Showing posts with label for reference. Show all posts
Showing posts with label for reference. Show all posts

Wednesday, July 24, 2013

How To Become A Feature Film Compositor


How To Become A Feature Film Compositor

by John Lafauce, Jr.

 

What skills are needed to become a compositor? >> Become a problem solver

 

Probably the most important skill you can possess, and I can't emphasize this enough, is the ability to problem solve and be resourceful. Often times on a project, image elements are assigned to you that don't fit together at all and it's the compositor's responsibility to make them seamless and believable when they are composited. Elements might come to you in a different color space, resolution, or format than what you are presently working in, and you have to make adjustments as needed. Objects may need to be rotoscoped or painted out of a shot, which in a smaller facility, where you are more of a generalist, you would be responsible for doing this. In a larger facility, the roles are more specialized and you would most likely be compositing only.

 

Well said soul but to be a good compositer its important that u work inextreme detail....and understand the overall process.that becomes allthe more important when u work on film.as one small pixel will zoom upto 2000 % on film.and even smallest mistake is visible..and that nevercomes easy.color knowlage is essential though to seamlessely integratethe invisible effects.i personaly feel that if u have overall knowlageof 3d and digilat painting it always helps.....and for 100,000 Rs i would say woohh too...may be one fine day i too c that on my paycheck...aameen----------------

 

I agree with u guyz, Good amount of detail and an eye to judge thedifficulty in a shot. As being one of the last stage we often lend updoing some very difficult task. At times it becomes difficult ratherimpossible to go back to CG and change it so we compositors have totwist our brains and deliver a quality output. Software knowledge isimportant but aesthetic skills, tricks which comes with experience areessential too.

 

Even i feel a person with a 3d bg can understand and is able to do good compositing.

 

> Hard working, loyal, want to learn a lot of stuff, creative ideas, etc.

… the usual wire removal, clor correction

... b4 after clips... explosions...

… match moving capability (for certain studios) in fact most of theplaces nowdays...

Well it depends on how u present your self and your work...etc...

 

 

--
Warm Regards,
Lakshmi Raj

Saturday, June 15, 2013

Creativity - What it is and what it takes ?

Meaning of Creativity

Definition

Creativity is the ability to think up and design new inventions, produce works of art, solve problems in new ways, or develop an idea based on an original, novel, or unconventional approach.

Description

Creativity is the ability to see something in a new way, to see and solve problems no one else may know exists, and to engage in mental and physical experiences that are new, unique, or different. Creativity is a critical aspect of a person's life, starting from inside the womb onward through adulthood.

Although many people equate creativity with intelligence, the two terms are not synonymous, and it is not necessary to have a genius-level IQ in order to be creative. While creative people do tend to have average or above-average scores on IQ tests, beyond an IQ of about 120 there is little correlation between intelligence and creativity. Researchers have found environment to be more important than heredity in influencing creativity, and a child's creativity can be either strongly encouraged or discouraged by early experiences at home and in school.

Standard intelligence tests measure convergent thinking, which is the ability to come up with a single correct answer. However, creativity involves divergent thinking, which is the ability to come up with new and unusual answers.

Creative individuals tend to share certain characteristics, including a tendency to be more impulsive or spontaneous than others. Nonconformity (not going along with the majority) can also be a sign of creativity. Many creative individuals are naturally unafraid of experimenting with new things; furthermore, creative people are often less susceptible to peer pressure, perhaps because they also tend to be self-reliant and unafraid to voice their true feelings even if those go against conventional wisdom.

Creativity in childhood is typically assessed through paper-and-pencil measures such as the Torrance Test of Creative Thinking. These tests are designed to measure divergent thinking, such as fluency, flexibility, originality, and elaboration. Signification criticisms have been raised about these tests as measures of creativity. First is the general problem that there are no universally accepted definitions of creativity. Second, critics of creativity tests argue that these tests do not measure creativity per se but instead reflect the specific abilities that are assessed by the tests. Third, the scores on these tests often depend partly on speed, which is not necessarily a criterion for creativity. A final consistent concern relates to the scoring of creativity tests, which by definition are somewhat subjective. Thus, the reliability of such tests is commonly questioned.

Saturday, May 18, 2013

Optimizing Images for Web

Optimizing Images for web using Photoshop
Links : http://sixrevisions.com/web_design/comprehensive-guide-saving-images-for-web/


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~X0X~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Thursday, May 9, 2013

A few scanning tips


A few scanning tips

Printing Resolution

Printing acts the opposite of video in almost every way.

Image program and printer driver options can cause results to vary. The software may have various options to make the printed size be different, to fill the page for example, or to scale the image. It is difficult to specify exactly what will happen in every case, but if it isn't going well, look for some obscure option setting affecting print size.

In most programs like Photoshop or PhotoImpact or PhotoDeluxe, the scans are printed at real size. That is, if you scan a 6x4 inch photo, and then select the menu FILE - PRINT, it will print 6x4 inches on the paper too. Resolution does NOT determine image size on the printer as it does with video. The size of the original scan area determines printer image size. Lower resolution can look fuzzy, and higher resolution may look better, but the printed size will be the same at any scan resolution (scaled resolution is something quite different, discussed below). Printed resolution works like we think of resolution, as increasing image detail instead of image size. But even so, like video, there are limits; we are still very much limited by the capabilities of our output device.

Printed pages are normally a standard size defined in inches, 8.5 x 11 say, so printed image size in inches is very meaningful too. The printer driver will try to print the image at its original size in inches, unless told otherwise. Conversely, video screens don't care about inches. If you have a 400x400 image, video monitors will show it as a 400x400 image (unless the viewing software's goal in life is to make it fit or fill the screen, etc). Notice that both techniques do maintain the same relative size of the image to the total size of the page or screen.

To print images at the actual original size on the paper, the value for resolution entered into the scanner's TWAIN driver is both the scanning resolution and the printing resolution (assuming 100% Scale Factor). The same resolution (same dots per inch) is what makes the image print at actual real size. Scan and print, that's about all there is to it, unless your printing software interferes with its own notions about how it should be done. TIF and JPG image files store this resolution number in the file for later use (GIF files do not), and good applications will honor it. Some applications do not.

But when we want to print the scanned image at a different size on the paper, there are special techniques and concerns for scaling the image size. Since this is the norm instead of the exception, we'll give it special attention here. Scaling is a word that means stretching or shrinking the image to fit a specified area, and it is accomplished by simply changing the value of the number used as resolution when the printer calculates the spacing of the dots on the paper. Nothing else happens, except that the size of the printed image changes. It's a simple operation, but the concept is a little abstract. It's much harder to explain than to do. I'll spend some time on it if you're interested in the whys of it. Don't let all my words scare you away, that's just my own shortcoming, and scaling is really quite easy.

