BUL Computer Tips

BUL Computer Tips for Graphics

Questions or suggestions about this page should be sent to mspisar@flip.eecs.umich.edu

Question index:

How do I record an animation onto videotape?
How do I scan an image?
How do I use the Microtek ScanMaker E3 under Linux?
How can I use Matlab to create a sequence of image files?
How do I create movies in the format ...
How do I strip frames from a movie in the format ...
- Animated GIF?
- FLI?
How do I play movies in the format ...
What is the advantage of using the format ...

Q: How do I record an animation onto videotape?
A: The first step is to create FLI animations from your images. Refer to "How do I create movies in the FLI format?" for help on creating such movies. When creating these movies, keep the number of colors used below 230, because Windows uses up 20 of the colors in the colormap, and the animation player program needs a few extra colors as well, so not all 256 colors will be available during the playback. Also, make sure all of your images use the same colormap, and use the '-m' option in fbm2fli to use a single colormap for the entire movie. If each frame has it's own colormap, the movie playback will flicker as the colormap changes.

After you have created the movies, transfer them to the c:\temp directory on Uticaclub. The following instructions assume you are working with Uticaclub under Windows95 in 800x600 256-color mode. High-color and True-color modes will also work as long as you do not select the fade-in/out options (in the higher color modes, the fade options will cause the entire movie to be black.)

Run the "Autodesk Animation Player," listed under the Utilities program menu.
Select the menu option in this program to create a new script.
Set the first image in the script to be c:\util\waaplay\black.bmp. This is a 640x480 black image. The settings for this image should be:
- Use full screen
- Color cycling OK
- Use all colors
- Loops: 1
- Pause at end: 10 (set this number to the number of pre-black seconds)
- Transition: Cut

Set the second image in the script to be your FLI movie. The settings for this movie should be:

