FAQs

Q.    Will VideoPoint run on a server?

A.      Yes.

Q.    How many copies do I need to buy if I'm running on server?

A.   You need to buy as many student copies as you have concurrent uses. So if you have a class of 20 students and 10 computers, you would only need to buy 10 copies since only 10 copies could ever be open at once. There is no formal copy protection; we rely on people's honesty.

return to top

Q.    What is saved in a .vpt file (VideoPoint data file)?

A.   Only the data and a pointer to the movie is saved. If you open the data file and the movie is not in the same place as before, VideoPoint will first look in the same directory of the .vpt file and then will ask you to locate the movie on your own. This allows for students to save analysis of these movies onto floppies without having to copy the movie all the time.

Q.    Is the CD for the Teacher's Kit different than that for the Student Kit?

A.   No. The Teachers Kit is the same CD with a manual. Part of the cost is considered a license to purchase the Student Kits.

Q.    I want to copy parts of the manual. Do I need permission?

A.   Not as long as it is for use in your classroom or for a VideoPoint workshop. If your copy center needs written permission before making the copies, please contact us with their basic requirements.

return to top

Q.    I click on the movie to locate a feature, and nothing happens.

A.

	Click on this toolbar button.
	If the cursor changes to this, clicking on the movie will locate the selected video point (as indicated in the bottom right part of the movie).
	If the cursor remains, all video points have been located on this frame. Advance the movie one frame and try again.

return to top

Q.   The motion in the movie is on an incline.

A.   If the motion that you are analyzing is on an incline, it is possible to move and rotate the origin such that motion on the incline is reported in just the x or y direction. This makes for simpler analysis.

Q.  How do I move an origin?

A.   If the origin has been located and is visible on the movie (like "Origin 1" is by default), drag the origin on the movie to the new location by clicking on the video point where the lines intersect

The Selected Origin is rotated by dragging it's handle around to the desired angle.

return to top

Q.  How do I rotate an origin's coordinate system?

A.   Select the origin once it has been located on the movie; drag the handle (the dot on the outer edge of the circle) of the origin around until the x-axis:origin; points in the desired direction. The angle of rotation between 0 and 360 degrees will be displayed in the Coordinate System Window where the header of the coordinate system you are rotating is located.

Q.   I want to reported data in two coordinate systems.

A.   You can do this by cloning the point series and associating the clone with a second coordinate system (see section 3.2 in Chapter 3). Once this is done, data are reported for both the original point series relative to its coordinate system and its clone relative to a second coordinate system.

Q.  How can I adjust for the camera moving when the movie was filmed?

A.   You can define a moving coordinate system that allows you to compensate for camera motions by selecting a familiar background feature as the origin in each frame. This feature allows you to analyze motions in frame sequences made from video images recorded with a hand-held camera. In short, the position of an object relative to any other object on the screen can be reported on a frame-by-frame basis.

In these three frames of a Saturn IV rocket launch the camera is panning. If an origin that is fixed relative to the screen is used, the data for the rocket's location will be useless. If a moving origin is located at the top of the launch tower on a frame-by-frame basis, meaningful data for the rocket's acceleration relative to the tower which is fixed to the ground, can be obtained.

In the graph above, the rocket appearts to stand still when coordinate data for its position is reported relative to an origin fixed on the screen. The graph on the right shows the rocket nose's position recorded relative to a moving origin that follows the location of the top of the launch tower as the tower appears to move down the screen in each successive frame.

return to top

Q.  The camera panned so much that my origin disappears.

A.   In some cases defining coordinate transformations allows you to obtain intelligible motion data of a feature or object of interest that is tracked with a panning camera. For example, the path of a broad jumper or basketball player performing a slam dunk can be reconstructed even if the camera is sweeping so broadly that background features change completely.