Scaling is an extremely important working tool for printing. There are two ways to scale an image, either during the scan, or after the scan. The next page covers scaling AFTER the scan. 
Courtesy : Link

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Further for better understanding, you can go to the following Link, for the PDF file on Image Resolution Explained - Link

Pick Your Scanning Resolution


                           Pick Your Scanning Resolution
                                    Calculate the Best Resolution for Your Scans


   The best resolution to scan your images depends on how it will be used - on screen or in print. For the purposes of this article, we'll assume the image is for print. To find your best scanning resolution you need to calculate the final image resolution, that is, the resolution of your image when placed in your file for printing. Your scanning resolution and final image resolution may be the same, but not necessarily. For most users, it is rare that you will need thousands or even more than 200-400 pixels of resolution.


To reach your final image resolution, you need to know 5 things before scanning:

Related Resources
• Resolution Basics for Beginners
• Resolution in Detail - SPI, PPI, DPI, LPI
• Scanning Resolution, Scaling, File Size

  1. Size of the original image.
    Are you scanning a 4x5 inch photograph, for example?
  2. Size of the printed image.
    Will your image be reduced or enlarged?
  3. LPI at which your image will be printed.
    LPI depends on the printer you are using. The type of paper, also affects LPI. Your final image resolution is usually about double the LPI at which the image will be printed.
  4. Resolution Multiplier.
    This number is usually 2. You can see how it is used below. For most purposes, 2 is the magic number.
  5. Editing of the scanned image.
    It is often better to scan at a somewhat higher resolution than you'll eventually need to allow for pixels lost during image editing. Just remember, the higher the resolution, the larger the file size. Don't go overboard.
For commercial printing, if you don't know the required LPI, talk to your print shop. They can help you come up with the right numbers based on the specifications of your print project. Here's a quick reference list or see the more detailed chart in the Resolution Inch by Inch article on this site:
screen printing 35-65
laser printer/photocopier 50-110
offset printing 60-150+
standard for newsprint 85
standard for offset printing 133

To calculate your final image resolution:

  • If the size of your image will not change, multiply LPI by your resolution multiplier. LPI x 2
    For example, an image destined for a brochure on uncoated paper using offset printing needs a final resolution of approximately 133 x 2 or 266.
  • If the size of your image will change, multiply LPI by your resolution multiplier and then by the size of your final image as a percentage of the original. LPI x 2 x %
    For example, if our image for the brochure, above, is to be enlarged by 25%, then it needs a final resolution of 133 x 2 x 125% or 333. If it is to be reduced by 25%, then it needs a final resolution of 133 x 2 x 75% or 200. (I've rounded up in each case.) Don't know the percentages? Use the scanning calculator, below.
You may be fine scanning at the final resolution arrived at by one of the formulas, above. However, since most scanned images need some type of image editing, scanning at a slightly higher resolution is often advantageous. It is easier to "throw out" unneeded resolution after scanning than it is to restore lost resolution after the fact. Just remember to reduce your image to the final resolution before placing it in your document for printing.
Having trouble doing the math? Try this simple scan resolution calculator:


Having trouble doing the math? Try this simple scan resolution calculator:
Size of the original image:
WIDTH:
HEIGHT:
Size of the printed image:
WIDTH:
HEIGHT:
LPI at which your image will be printed:
Resolution Multipler
(2 is standard, for most uses no need to change this number):
<--choose one--="">

The image will be enlarged/reduced by: percent Your scanning resolution should be at least:
samples per inch


Getting Started Scanning...

• Introduction
• Part 1: DPI for Email & Web
• Part 2: Saving Your Scans



Part 3: Printing Digital Photos & Scans

The information here pertains to both color and grayscale images that will be printed on an inkjet printer. It does not apply to printing line art images. For more discussion on printing line art images see Scanning Line Art by Desktop Publishing Guide Jacci Howard Bear. (Continued below...)



PPI vs. DPI
The first important point to understand is the differences between PPI (pixels per inch) and DPI (dots per inch). Many software programs and scanner interfaces use these two terms interchangeably but that's not exactly accurate. As a general rule, and for the purposes of this article, the term PPI should be used when referring to image resolution, and the term DPI should be used when referring to printing resolution. How can you remember this? Monitors display pixels, and printers produce dots. So, as you follow this discussion, whether your software calls it DPI or PPI, when I talk about the resolution of your image, I will be using the terminology PPI.

Pixel Resolution
Digital photos and scans are all bitmap graphic types, no matter what format the image is saved to. That means they are made up of a grid of individual, tiny pixels with each pixel representing a single color in the image. If you zoom in on one of these images (see example) in your imaging software you'll be able to see the individual squares of color.
Bitmap images are made up of individual pixels.
What this means in relation to printing is that these types of images are resolution-dependent. In other words, the quality of the print and the size of the printed image is limited by the number of pixels in the image. You can't increase one value without effectively decreasing the other.

Inkjet Printer DPI
Today's inkjet printers have three standard output settings:


normal:
300 x 300 or 320 x 320 dpi
high quality:
600 x 600 or 720 x 720 dpi, 1440 x 720
photo quality:
1200 x 1200, 1440 x 1440 dpi, 2880 x 1440 and up

You might also have a draft or economy setting, but you should never use this setting for printing images. It's primarily used for printing text and rough drafts.
A popular myth is that it's necessary to scan an image at the same resolution that you will be using to print. In the case of color and grayscale images, this only results in excessive file sizes. At one time, a good general rule for inkjet printing was that you needed half to one-third of the PPI of the printer's DPI setting that you intend to use. So if you're using your printer's "normal" setting (300 dpi), your image needed to have at least 150 ppi. When using the higher quality printer settings (720 dpi and up), you can bring the PPI down to about 1/3 of the output resolution. That would be 240 ppi for your printer's 720 dpi setting.
Since writing this article photo printers have gotten much better--the dots are smaller and more compact--but still, you rarely need your image resolution to be higher than 240-300 ppi for inkjet printing.

Calculating Image Size and PPI
So how do you calculate the image size you need? It's simpler than you might think. First, decide what size you want your printed image to be, then multiply the height and width in inches by the PPI you need for the print resolution you'll be using. Here's a reference chart for some common print sizes:
 
Printer Quality
Draft/Economy
Standard
High/Photo
Print @
300 - 320 dpi
600 - 720 dpi
1200 - 2880+ dpi
Scan @
150 ppi
150-240 ppi
240-360 ppi
Printed Size
Actual Pixel Dimensions (Average)
2" x 3"
300 x 450 pixels
400 x 600 pixels
600 x 900 pixels
4" x 6"
600 x 900 pixels
800 x 1200 pixels
1200 x 1800 pixels
5" x 7"
750 x 1050 pixels
1000 x 1400 pixels
1500 x 2100 pixels
8" x 10"
1200 x 1500 pixels
1600 x 2000 pixels
2400 x 3000 pixels
Generally, printing photographic images requires more pixels per inch. For photo printing, you will get better results using the higher end of the scale. For scanned illustrations, cartoons, or paintings, you can use a lower PPI and not see a noticeable change in the printed quality.