Load into memory
Use full screen
Color cycling ok
Use all colors
Loops: 1
Pause at end: 1 (This is the amount of time spent on the last frame. If you want the fade-out to occur after the movie has complete, make sure this number is larger than the fade time.)
Transitions: fade from/to black in 0.5 seconds
If you have additional movies, repeat steps 3 and 4 as necessary.
Set the last image in the script to be c:\util\waaplay\black.bmp. The settings for this image should be:
- Use full screen
- Loops: 1
- Pause at end: 30 (set this number to the number of post-black seconds)
- Transition: Cut
After you have defined the script, press the ">>" menu item to test the movie.
Once you have things set up the way you like, connect the S-video connector from the VCR to the S-video output on the computer. The S-video output is on the same card as the monitor connector. Be sure to connect the connector to the S-video output on the video card, and not to the S-video input which is on the card next to the video card.. Turn on the VCR and TV monitor. Then select the Display control panel, select the Settings tab, and click on the "Television" check box. If this box is grayed, it is because you either do not have the S-video connector plugged into the video output, or the display resolution is set higher than 800x600. Fix both of these, and then restart the Display control panel.
At this point the Windows display should appear on the TV monitor. Return to the Autodesk Animation Player, click on the ">>" menu item and move the cursor to the bottom right corner of the display so that it is not visible. Then press the play and record buttons on the VCR to start the recording.
When you are finished with the recording, disconnect the S-video connector from the computer's S-video out connector. If this remains plugged in, the computer monitor will fail to sync properly when running Linux.
Q: How do I scan an image?
A: If you need to use a Linux machine, see the next section. Otherwise, on uticaclub (PC in the computer lab), use PhotoImpact (under Programs).
Q: How do I use the Microtek ScanMaker E3 under Linux?
A: There is a program called "mtekscan" which will scan an image from the scanner, and output the results to a file, or to STDOUT which can be piped into xv. Using the '-h' switch with this program gives the following help information:
```
mtekscan: User level driver for Microtek scanners v0.1.
  Copyright (c) 1996 by Jan Schoenepauck / Fast Forward Productions

Usage: mtekscan [options] 

Valid options:
  -a      halftone scan                   -s <n>  shadow adjustment      
  -b      line art (B/W) scan             -l <n>  highlight adjustment   
  -c      color scan                      -m <n>  midtone adjustment     
  -g      grayscale scan                                                 
  -d <n>  digital brightness adjustment   -t      transparency scanning  
  -k <n>  contrast adjustment             -p      prescan mode           
  -e <n>  exposure time                   -v <n>  scanning velocity      
  -H <n>  halftone pattern                -L <n>  set paper length       
  -n      reverse colors (negative)       -V      verbose mode           
  -r <n>  scanning resolution             -B      disable backtracking   
  -h      this help screen                -C      disable calibration    
  -i      scanner information (short)     -P      same as -p -C          
  -I      scanner information (long)                                     
                                                                         
  -f <x1> <y1> <x2> <y2>  scanning frame coordinates                     
  -o <file>               write to <file> instead of stdout              
```
For example, to scan an 8.5" by 11" grayscale image based at (0",0") with a 150 dpi resolution, sending the output to xv for processing and/or format conversion:
```
mtekscan -g -r 150 -f 0 0 8.499 11 | xv -
```
To scan a 5" by 3" color image based at (2",1") with a 300 dpi resolution, sending the output to xv for processing and/or format conversion:
```
mtekscan -c -r 300 -f 2 1 7 4 | xv -
```
To scan an 8.5" by 11" lineart (i.e. black/white) image based at (0",0") with a 300 dpi resolution, sending the output to xv for processing and/or format conversion:
```
mtekscan -b -r 300 -f 0 0 8.499 11 | xv -
```
Q: How can I use Matlab to create a sequence of image files?
A: Set up an m-file to create the images that you want inside of a for loop, and then modify it to match the following structure:
```
    % Force this figure to have it's own full colormap
    set(gcf,'ShareColors','off');
    for frame=1:30,
      % Insert your code here to draw the image in the current figure

      % using getframe(gcf) gets the entire figure
      % using getframe(gca) gets just the region inside the axis
      [x,map]=getframe(gcf);
      gifwrite(x,map,sprintf('/tmp/movie/frame%03d.gif',frame));
    end
```
This will create a set of GIF format images in the directory /tmp/movie. The image file names will have the form frame001.gif, where the number will be incremented for each subsequent frame.
The above method has a few drawbacks. One is that the getframe function in matlab takes whatever is on the screen in the figure window, even if that window is overlapped by another. This means that while making movies in this manner, you must be careful not to disturb the figure window. Another drawback is that you may or may not be limited in the colormap you use (although the sharecolors=off option should fix this.)
A better method is to create the image data in a matrix, and then write this matrix to an image file, with no drawing in a figure window. This method can be done in the background without any worries about others disturbing the process. For example:
```
    for frame=1:30,
      % Replace this code with your image creation routines
      % The image must have values from 1 to length(map)
      eval(['load frame' num2str(frame)]);
      im=grayimage(data);
      map=gray(256);dB=40;
      im=(20*log10(im/scale)+dB)*(size(map,1)/dB);

      % Then call this function to write the file
      gifwrite2(im,map,sprintf('frame%03d.gif',frame));
    end
```
Q: How do I create movies in the format ...
A: Select format type below to goto appropriate question
Q: How do I create movies in the animated GIF format?
A: Assuming that you have a directory containing a series of GIF format images with sequential (in an alphabetical sense) file names, you may use the program "gifmerge" to merge them into an animated GIF file. The syntax for this command is:
```
    gifmerge -l0 -25 *.gif > ../anim.gif
```
The "-l0" switch sets the number times the animation will sequence through the set of frames (0 means loop forever). The "-25" sets the delay between frames, in 1/100 of a second. Using the value 25 will result in a display rate of 4 frames per second. Run "gifmerge" without options for more help information, or refer to the web page describing the program.
Q: How do I create movies in the MPEG format?
A: Assuming that you have a directory containing a series of images with sequential file names, you may use the program "mpeg_encode" to turn them into an MPEG file. The syntax for this command is:
```
    mpeg_encode parameters.txt
```
To run this first requires that you set up a file "parameters.txt" which contains the options for the program. The following file may be used as a starting point
```
    PATTERN IBBPBBPBBPBBPB
    IQSCALE 6
    PQSCALE 6
    BQSCALE 6
    OUTPUT /tmp/apr1996.mpeg
    INPUT_DIR /tmp/movie
    INPUT
    frame*.gif [001-100]
    END_INPUT
    BASE_FILE_FORMAT PNM
    INPUT_CONVERT giftopnm *
    GOP_SIZE 100
    SLICES_PER_FRAME 1
    PIXEL FULL
    RANGE 2
    PSEARCH_ALG LOGARITHMIC
    BSEARCH_ALG CROSS2
    REFERENCE_FRAME DECODED
```
The following is a brief description of some key options. A more thorough description of the options specified in this file are described in the program's man page, and at http://bmrc.berkeley.edu/projects/mpeg/mpeg_encode.html.