In these four movie frames, the camera pans so much that no stationary object is in the frames at all times. In this case the origin (marked by a white circle) is located at one of the red tiles against the wall in the first few frames. Before the red tile disappears the origin is transformed to the bottom left corner of the black rectangular heat vent (as marked by a white square). VideoPoint corrects for the motion by storing the difference in the coordinates of the red wall tile and the black heat vent based on the locations of these features in a frame in which both are visible.

Four frames from a movie where the camera pans so much that no stationary object is in the frames at all times. In this case the origin (marked by a white circle) is located at one of the red tiles against the wall in the first few frames. Before the red tile disappears, the origin is transformed to the bottom left corner of the black rectangular heat vent (as marked by a white square). VideoPoint corrects for the motion by storing the difference in the coordinates of the red wall tile and the black heat vent based on the locations of these features in a frame in which both are visible.

return to top

Q.  How can I define a coordinate transformation?

A.   Step back to the last time that the origin was in the movie. Here you will want to translate/transform the origin to a different object that is stationary but will not leave the field of view immediately. In order to get good data, the second object must be in the same plane as the first origin. In general, it is wise to choose this transformed/translated origin such that it does not leave the field of view at all. However, you can transform/translate origins as many times as necessary.

Q.  The camera zoomed when the movie was filmed.

A.   You can define scale factors on a frame-by-frame basis to follow the motion of an object with a camera that is zooming in or out or to track the motion of an object moving off into the distance. For example, the acceleration of the lunar module as it ascended from the moon during one of the Apollo missions can be determined even though the camera had been programmed to zoom back rapidly. The horizontal acceleration of a drag racer can be determined even though two different fixed cameras have been used to follow it as the dragster moves off into the distance and noticeably changes its size on every frame.

Q.  How do I compensate for a zooming camera?

A.  If the camera zooms while filming the movie, the pixels/meters calibration ratio will change on every frame of the movie. In this case, you will need to set up a scale for your coordinate system that is relocated on every frame.

If you have already created a scale:
1) Choose Options->Change Scale....

In the case of a zooming camera it is necessary to open this dialog to change the scale from a .fixed scale type to a Frame-by-Frame scale type.

2) Select the scale that you want to change (usually there is only one).

3) Change the scale type to "Frame-by-Frame". This changes the video points that make up the ends of the scales from fixed video points to frame-by-frame video points. These video points, usually named Scale1A and Scale1B (the numbers change with more scales) will now need to be selected on every frame. They will automatically be selected just like any other normal video point.

If you have not yet created a scale:

1) Choose Movie->Scale Movie...
2) Enter the known length of the object in the movie.
3) Select the origin of the frame that you want to scale.
4) Change the scale type to "Frame-by-Frame." This changes the video points that make up the ends of the scales from fixed video points to frame-by-frame video points.

Q.  The motions are in two planes.

A.   Each frame can be scaled by a different scale object. When all of the motion is in one plane that is perpendicular to the camera axis, only one scale object is necessary. However, if there are motions in two planes, you need to set up two scale objects and two coordinate systems because the objects in the plane that is farther away from the camera will appear smaller. One origin and one scale should be in one plane, the other scale object and origin should be in the other plane. Note that you will have to choose two different objects in the movie to use as your known lengths; one is needed for each scale.

When motions take place in two planes in a movie, two coordinate systems can be set up by using different scale objects within the movie frame.

Q.   The camera had a zoom lens or a wide angle lens.

A.   If the lens on the camera is a fixed lens, no significant distortion should be present. However, most commercial video cameras have variable focal length zoom lenses that range from wide angle to about 10X zoom. Although distortions are usually negligible in modern zoom lens systems, the zoom lenses in some low cost cameras may cause radial image distortions. Radial distortions can give a pincushion and/or barrel shape to a rectangular image. We have found that distortion is minimized when the camera is located fairly far away from the object of interest and then the zoom is set about halfway in, so the motion fills about 3/4 of the screen.

Distortion is quite noticeable when looking at a large motion (i.e., a bungee jumper) through a wide angle lens. Your data for the top and bottom of the motion can be noticeably inconsistent (10% error).