Why is my scan so HUGE?
Keep in mind that when you scan an image at anything higher than 72-100 ppi, it's going to display much larger on your monitor than the original picture or page. That's because your monitor can only display a fixed number of pixels per inch, usually 96. When you place the image into a page layout program, most software will be able to read the PPI information and will place your image into the page at the proper size. Some of the low-end printing software may not, however, and you'll have a very large image placed into your page. When that happens, you'll have to size the image in your layout program to the printed size you based your calculations on.

But I don't know what the output will be!
Whenever you increase or decrease the pixel dimensions of an image you lose some quality, but it's less destructive to reduce pixel dimensions rather than increase pixel dimensions. If you're scanning for archival, or you don't know what the intended output will be, it's better to scan at a higher resolution and reduce the image later. However, it is of no advantage to scan higher than your scanner's maximum optical resolution.

PPI and Digital Cameras
When you're printing images from a digital camera, you have to calculate things a little differently. With a digital camera, you have a fixed number of pixels which is the maximum your camera is able to capture. Today's lower-priced cameras have an average maximum resolution of 1200 x 1600 pixels (2 Megapixels). In this case, we have the pixel dimensions, so we need to work in reverse to figure out the best quality setting and printed size.

Using the formula above we can calculate that with 1200 x 1600 pixels we can either get a normal quality print at 8 by 10 inches, a high quality print at 5 by 7 inches, or a photo quality print at 4 by 5 inches maximum. That's not a very big image for a photo-quality print, so if you thought you'd be producing photo-quality 8 by 10 prints with your $200 digital camera and inkjet printer, you're probably going to be somewhat disappointed. As you can see, there is a choice to be made as to whether you want to sacrifice printed size or image quality. If you plan to print many 8 by 10 prints, you should have at least a 3 megapixel camera, although I have gotten acceptable results printing 8 by 10s from a 2 megapixel original on a new photo-printer.
To review:
  • PPI refers to image resolution
  • DPI refers to printer resolution
  • Inkjets need one-half to one-third the PPI of the printer's DPI setting
  • Print size and print quality have an inverse relationship
  • Print size and print quality are dependent on pixel resolution
Link Courtesy : Link

````````````````````````````````````X0X````````````````````````````````````````

Adjusting image sharpness and blur


Sharpening recommendations

Sharpening enhances the definition of edges in an image. Whether your images come from a digital camera or a scanner, most images can benefit from sharpening. The degree of sharpening needed varies depending on the quality of the digital camera or scanner. Keep in mind that sharpening cannot correct a severely blurred image.
Tips for better sharpening:
  • Sharpen your image on a separate layer so that you can resharpen it later to output to a different medium.
  • If you sharpen your image on a separate layer, set the layer’s blending mode to Luminosity to avoid color shifts along edges.
  • Sharpening increases image contrast. If you find that highlights or shadows are clipped after you sharpen, use the layer blending controls (if you sharpen a separate layer) to prevent sharpening in highlights and shadows. See Specify a tonal range for blending layers.
  • Reduce image noise before sharpening so that you don’t intensify the noise.
  • Sharpen your image multiple times in small amounts. Sharpen the first time to correct blur caused by capturing your image (scanning it or taking it with your digital camera). After you’ve color corrected and sized your image, sharpen it again (or a copy of it) to add the appropriate amount of sharpening for your output medium.
  • If possible, judge your sharpening by outputting it to the final medium. The amount of sharpening needed varies among output media.
Use the Unsharp Mask (USM) filter or the Smart Sharpen filter for better control when sharpening your images. Although Photoshop also has the Sharpen, Sharpen Edges, and Sharpen More filter options, these filters are automatic and do not provide controls and options.
You can sharpen your entire image or just a portion using a selection or mask. Because the Unsharp Mask and Smart Sharpen filters can be applied to only one layer at a time, you might need to merge layers or flatten your file to sharpen all image layers in a multilayered file.
Note:
The name Unsharp Mask comes from a darkroom technique used in traditional film‑based photography. The filter sharpens images rather than the opposite.

Sharpen using Smart Sharpen

The Smart Sharpen filter has sharpening controls not available with the Unsharp Mask filter. You can set the sharpening algorithm or control the amount of sharpening that occurs in shadow and highlight areas.
  1. Zoom the document window to 100% to get an accurate view of the sharpening.
  2. Choose Filter > Sharpen > Smart Sharpen.
  3. Set the controls in the Sharpen tabs:
    Amount
    Sets the amount of sharpening. A higher value increases the contrast between edge pixels, giving the appearance of greater sharpness.
    Radius
    Determines the number of pixels surrounding the edge pixels affected by the sharpening. The greater the radius value, the wider the edge effects and the more obvious the sharpening.
    Remove
    Sets the sharpening algorithm used to sharpen the image. Gaussian Blur is the method used by the Unsharp Mask filter. Lens Blur detects the edges and detail in an image, and provides finer sharpening of detail and reduced sharpening halos. Motion Blur attempts to reduce the effects of blur due to camera or subject movement. Set the Angle control if you choose Motion Blur.
    Angle
    Sets the direction of motion for the Motion Blur option of the Remove control.
    More Accurate
    Processes the file more slowly for a more accurate removal of blurring.
  4. Adjust sharpening of dark and light areas using in the Shadow and Highlight tabs. (Click the Advanced button to display the tabs). If the dark or light sharpening halos appear too strong you can reduce them with these controls, which are only available for 8‑bits and 16‑bits-per-channel images:
    Fade Amount
    Adjusts the amount of sharpening in the highlights or shadows.
    Tonal Width
    Controls the range of tones in the shadows or highlights that are modified. Move the slider to the left or right to decrease or increase the Tonal Width value. Smaller values restrict the adjustments to only the darker regions for shadow correction and only the lighter regions for highlight correction.
    Radius
    Controls the size of the area around each pixel that is used to determine whether a pixel is in the shadows or highlights. Moving the slider to the left specifies a smaller area, and moving it to the right specifies a larger area.
  5. Click OK.

Sharpen using Unsharp Mask

The Unsharp Mask sharpens an image by increasing contrast along the edges in an image. The Unsharp Mask does not detect edges in an image. Instead, it locates pixels that differ in value from surrounding pixels by the threshold you specify. It then increases the contrast of neighboring pixels by the amount you specify. So, for neighboring pixels the lighter pixels get lighter and the darker pixels get darker.
In addition, you specify the radius of the region to which each pixel is compared. The greater the radius, the larger the edge effects.


Original image, and Unsharp Mask applied
The degree of sharpening applied to an image is often a matter of personal choice. Keep in mind that oversharpening an image produces a halo effect around the edges.


Oversharpening an image produces a halo effect around the edges.
The effects of the Unsharp Mask filter are more pronounced on‑screen than in high-resolution output. If your final destination is print, experiment to determine what settings work best for your image.
  1. (Optional) If your image is multilayered, select the layer containing the image you want to sharpen. You can apply Unsharp Mask to only one layer at a time, even if layers are linked or grouped. You can merge the layers before applying the Unsharp Mask filter.
  2. Choose Filter > Sharpen > Unsharp Mask. Make sure the Preview option is selected.