PATTERN IBBPBBPBBPBBPB
This option selects the order of the frame types output to the MPEG file. The three frame types are I (), B (), and P (). The compressed size of the I frames is generally the largest (least efficient), and the P frames the smallest (most efficient).
[I,P,B]QSCALE 6
This selects the compression for each of the frame types. Lower numbers here will yield higher quality images with less compression.
OUTPUT
This specifies the MPEG file to be created.
INPUT_DIR
This specifies the directory in which the input frames may be found.
INPUT...END_INPUT
The ordered list of files must be between the INPUT and END_INPUT lines. The above example uses a shortcut to specify that the '*' should be replaced sequentially with '001', '002', ... '100'. Other shortcuts exist also; refer to the documentation for more details.
INPUT_CONVERT giftopnm *
This specifies the command used to convert the files stored on the disk into the PNM (or other acceptable types) file format required by the encoder.
Information on the MPEG file format can be found at http://www.crs4.it/HTML/LUIGI/MPEG/mpegfaq.html.
Q: How do I create movies in the Quicktime format?
A: First, create an animated GIF file using the "gifmerge" program. Then, on the Mac, run the "Graphic Converter" program from the Utilities..Multimedia folder. From this program, load in the animated GIF file, and then select Save As..MooVer (Quicktime). (I think this program requires the image width to be a multiple of 8 to prevent getting a while line down the right side of the resulting Quicktime movie.) Select Picture..Speed to set the speed of the movie to a reasonable value. The best compression type for GIF-type images seems to be Graphics--Color.
Q: How do I create movies in the AVI format?
A: First, create an animated GIF, convert this to a Quicktime movie, without using any compression. Then, run the "VfW Converter" program from the Utilities..Multimedia..VfW folder on the Mac. Use either the Microsoft RLE or Microsoft Video compression formats. (If these formats do not appear as options, then copy the "Microsoft Compressors" from the VfW folder into the System Folder and reboot the machine.) The RLE format gives the best results for GIF-type images, while the Video format gives better compression but poorer results.
Q: How do I create movies in the FLI format?
A: Assuming that you have a directory containing a series of GIF format images with sequential (in an alphabetical sense) file names, you may use the program "fbm2fli" to merge them into an FLI file. First, determine the size of your images, in pixels, using a program such as "xv". The syntax for this command is (assuming your GIF files are in the directory /tmp/movie):
```
    ls /tmp/movie/*.gif > /tmp/filelist
    fbm2fli -rx 111 -ry 222 -s 10 /tmp/filelist /tmp/movie.fli
```
The "-rx 111" switch sets the width dimension of the images, in pixels. The "-ry 222" switch sets the height dimension of the images, in pixels. If you do not set these two values, the program will add a black border around your image. The "-s 10" switch sets the default frame rate, in frames per second. Run "fbm2fli" without options for more help information.
How do I strip frames from a movie in the format ...
- Animated GIF?
- FLI?
Q: How do I strip frames from a movie in the Animated GIF format?
A: There is a program called "gifasm" that will perform this function. This program is probably found in the directory /usr/coolstuff/gif. To use this program, first create a directory for the frames, and then run gifasm with the -d option to extract the frames to this directory. The files created will have the frame number (base 0) appended to the file name, and will be in GIF format. For example, to strip the frames from a movie file called movie.gif into the files /tmp/movie/frame???.gif:
```
   mkdir /tmp/movie
   /usr/coolstuff/gif/gifasm -d /tmp/movie/frame movie.gif
```
You may also use the giftopnm program with the -image option to strip individual frames of a GIF movie into pnm-format files. View the man page on giftopnm for more details. For example, to strip the 10th frame from movie.gif into frame010.ppm:
```
   giftopnm -image 10 movie.gif > frame010.ppm
```
Q: How do I strip frames from a movie in the FLI format?
A: There is a program called "unfli" which will perform this function. Run unfli without options to get some help information, or view the man page for more details. For example, to strip the frames from movie.fli into GIF files in the directory /tmp/movie:
```
  mkdir /tmp/movie
  unfli -G movie.fli /tmp/movie/frame gif
```
Q: How do I play movies in the format ...
A: Select format type below to goto appropriate question
Q: How do I play animated GIF movies?
A: Animated GIF movies are supported natively by many web navigator programs, so simply opening the file using, for example, Netscape will play the movie. You may also use the <IMG SRG="anim.gif"> HTML markup to include these animations within a web page.
Alternatively, the "xanim" program allows you to play these animations, control the frame rate, pause, and step frame by frame.
Q: How do I play MPEG movies?
A: The "mpeg_play" program will play these movies in Unix. The "xanim" program will also play them in Unix, but its support is currently somewhat lacking. In Windows, assuming that the proper drivers are installed, the Media Player application will play MPEG movies. On the Mac, just double-click on the file and it will magically play.
Q: How do I play Quicktime movies?
A: The "xanim" program will play these movies in Unix. Media Player in Windows may be able to play this format if the correct drivers are installed. On the Mac, just double-click on the file and it will magically play.
Q: How do I play AVI movies?
A: The "xanim" program will play these movies in Unix. Media Player in Windows will play this format. I don't know about the Mac.
Q: How do I play FLI movies?
A: The "xanim" program will play these movies in Unix. There are several viewer programs that will run on PCs to view these files. Some of them are installed on Lager. I don't know about the Mac.
Q: What is the advantage/disadvantage of using the format ...
A:Select format type below to goto appropriate question
Q: What is the advantage/disadvantage of using the animated GIF format?
A: This format gives fair compression, but not as good as what may be achieved with the other formats. However, since it uses a paletted image, the animation displays quite well even on 256 color displays. Also, since this format is natively supported in web browsers, it is convenient for adding animations to web pages.
Q: What is the advantage/disadvantage of using the MPEG format?
A: This format provides a good compression ratio, but decompresses slowly unless you have some hardware assistance. Also, the dithering used on 256 color displays can reduce the image quality.
Q: What is the advantage/disadvantage of using the Quicktime format?
A: This is the native format on Macs. The image quality and compression ratio is similar to MPEG.
Q: What is the advantage/disadvantage of using the AVI format?
A: This is the native format in Windows. The image quality and compression ratio is similar to MPEG.
Q: What is the advantage/disadvantage of using the FLI format?
A: These files aren't as compressed as the GIF movies, but they will play at faster frame rates.

Back to Internal Use Page