Q.  I'm only interested in part of the movie.

A.   If you are only interested in a small section of the movie, you can shorten the movie by cutting out frames using software such as Apple's MoviePlayer (Macintosh) or Adobe Premiere (Windows or Macintosh) before analyzing the movie. Alternatively, you can skip the uninteresting frames when you locate features or objects of interest.

Q.  The movie is a time-lapse or slow motion frame sequence.

A.   If the movie does not play in real time even when the Movie->Play All Frames menu entry is unchecked, the original movie may be either a time-lapse or a slow motion movie. Also, movies that have been digitally edited and saved at a different frame rate than the original will not play in "real time". In these cases, you can override the time code that is read by VideoPoint to display the time. To do this, choose Movie->Select Frame Rate.... This prompts you to enter the number of frames per second that you think the movie really is. Every frame in the movie will then have a time that is equivalent to (seconds/frame) * (frame #).

Q.  It looks as if a frame is missing.

A.   Sometimes, when you take data, it looks as if a frame is missing from the movie. If the time code of the movie corresponds to this missing frame (i.e., there is a larger than normal step between the previous time to the next time), do nothing. VideoPoint will just use the time code on the movie. The dropped frame is common with less expensive capture cards that are set to capture at a higher frame rate than it can handle. Though there might be an extra space in anything plotted against time, the data will still be accurate.

If, however, the time code does not reflect a missing frame (i.e., the time step is the same throughout the whole movie), this is a problem. Since there is no way to determine the real time of the frame following the dropped frame, some error will be present. This situation generally appears on files converted from Video for Windows (.AVI) to QuickTime for Windows (.MOV) since Video for Windows does not associate time codes with individual frames. This is one of the reasons we chose to use the QuickTime format. If you are capturing the movie, you can alleviate the problem by reducing the capture rate in frames/second of your capture card. If the movie you are using has been transformed to QuickTime format from a commercially made .AVI file, there isn't much you can do.

Q.  I notice that some of the frames are duplicated.

A.   If the movie has been taken from a video tape made from a film (movies, sports highlights), look at the next entry which deals with problems also created with dubbing commercial films into videotapes. If your movie was digitized from a live source, your digitizing card or software may have duplicated the frames to keep the frame rate constant. In this case, the movie can probably be analyzed if you skip duplicated frames when taking data.

Q.  The movie is from a commercial film dubbed onto a videotape.

A.   Movie film is normally recorded at 24 frames/second. Videotape is normally recorded at 30 frames per second. When 24 fps film is transferred to video tape, every 4th frame is usually duplicated. Though you won't notice this duplicated frame while you are watching the video tape, its presence is problematic when analyzing digitized frame data. If changes are happening slowly in a movie, the simplest way to assure that the time interval between frames is constant and no repeated frames are present is to digitize the "commercial" videotape at 6 fps "every fifth frame". If any other frame rates are used in digitization, special problems will be created.

If you want to digitize a commercial videotape at a higher frame rate the best way to do this is to compress the movie at 30 fps and then remove the extra frames "by hand" before re-compressing it. If you want the maximum frame rate you should re-compress the movie at 24 fps. You can also re-compress the movie at 12 fps, 8 fps or 6 fps, and so on.

If the commercial video is digitized at 30 fps and re-compressed at 24 fps...

It is tempting, when digitizing a commercial movie, to consider digitizing it at 30 fps without removing extra frames and then recompressing it at 24 fps. Don't do this. If you are lucky, this will work beautifully. However you only have a 1/6 chance of this process working. It would depend on which frame your re-compression software drops to get the movie down to 24 fps. If it happened to drop the extra frame as the first dropped frame, everything will be correct. However, dropping any other frame first will lead to incorrect frame times.

As indicated above, the best way to handle the situation is to remove the extra frames "by hand" while digitizing at 30 fps and then recompress the frames at 24 fps.