    Click the image in the preview window and hold down the mouse to see how the image looks without the sharpening. Drag in the preview window to see different parts of the image, and click + or – to zoom in or out.
    Although there is a preview window in the Unsharp Mask dialog box, it’s best to move the dialog box so you can preview the effects of the filter in the document window.
  3. Drag the Radius slider or enter a value to determine the number of pixels surrounding the edge pixels that affect the sharpening. The greater the radius value, the wider the edge effects. And the wider the edge effects, the more obvious the sharpening.
    The Radius value varies according to the subject matter, the size of the final reproduction, and the output method. For high-resolution images, a Radius value between 1 and 2 is usually recommended. A lower value sharpens only the edge pixels, whereas a higher value sharpens a wider band of pixels. This effect is much less noticeable in print than on‑screen, because a 2‑pixel radius represents a smaller area in a high-resolution printed image.
  4. Drag the Amount slider or enter a value to determine how much to increase the contrast of pixels. For high-resolution printed images, an amount between 150% and 200% is usually recommended.
  5. Drag the Threshold slider or enter a value to determine how different the sharpened pixels must be from the surrounding area before they are considered edge pixels and sharpened by the filter. For instance, a threshold of 4 affects all pixels that have tonal values that differ by a value of 4 or more, on a scale of 0 to 255. So, if adjacent pixels have tonal values of 128 and 129, they are not affected. To avoid introducing noise or posterization (in images with flesh tones, for example), use an edge mask or try experimenting with Threshold values between 2 and 20. The default Threshold value (0) sharpens all pixels in the image.
If applying Unsharp Mask makes already bright colors appear overly saturated, choose Edit > Fade Unsharp Mask and choose Luminosity from the Mode menu.

Sharpen selectively

You can sharpen parts of your image by using a mask or a selection to prevent sharpening in certain parts of your image. For example, you can use an edge mask with the Unsharp Mask filter on a portrait to sharpen the eyes, mouth, nose, and outline of the head, but not the texture of the skin.


Using an edge mask to apply the Unsharp Mask only to specific features in an image

Sharpen a selection

  1. With the image layer selected in the Layers panel, draw a selection.
  2. Choose Filter > Sharpen > Unsharp Mask. Adjust the options and click OK.
    Only the selection is sharpened, leaving the rest of the image untouched.

Sharpen an image using an edge mask

  1. Create a mask to apply sharpening selectively. There are many ways to create an edge mask. Use your favorite method, or try this one:
    • Open the Channels panel and select the channel that displays the grayscale image with the greatest contrast in the document window. Often, this is the green or the red channel.


    Selecting a channel with the greatest contrast
    • Duplicate the selected channel.
    • With the duplicate channel selected, choose Filter > Stylize > Find Edges.
    • Choose Image > Adjustments > Invert to invert the image.


    Find Edges filter applied and image inverted
    • With the inverted image still selected, choose Filter > Other > Maximum. Set the radius to a low number and click OK to thicken the edges and randomize the pixels.
    • Choose Filter > Noise > Median. Set the radius to a low number and click OK. This averages the neighboring pixels.
    • Choose Image > Adjustment > Levels and set the black point high to get rid of random pixels. If necessary, you can also paint with black to retouch the final edge mask.


    Setting the black point high in Levels to eliminate random pixels in the edge mask
    • Choose Filter > Blur > Gaussian Blur to feather the edges.
    Note:
    The Maximum, the Median, and the Gaussian Blur filters soften the edge mask so that the sharpening effects blend better in the final image. Although all three filters are used in this procedure, you can experiment using only one or two.
  2. In the Channels panel, Ctrl-click (Windows) or Command-click (Mac OS) the duplicate channel to make the edge mask a selection.
  3. In the Layers panel, select the image layer. Make sure the selection is still visible on the image.
  4. Choose Select > Inverse.
  5. With the selection active on the image layer, choose Filter > Sharpen > Unsharp Mask. Set the desired options and click OK.
    To view your results, select the RGB channel in the Channels panel and deselect the selection in the image.
    You can create an action to conveniently apply all the steps in the procedure.

Add lens blur

Adds blur to an image to give the effect of a narrower depth of field so that some objects in the image stay in focus and others areas are blurred. You can use a simple selection to determine which areas are blurred, or you can provide a separate alpha channel depth map to describe exactly how you want the blur added.
The Lens Blur filter uses the depth map to determine the position of pixels in an image. With a depth map selected, you can also use the crosshair cursor to set the starting point of a given blur. You can use alpha channels and layer masks to create depth maps; black areas in an alpha channel are treated as though they’re at the front of the photo, and white areas are treated as if they’re far in the distance.
For a gradual blurring effect (none at the bottom to maximum at the top), create a new alpha channel and apply a gradient so that the channel is white at the top of the image and black at the bottom. Then select the Lens Blur filter and choose the alpha channel from the Source pop‑up menu. To change the direction of the gradient, select the Invert check box.
The way the blur appears depends on the iris shape you choose. The number of blades determines the iris shape. You can change blades of an iris by curving them (making them more circular) or rotating them. You can also reduce or magnify the preview by clicking the minus button or the plus button.
  1. Choose Filter > Blur > Lens Blur.
  2. For Preview, Choose Faster to generate quicker previews. Choose More Accurate to view the final version of the image. More Accurate previews take longer to generate.
  3. For Depth Map, choose a source (if you have one) from the Source pop‑up menu. Drag the Blur Focal Distance slider to set the depth at which pixels are in focus. For example, if you set focal distance to 100, pixels at 1 and at 255 are completely blurred, and pixels closer to 100 are blurred less. If you click in the preview image, the Blur Focal Distance slider changes to reflect the clicked location and brings the depth of the clicked location into focus.
  4. To invert the selection or alpha channel you’re using as the depth map source, select Invert.
  5. Choose an iris from the Shape pop‑up menu. If you wish, drag the Blade Curvature slider to smooth the edges of the iris, or drag the Rotation slider to rotate it. To add more blur, drag the Radius slider.
  6. For Specular Highlight, drag the Threshold slider to select a brightness cutoff; all pixels brighter than the cutoff value are treated as specular highlights. To increase the brightness of the highlights, drag the Brightness slider.
  7. To add noise to an image, choose Uniform or Gaussian. To add noise without affecting color. choose Monochromatic. Drag the Amount slider to increase or decrease noise.
    Blurring removes film grain and noise from the original image. To make the image look realistic and unretouched, you can return some of the removed noise to the image.
  8. Click OK to apply the changes to your image.

Blur image areas

The Blur tool softens hard edges or reduces detail in an image. The more you paint over an area with the tool, the blurrier it becomes.
  1. Select the Blur tool .
  2. Do the following in the options bar:
    • Choose a brush tip and set options for the blending mode and strength in the options bar.
    • Select Sample All Layers in the options bar to blur using data from all visible layers. Deselect this option and the tool uses data from only the active layer.
  3. Drag over the part of the image you want to blur.

Sharpen image areas

The Sharpen tool increases contrast along edges to increase apparent sharpness. The more you paint over an area with the tool, the more sharpening increases.
  1. Select the Sharpen tool . (If the tool isn’t visible, hold down the Blur  tool.)
  2. Do the following in the options bar:
    • Choose a brush tip and set options for the blending mode and strength.
