to only include the window values. Then you could use your midtone and shadow controls to control the rest of the image.
Fig. 2.36 In this image, the Luma Ranges are basically unedited. Notice what parts of the grayscale image the Symphony considers shadows, midtones, and highlights in its default mode.
Fig. 2.37 Here is the same grayscale image with no color correction, but the Luma Ranges have been edited so that there is much less of the image that is considered by the software to be midtones. This would be useful if you wanted to make an adjustment that affected only a very small range of tones in the absolute middle of the picture.
Alternative to Luma Range
Experienced colorists get a lot of this same ability to isolate a very specific tonal range by creating a matte that qualifies a specific tonal range (Figure 2.38). Because this type of correction is really defined as a secondary correction, we’ll discuss it further in the next chapter.
Definition
qualify: This term means that an area of the picture is specifically isolated for a correction by any number of methods. You could qualify something for correction using its hue, chroma strength, or tonal value. You could also qualify an area of the image using a window or garbage matte. For example: “I qualified the brightest highlights by making a matte of everything over 90IRE and added a bit of yellow to them.”
Fig. 2.38 A key that was created in the Color Secondary Room to qualify the highlights of the “brian_interview_overexposed” shot. The areas defined in white would be the only areas qualified for the correction.
Thinking about the Budget
One consideration before you start to play with the Luma Range editing capabilities is that—like most things—color correction is usually done on a deadline and with a budget.
To stay on budget or deadline with your corrections, you should stick with the main or primary color correction capabilities of your application. A sure way to kill that budget is to define the specific Luma Range of every shot, or to add color effects or secondary color correction to every shot. Secondary color corrections and Luma Range definition are fantastic tools that help you accomplish specific tasks, but you need to consider how long you have to grade the entire project and how much time you can devote to each shot. Hopefully, you can make each shot of a longer form project look pretty good in under a minute. I typically had about two days to grade a 600-shot, 48-minute documentary. That works out to about a minute and a half per shot or 20 minutes of color correction work for each finished minute of programming. American dramatic primetime shows are usually in the range of 1000 shots in a one-hour show and are usually graded in 12 to 16 hours, averaging a bit better than a shot a minute. Reality shows are closer to 1200 shots in an hour-long show and only usually budget for a single day, including laying it off to tape, which works out to about 170 shots an hour or close to 3 shots a minute. Color correction for digital intermediates can vary greatly, but can average about 20 minutes (about two reels) per day.
Neal Kassner, a colorist for CBS’s 48 Hours estimates that he has about 16 hours to color correct that show’s 1200 to 1500 shots per episode. That’s 75 to 90 shots an hour. Other colorists I’ve spoken to have mentioned averages for a nationally telecast documentary as 6 to 8 minutes to correct one minute of finished program time. On spots, the average is 3 to 8 hours for a single 30-second spot or a series of spots based on the same material. Some facilities expect certain output from their colorists, such as 100 shots an hour.
Craig Leffel, of Chicago’s Optimus, says that it depends if the corrections are from tape or server or if the OCN has to actually be racked on the telecine. For film negative, the average is 4 to 6 shots an hour if there are only a few shots per reel.
Legendary colorist Bob Festa, formerly of R!OT in Santa Monica, says that for spots he corrects off the telecine, the average is 10 shots an hour. “Unfortunately, in today’s world, I’m still racking up film on a day-by-day basis. Today I was working with dailies rolls and we had 30 shots in 3 hours basically. So that’s pretty much to my formula of 10 shots per hour.” (Most colorists working from telecine have an assistant that threads up the telecine for them.) In a new interview for this edition of the book, Festa says that the majority of his work is spots, and he grades a typical 30-second spot in 3 to 5 hours. His company, New Hat, also does DI work, which typically takes 40 to 80 hours to grade a midlevel 2-hour feature film. Festa’s company doesn’t do TV series regularly, but his colleagues at his former home, R!OT, generally grade a 30-minute HBO series in about 8 hours.
Pankaj Bajpai, who is also featured throughout this book, grades top-tier episodic television at Encore in Hollywood. According to Bajpai, “It depends on the project; it depends on the budget for the project because you can do an episodic in ten or 14 hours. We average about 16 to 18 hours on an hour long episodic but you can go up to 25 or 30 hours on a budget. I have done that obviously on a show like Rome where you’re not gonna get it done in 20 hours because every detail is just finessed to the nth degree so it go 25 to 30 hours. If it’s a period drama like Carnivale from HBO, it was a period drama set in the Dust Bowl era and everything was heavily processed on color, and we did it in 24 to 26 hours per episode, but it was an enormous amount of work to create a time and place. In Santa Clarita, they would shoot hills of green for the Dust Bowl. The reality of it is you have to be able to manage your time, and the way you do that in my view is that you don’t spend a lot of time talking about it, you show options and find out if this is what they are talking about and then move, you have to constantly keep moving on and I find that if you get stuck, one thing and people start to obsess on one frame then you’re hosed. So you’ve gotta keep moving and move in way that people don’t feel rushed. That’s the other aspect of coloring. I don’t think were talking as much about color here but about the other aspect of coloring which is important to know.”
Some of these numbers have changed somewhat over the years, as colorists are transitioning from a workflow that was originally almost entirely “straight off the telecine” to a current workflow where telecine transfers get transferred “flat” to either a digital disk recorder, some kind of a server as a file or to a tape format like D5, then the colorist basically does a tape-to-tape color correction or color corrects from a file. Back in the day, a rule of thumb for telecine transfers was 1 hour to grade 11 minutes (one 1000-foot 35 mm reel).
In even more recent history, the technology and workflow of color correction has transitioned yet again. Instead of the telecine transfers or film scans, colorists are primarily seeing file-based footage from cameras such as the RED, ARRI ALEXA, and other HD and digital cinema cameras. Festa claims he hasn’t touched a roll of OCN since 2007 or 2008.
Regardless of the technology or the originating format, the trick to grading an entire project on budget is to leave enough extra time to work on the shots that really need the additional attention.
Until you get more experienced at estimating how long you need to really tweak an entire project, try to get a first pass at all the shots done in half of your budgeted time. Then use the second half of the time to polish the overall corrections and devote extra time to “trouble” shots or those that have high emotional significance or importance to the story,
Also, don’t forget to leave time for revisions, especially if you don’t have absolutely every single decision maker in the session.
Definition
OCN: Original Camera Negative.
racked: Physically placing the spool or reel of film on the telecine and threading it.
telecine: This is the machine—or sometimes used as a description of the process—that transfers film to video in real time. The telecine feeds the image to the color correction hardware. I’ve heard at least three different pronunciations and everyone will tell you that the way they pronounce it is correct. TELL-uh-sin-ee. Tell-uh-SEEN. TELL-uh-sinuh. Most of the interviewees, including Bob Festa, who’s probably been at it longer than anyone else, pronounced it “TELL-uh-sin-ee,” with the heavy accent on “tell” and a lesser accent on “sin.” The other way to transfer film to video (or data, actually) is with a film scanner, which, as of the writing of this book, is not real time, but is getting close.