    • Select Sample All Layers to sharpen using data from all visible layers. If this is deselected, the tool uses data from only the active layer.
    • Select Protect Detail to enhance details and minimize pixelated artifacts. Deselect this option if you want to produce more exaggerated sharpening effects.
  3. Drag over the part of the image you want to sharpen.      

    .....................................................................X 0 X.....................................................................   

Wednesday, May 8, 2013

Printing Images to a Commercial Printing Press


Photoshop / Preparing images for press

    Preparing images for press
    Set output options
    Print separations from Photoshop
    Prepare an image with spot channels for printing from another application
    Create a color trap
    Determine scan resolution for printing


From Photoshop, you can prepare image files for offset lithography, digital printing, gravure, and other commercial printing processes.
Generally, your workflow depends on the capabilities of the prepress facility. Before you begin a workflow for commercial printing, contact the prepress staff to learn their requirements. For example, they may not want you to convert to CMYK at any point because they may need to use prepress-specific settings. Here are some possible scenarios for preparing your image files to achieve predictable printing results:
  • Work entirely in RGB mode and make sure that the image file is tagged with the RGB working space profile. If your printer or prepress staff use a color management system, they should be able to use your file’s profile to make an accurate conversion to CMYK before producing the film and printing plates.
  • Work in RGB mode until you finish editing your image. Then convert the image to CMYK mode and make any additional color and tonal adjustments. Especially check the highlights and shadows of the image. Use Levels, Curves, or Hue/Saturation adjustment layers to make corrections. These adjustments should be very minor. Flatten the file if necessary, then send the CMYK file to the professional printer.
  • Place your RGB or CMYK image in Adobe InDesign or Adobe Illustrator. In general, most images printed on a commercial press are not printed directly from Photoshop but from a page-layout program like Adobe InDesign or an illustration program like Adobe Illustrator. For more information on importing Photoshop files into Adobe InDesign or Adobe Illustrator, see Adobe InDesign Help or the Adobe Illustrator Help.
    Here are a few issues to keep in mind when you work on an image intended for commercial printing:
  • If you know the characteristics of the press, you can specify the highlight and shadow output to preserve certain details.
  • If you use a desktop printer to preview the appearance of the final printed piece, keep in mind that a desktop printer cannot faithfully replicate the output of a commercial printing press. A professional color proof gives a more accurate preview of the final printed piece.
  • If you have a profile from a commercial press, you can choose it with the Proof Setup command and then view a soft proof using the Proof Colors command. Use this method to preview the final printed piece on your monitor.
    Note: Some printers may prefer to receive your documents in PDF format, especially if the documents need to conform to PDF/X standards. See Save in Photoshop PDF format.

Set output options

If you are preparing your images for commercial printing directly from Photoshop, you can select and preview a variety of page marks and other output options using the Print command. Generally, these output options should be specified only by prepress professionals or people knowledgeable about the commercial printing process.



Page marks
A.Gradient tint bar
B.Label
C.Registration marks
D.Progressive color bar
E.Corner crop mark
F.Center crop mark
G.Description
H.Star target
  1. Choose File > Print.
  2. Choose Output from the pop‑up menu.
  3. Set one or more of the following options:
    Calibration Bars
    Prints an 11‑step grayscale, a transition in density from 0 to 100% in 10% increments. With a CMYK color separation, a gradient tint bar is printed to the left of each CMYK plate, and a progressive color bar to the right.
    Note: Calibration bars, registration marks, crop marks, and labels are printed only if the paper is larger than the printed image.
    Registration Marks
    Prints registration marks on the image (including bull’s-eyes and star targets). These marks are used primarily for aligning color separations on PostScript printers.
    Corner Crop Marks
    Prints crop marks where the page is to be trimmed. You can print crop marks at the corners. On PostScript printers, selecting this option will also print star targets.
    Center Crop Marks
    Prints crop marks where the page is to be trimmed. You can print crop marks at the center of each edge.
    Description
    Prints any description text entered in the File Infodialog box, up to about 300 characters. Description text is always printed in 9‑point Helvetica plain type.
    Labels
    Prints the file name above the image. If printing separations, the separation name is printed as part of the label.
    Emulsion Down
    Makes type readable when the emulsion is down—that is, when the photosensitive layer on a piece of film or photographic paper is facing away from you. Normally, images printed on paper are printed with emulsion up, with type readable when the photosensitive layer faces you. Images printed on film are often printed with emulsion down.
    Negative
    Prints an inverted version of the entire output, including all masks and any background color. Unlike the Invert command in the Image menu, the Negative option converts the output, not the on‑screen image, to a negative. If you print separations directly to film, you probably want a negative, although in many countries film positives are common. Check with your print shop to determine which is required. To determine the emulsion side, examine the film under a bright light after it has been developed. The dull side is the emulsion; the shiny side is the base. Check whether your print shop requires film with positive emulsion up, negative emulsion up, positive emulsion down, or negative emulsion down.
    Background
    Selects a background color to be printed on the page outside the image area. For example, a black or colored background may be desirable for slides printed to a film recorder. To use this option, click Background, and then select a color from the Color Picker. This is a printing option only; it does not affect the image itself.
    Border
    Prints a black border around an image. Type a number and choose a unit value to specify the width of the border.
    Bleed
    Prints crop marks inside rather than outside the image. Use this option to trim the image within the graphic. Type a number and choose a unit value to specify the width of the bleed.
    Interpolation
    Reduces the jagged appearance of a low-resolution image by automatically resampling up while printing (on PostScript printers). Resampling may reduce the sharpness of the image quality.
For information on the Include Vector Data option, see Print vector data.

Print separations from Photoshop

When preparing your image for prepress and working with CMYK images or images with spot colors, you can print each color channel as a separate page.
Note: Separations from CMYK, Duotone, or multi-channel documents printed on non-PostScript printers may not be identical to those printed on PostScript printers.


Each color channel printed as a separate page.
Note: If you are printing an image from another application and want to print spot channels to spot color plates, you must first save the file in DCS 2.0 format. DCS 2.0 preserves spot channels. This format is supported by applications such as Adobe InDesign and QuarkXPress.
  1. Make sure that your document is in CMYK Color, Multichannel, or Duotone mode, and then choose File > Print.
  2. Choose Separations from the Color Handling drop-down menu.
    Note: Depending on the designated printer and printer drivers on your computer, these options may also appear in the Print Settings dialog box. In Windows, click the Properties button to access the printer driver options; in Mac OS, use the pop-up menu in the Print Settings dialog box that appears.
  3. Click Print. Separations are printed for each of the colors in the image.

Prepare an image with spot channels for printing from another application

  1. If the image is a duotone, convert it to Multichannel color mode.
  2. Save the image in DCS 2.0 format.
  3. In the DCS 2.0 Format dialog box, deselect the Include Halftone Screen and the Include Transfer Function options.
  4. Open or import the image in Photoshop, and set your screen angles. Make sure that you’ve communicated to the printer the spot color you want for each of the color plates.
    Note: You can place a PSD file containing spot colors directly in Illustrator or InDesign without special preparation.

Create a color trap

A trap is an overlap that prevents tiny gaps in the printed image caused by a slight misregistration on press. Contact your service provider before you do any trapping. In most cases, your print shop determines whether trapping is needed. If so, the print shop staff will tell you what values to enter in the Trap dialog box.


Trapping to correct misalignment
A.Misregistration with no trap
B.Misregistration with trap
Trapping is intended to correct the misalignment of solid colors. In general, you don’t need traps for continuous-tone images such as photographs. Excessive trapping may produce an outline effect. These problems may not be visible on-screen and might show up only in print. Photoshop uses standard rules for trapping:
  • All colors spread under black.
  • Lighter colors spread under darker colors.
  • Yellow spreads under cyan, magenta, and black.
  • Pure cyan and pure magenta spread under each other equally.
  1. Choose Edit > Convert To Profile to view the image in the color space of the output device. See Convert document colors to another profile (Photoshop).
  2. Choose Image > Trap.
  3. For Width, enter the trapping value provided by your print shop. Then select a unit of measurement, and click OK. Consult your print shop to determine how much misregistration to expect.

Determine scan resolution for printing

You can use a number of techniques to determine the resolution at which to scan a photograph. If you are scanning photos to be printed, and you know the exact size and the printing screen frequency, you can use the following techniques to determine the scanning resolution. Often it is easiest to scan at your scanner's maximum optical resolution and then resize the image in Photoshop later.

Estimate scan resolution

You can determine the resolution for your scan using the original and final image dimensions and the resolution of your output device. Scan resolution translates into image resolution when you open the scanned image in Photoshop.
  1. Do one of the following:
    • For laser printers and imagesetters, multiply the printer’s screen frequency by 2. To determine your printer’s screen frequency, check your printer documentation or consult your service provider.
    • For inkjet printers, check your printer documentation for the optimal resolution. Many dye sublimation printers and devices that print directly onto photographic paper have an optimal resolution of 300 to 400 dpi.
  2. Determine the ratio of the final image dimensions to the original image dimensions. For example, the ratio of a 6‑by‑9‑inch final image to a 2‑by‑3‑inch original image is 3:1.
  3. Multiply the result of step 1 by the result of step 2. For example, suppose you are printing to an imagesetter with a screen frequency of 85 lpi and the ratio of the final image to the original is 3:1. First multiply 85 (the screen frequency) by 2 to get 170. Then multiply 170 by 3 to get a scan resolution of 510 ppi. If you are printing to an inkjet printer with an optimal resolution of 300 dpi, multiply 300 by 3 to get a scan resolution of 900.
    Note: Different color separation procedures might require different ratios of image resolution to screen frequency. It’s a good idea to check with your service provider or print shop before you scan the image.

Calculate the file size before scanning an image

You can create a dummy file to predict the file size needed for the final output of your scan.
  1. In Photoshop, choose File > New.
  2. Enter the width, height, and resolution of your final printed image. The resolution should be 1.5 to 2 times the screen frequency you will use to print. Make sure that the mode you plan to scan in is selected. The New dialog box displays the file size. For example, suppose you want the final image to be 4 inches wide and 5 inches high. You plan to print it with a 150‑line screen using a 2:1 ratio, so you set the resolution to 300. The resulting file size is 5.15 MB.
    To produce the scan, enter the resulting file size in your scanner settings. Don’t worry about resolution or image dimensions. After you have scanned the image and imported it into Photoshop, use the Image Size command (with the Resample Image option deselected) to enter the correct width and height for the image.
More Help topics

Tuesday, March 19, 2013

The Special Visual Effects of Willow


The Special Visual Effects of Willow
by David Hutchinson

It's a bright, sunny day up at George Lucas' San Rafael effects facility, Industrial Light & Magic, but ILM matte artist Chris Evans invites me indoors.  "Come inside," he says, "and watch us make it rain."  One of ILM's large stages has been draped in black Duvetyne and vast sheets of black plastic cover the floor.  Far overhead, water sprinklers are cascading sheets of "rain" onto the stage floor, where a crew waits with brooms and squeegees to push the water into floor drains.
Chris Evans stretches out on the floor to peer through the VistaVision viewfinder on one of ILM's cameras.  The lens points up into the rain at a steep angle.
Evans sees me shielding my eyes from the bright lights, which are shining directly toward us through the spray of rain.  "Backlighting makes the droplets visible against the background," he explains.  "This rain element will be composited with a matte painting of the Nockmaar castle.  It's a very dramatic shot, since the camera will appear to tilt up the castle walls into the rain."
Several crew members approach Evans with suggestions about placing the wind machines so that the rain appears to move across the stage in sheets, just the way a real rainstorm does on a windy day.  It is standard procedure at ILM, and one of the secrets of their success, to invite comment and suggestions from everyone on the crew.
As the stage technicians set up for another take, Evans walks me back over to ILM's main building.  The matte paint-ing department is located upstairs in the building's rear.  Evans' associates are hard at work painting scenes on glass panels which are approximately two feet by six feet.  Some paintings have blank areas in which a live action shot (called a "plate") will be inserted.  For example, one painting is a vast panorama of a landscape; a small circle in the center remains clear unpainted glass.  "A live action plate of the Nelwyn village, which George shot in Wales, will be rear-projected into that painting's center," Evans announces.  Other paintings are complete themselves and are used to establish locales or serve as transition between scenes.
The studio walls are lined with storyboard sketches and landscape photographs, frame blow-ups from the miles of film shot by the location unit for the matte department.  A large organizational chart detailing the progress of each shot in the department takes up another wall.
"Last summer," Evans continues, "before the main production units were on location in Wales and New Zealand shooting principal photography, we discussed the kind of shots we were looking for.  But, the basic idea was that they would simply go out and just shoot tons and tons of plates, not knowing what great shots there might be in the making.  The idea was not to lock ourselves into predetermined designs, which would limit our vision.  What we didn't get on location we could re-stage back here, which we are doing a great deal.
"All the plates arrived in September, and we went through them with George, who had a very rough assemblage of the picture.  As we ran through the film, he would say that he wanted a matte shot here and one there.  'This is a plate we got in New Zealand, I want you to add this, and this is something we didn't get so I want you to create the entire thing from scratch locally.'  Essentially, we got a whole shopping list from George, but he let us design the shots.
"For example," Evans notes, "the castles were never fully designed, especially the Nockmaar one.  The location unit in Wales built a section of it and a drawbridge, but when it came to what the entire castle and surrounding landscape actually looks like from different angles, then we got to design it, which is really very nice.  It would have been nicer, if we had designed it before they went to England to shott the plates.  If someone had some idea of exactly how big this thing was at Tir Asleen, it would have been a little simpler."
It's January as I talk with Evans, about four months before Willow opens in theaters nationwide, and about two-and-a-half months before ILM must complete hundreds of individual special effects shots.  "We have about 30 matte shots to do," Evans says.  "On Return of the Jedi, we had 44.  As supervisor, I am personally completing three paintings.  Our other artists are Mike Pangrazio, Sean Joyce, Caroleen Green and Paul Paul Sen, who is new, though he did work on Tucker.  Craig Barron is supervising all the photography involving the matte shots, his assistants are Wade Childress and Randy Johnson.  Paul Houston, who is the film's chief model maker, has done a tremendous amount of work for us; we have a number of shots which combine miniatures with matte paintings.
"Mike Pangrazio is working on a painting now that will be combined with miniatures.  It's a big establishing shot of the Nockmaar castle.  We have some miniature towers that will have banners and flags blowing in the wind.  It will add life and movement to an otherwise static shot.  The model allows us to move the camera slightly during the shot and get a sense of changing perspective.  Also, the model can be lit to dive us a depth and quality of texture.  It adds enormously to the painting's reality."
Some of the storyboard drawings look as complicated as they are spectacular.  I point to one showing the vast maze around Tir Asleen.  "The idea for this one is to start with the camera very tight on the plate, which is surrounded by a painting and a miniature.  The painting is actually a bas relief.  It's flat at the point where it joins the plate and then it immediately goes off into three dimensions.  So, when we do a big pull back and tilt, the audience will be aware of parallax shifts in the foreground elements, and it will look very three-dimensional.  Essentially, we are rear-projecting a plate into a multi-plane miniature.  It should give the feeling of a live-action helicopter shot."
In Evans' corner of the studio rests a large easel with a nearly completed panorama of the Nelwyn Valley.  "This is the last shot in the movie," he says.  "George wants it to be the landscape equivalent to the warehouse at the end of Raiders of the Lost Ark where you're in tight and you gradually pull back and face out and the titles come up.  Willow has returned, he has been reunited with his family and to show that peace and happiness has come back to the land again, we do this big pull back.  Actually, there are two plates which will be rear-projected into this paining: one of the village and another of a rushing river.  Then, we'll add some moving clouds and some rays of light."
The scene evokes a very 19th century romantic style.  "The Nelwyns could sell this shot as a tourist postcard," I joke.
Evans smiles.  "That's what George wants--very artistic panoramic landscapes in the style of the Hudson River school.  George wants a fairy tale quality, so the landscape has a misty, golden glowing look to it."
Pointing out the other paintings in progress, Evans tours the room.  "Sean Joyce is working on a painting of Bavmorda's throne room.  It's a latent image shot, which means the scene is composited entirely in-camera, giving us first generation quality.  It's a sunset shot.  This will be a bipack element and there will be another bipack element of trees.  Behind that will be a sunset, which we will either paint or photograph.
Evans then singles out "another full painting with a bipack element of just these guys walking across a log, everything else will be painted.  Mike Pangrazio is starting a snow camp shot, which is going to be very beautiful.  It's another latent image composite, this time with a tent--a miniature tent with lights against a moonlit landscape.  Sean Joyce will be doing the fairy forest shot with the gigantic tree trunk; we will probably compose that shot in front projection.  We can set up a miniature tree trunk and flicker the light against the scene.  In front projection, you can shoot it and see it through the camera all at once.
"All in all, we designed 40 shots, but George cut out 10.  Only 20% are plates that were actually shot on location in England or New Zealand; the rest we have shot here on the stage or locally.  So, about 40 plates were shot for matte paintings and about fiver were used.  But there are many other special effects plates that were photographed on location."
"When do you hope to finish all this?" I wonder.
"Actually, we are fairly well ahead of the game," Evans smiles confidently.  "We should be wrapping by march and finals are due in April.  So even if we go a month over schedule, we're still two weeks ahead."
In his quest to populate a Nelwyn village, George Lucas put out a casting call all over Europe to gather together more than 200 little people, including star Warwick Davis, who stands 3' 4" tall, and Billy Barty who towers over the youngster at 3' 9".  But, where would George Lucas find even one actor capable of playing brownie--an actor only 9 inches tall?
The solution turns out to ge quite easy: Hire Rick Overton, 6' 4" and Kevin Pollak, 5' 10" and let ILM handle the rest.  Of course, this means that director Ron Howard found himself directing scenes on location with imaginary brownies.  Young Warwick Davis had to pretend the brownies were in a given shot and depend upon ILM to shrink full-size actors down and insert them into the already lensed scenes.
Mike McAlister, an eight-year veteran of ILM, was attached to the main production unit during five months, overseeing the shooting of more than 700 special effects.  "I came onto Willow," McAlister reveals, "about the beginning of April '87, just before production started in England.  I was sent to England to supervise all the background plates photographed over there for about five months.  Dennis Muren and Phil Tippett each came over for a couple of weeks while the bulk of their stuff was being shot.  England was great for about the first three months, but the I began getting homesick for some good french fries.
"After we got back to America, Ron and George were busy putting together a rough cut of the film.  Originally, there were going to be many more brownie shots, but by the time they got finished assembling the film I was down to about 70 shots."
Then, it was time to call in Rick Overton and Kevin Pollak to start performing their roles.  The luckless brownies had an even more difficult challenge than Willow Ufgood or Madmartigan, because Overton and Pollak (who play Franjean and Rool) had only a giant blank blue stage on which to work.  Their entire performance would have to be created solely with their imaginations.  Of course, bout George Lucas and Ron Howard were on hand to direct the pair through some difficult scenes, though a large number of the scenes fell to Mike McAlister alone, and a few others to effects supervisor Dennis Muren.
"Rehearsals for the brownie sequences largely consisted of sitting through much of the dailies," says Overton.  "We discovered that there was a good bit of leeway in each plate so that we could develop little character bits and gags.  Ron was very clever, in the sense that he shot plates with different timings.  Other sequences were very difficult because they had to be very precise.  it was explained to us on the set that Willow sets a new record for matte work with a film using actors.  That is, live people, not miniatures matted into shots and working in the scene with a previously photographed cast.  We have done more matte work of this type than any other film that they can recall in history.  And we were there!  For five weeks, sometimes six days a week, all day, and those are long days when it's just Kevin and I facing a camera.  Of course, next time they might just dispense with all the rehearsal, get out a motion control rig, and stick a blue pylon up my wazoo."
They did all their work "on one of the largest blue screen stages ever built, too--about 70 by 70 feet," Overton recalls.  "They spent a great deal of time getting the lighting just right.  Mike McAllister said that one of the biggest problems in blue screen work is getting the lighting to match.  Even the hair style designed for us was a problem.  Because our hair was so frizzy, the early tests showed grainy lines, you know, just like you see on those early TV episodes of Star Trek when the Enterpriseis going around a planet and the ship's pylons disappear.  Finally, they changed our hair to a very spikey look and the mattes worked great.  The poor hair was just spraying and pinching, twisting and twirling--I though we might look like punk elves.  It says in the script, 'angry little haircuts,' and that's just what we wound up with.
"The first week was spent in rehearsal, working out character and roughing out timings with the plates.  They would project the plates for us, which had been marked in very roughly with little dots showing us where we should be.  Also, we were able to rough out the timing for the dialogue.
"George was very open-minded about our suggestions for character bits.  They really trusted us.  It was a very nice working situation.  They knew exactly what they wanted and how to get it.  Usually their directions were so good that you could get it on the first take--unless the scene was just something so complex you had to train your legs to walk the path of taped X's on the floor and watch for their hand signals for the dialogue.  That would take a few runs to rehearse.
"For a few sequences, we worked on oversize sets," he confesses.  "For example, when we have to pick the lock to Madmartigan's cage or the big wagon chase scene in the escape from Sorsha's camp, the sequence uses both blue screen shots and oversize sets.  They made a big wagon, which they rocked and bounced around; Kevin and I were able to do all these great slapstick bits with big apples rolling by and giant kernels of corn.  We're inside the wagon with the soldiers stomping around and then you curt to outside, in which we will be blue screen matted, and then back again to us on the oversize set.  Anyway, the wagon is being pulled by horses at a gallop and we're bouncing around, Rool [Pollak] is getting nauseous and falls off the wagon's side and I pull him back in.  On the full-size set, Kevin's hanging 25 feet off the ground, practically naked.  There is one take in which he is haning in there screaming, 'Help me!  Help me!"  And I'm just saying, 'What?  I can't hear you.  What?'
"The cutting back and forth between blue screen and full size helps to hide the effects technique," Overton explains.  "When you're looking straight down at Rool, for example, when he is hanging and swinging, ILM will be matting in the terrain whizzing by underneath, and then, back in the wagon, those are the straight oversize set shots."
In order to match eyelines with the characters in a plate," Overton says, "the directors watched us on the video monitor, telling us to spot on a light or a particular crossbeam in the grid overhead or one of the grips, whatever.  We had to time our lines to answers, so they would point at us to cue us for the next line.  Or they would say the line and then leave a little leeway so there would be no overlap on the soundtrack when we did our lines.  As it turned out, the sound quality was so echoey on that vast blue screen stage that we will have to re-record all of our dialogue.  We'll be looping our little brownie butts off--looping hell for a week.
"Ron Howard has the most incredible team of people up there at ILM; they'll do anything for him, bend over backwards and follow him to the end of the Earth.  Ron was born to be in the business; endless experience.  At one point, I grumbled a little during the big battle cry charge.  We were on gravel that was supposed to be the beach in front of the Nockmaar castle.  It was cold, and we were barefoot and the gravel hurt our feet.  Ron cheered us up by telling us about a take he did forHuckleberry Finn (1975) when he was Tom Sawyer; they made him run on heavy, sharp driveway gravel, barefoot and it was freeing out.  'OK!' I said.  'You win!'  The man has been doing this for years; what do I have to complain about?  I enjoyed working with him on Gung Ho, and Willow was a chance to repeat that experience."
The moving camera shots with the brownies are some of the most interesting blue screen shots ever done.  "We had a shot," Mike McAlister says, "in which a brownie was running along the top of a log that was three feet off the ground.  Then, he jumps off the log, lands on the ground, and tumbles.  I had to figure out how to photograph that.
"First of all, a brownie is supposed to be about 10 inches tall, and, say, the top of the log was 30 inches off the ground.  That means the log is three times as tall as the brownie.  But we had a six-foot brownie.  So, if we were to build a set that was to literally put the human through what the brownie is supposed to go through, the log would have to be three times higher than the human actor, which is 18 feet.  At that point, we would have been asking a human actor to run along the top of something that was 18 feet from the ground and then jump off and break all his arms and legs landing on the ground.
"Well, that's what mathematics tells you to do, but the reality of it is that even though we're dealing with a six-foot actor, he still only has to jump 30 inches.  The reason is simply that the laws of physics tell us that there's a certain amount of time for any object to fall 30 inches, and it is the same time for a brownie or a human.  Didn't Galileo prove that?" MacAlister smiles.
"It takes maybe four frames--a sixth of a second to fall 30 inches.  If our actor had jumped form 18 feet, it would have taken several times longer.  So, even though the actor was falling through a distance of only half his height; he was still airborn for the same length of time as a 10-inch brownie.  So, what we do is, he jumps off a 30-inch high platform and he's in the air for four or five frames.  Then, we use a process which we cal lpin blocking in which we re-photograph the film on an animation stand positioning the actor where he needs to be, so that it actually looks like the human is falling three times his height.
"Also, there is another big dolly shot, in which the brownies are running along the top of a log.  These are showcase shots in which we can really make the brownies look believable.  It's one of the most important jobs we have here at ILM, and that is to make these characters believably 10 inches high.  It's our job to establish their credibility.
"I had the greatest time working with Rick Overton and Kevin Pollak," admits McAlister.  "You have to remember I had spent months and months trying to place them into scenes that Ron was directing and saying, 'Well, what's that brownie going to do?'  Then, finally, seeing them perform and bring life and personality to their characters was a real thrill.  I guess all the anticipation built up in me and when I finally saw the brownies, they were even better that I imagined they could be."
Work on the battle at Tir Asleen with the two-headed Eborsisk as well as Bavmorda's terrible phantasmagoria of magic in her climactic struggle with Raziel has only begun at this writing.  Phil Tippett is bringing his go-motion expertise to bear with the fire-breathing Eborsisk and Dennis Muren is supervising the great duel of magic in Bavmorda's castle tower.
Stop motion is being used to bring mechanical objects to life and rod puppetry for a chair.  "The slightly mechanical look of stop motion is fine for animating mecahnical objects," says Dennis Muren, "but we'll try to introduce a little blur just to blend it in.  One of the things which are still working out at this late date is the look of Cherlindrea, the queen of the fairies.
"Transformations, too; such as the character of Raziel, who is variously a goat, a raven and a sort of otter.  Most of these transformations were hadled on the set, they are not an optical effect, except perhaps for Raziel's fist appearance.  It was an otter-like puppet in England, but George was never satisfied with her appearance, so a real animal was used.
"The other shot that is coming up is a very elaborate transformation of Raziel the goat into Raziel the woman.  It involves a great deal of image processing," Muren reports.  "you are really going to see something that you haven't seen before.  It won't look like a lot of bladders going and things being pushed out.  We're using our laser scanner to scan out frames of film, manipulate the image and scan it back in again.  It will be very interesting."
Confides Dennis Muren, "I really hope Willow works out.  George is playing it very straight, and I hope the audience is going to accept it that straight.  If you reallybelieve that there are all these little characters and this other world, then the film will seem credible.  And I really think you will believe you are there in the world ofWillow."
Hutchinson, David. "The Special Visual Effects of Willow."  The Official Willow Movie Magazine.  Ed. David McDonnell.  New York: Jacobs, 1988. 58-66.

Content Source : http://accad.osu.edu/~waynec/history/externalpages/index.html

Followers