7. Reference

7.1 Hypervideo File structure

An hypervideo file basically lists a set of sequences. Strictly speaking, an hypervideo file consists in a header block followed by a set of sequences. Each sequence is an ordered list of frames plus specific properties. Each frame corresponds to a position of the square in the window.

7.1.1 Header

Any hypervideo file starts with a monolithic header block. It can be used in a parser to ensure the file is an hypervideo structure and not something else.

The header block is followed by the filepath of the video file. Thus a third party program knows where to look for the video associated to the hypervideo data.

7.1.2 Sequence block

A sequence structure consists in :

1. a first line with the name of the sequence, sequence=...
2. a description line, description=...
3. an hypermedia link type, hypermedia_type=seek|url
4. an hypermedia link action, hypermedia_action=...
5. a visual effect description, visual_effect=...
6. the amount of keyframes
7. all keyframes

A keyframe structure consists in :

where :

• FRAME is the frame position in the video stream. It can't be negative. And it can't be higher than the amount of frames in the video.
• X1 : is the horizontal coordinate for the upper left corner of the square
• Y1 : is the vertical coordinate for the upper left corner of the square
• X2 : is the horizontal coordinate for the down-right corner of the square
• Y2 : is the vertical coordinate for the down-right corner of the square

The 7.1.4 paragraph exhibits a graphics of that.

Each member of this structure is separated with a comma except the last. The comma after FRAME is doubled since in fact there should have been an argument which is implicit. This argument may appear in future versions.

There are normally no blank lines between two sequence blocks.

7.1.3 Tell me more

As everything is in text format, the number 40 for example is represented by the ascii code '4' and the ascii code '0'. To read an HPV file, you need a parser.

Now a little detail about the keyframe structure. As any keyframe position corresponds to a square position, hence 4 points, and that the square is always directed horizontally and vertically, 2 points (4 coordinates in the 2D-plane) are necessary and suffisant to fully determine the position of the square in the window. This is the reason why we never mention the number of points of the hotspot.

If for any reason, an extension with advanced polygonal features was released, the argument after FRAME would be the number of points and the list of the n points (X1,Y1),(X2,Y2),...(Xn,Yn) would replace the current 4-coordinate substructure. The keyframe structure would consist in :

7.1.4 Coordinates 1/1000

As we deal with positions, we must talk about reference coordinates.

The system of coordinates is neither based on the real size of the video file itself, which is 352*288 for an ISO-MPEG1 video, nor on the displayed size of the video, which is variable as you stretch the window. It's a good idea to use abstract coordinates. The abstract system is a virtual XY-space between 0 and 1000 in the two directions. Hence the point (352,288) in the video mentioned earlier corresponds to the point (1000,1000) in our XY-space, while the point (0,0) in the video file corresponds to the point (0,0) in our XY-space.

The following graphics shows two things : the virtual XY-space and the geometrical position of the two points (X1,Y1) and (X2,Y2) of the square in the keyframe structure.

A piece of the hotspot may be outside the window, depending on you. In that case, the coordinates of those corners will be either greater than 1000, or lower than 0, ie negative. At run-time, a smart coordinate parser would clip the outside part so that it doesn't break the program just for that reason.

The corner (X2,Y2) is always more to the right and more to the bottom than the corner (X1,Y1), hence there is no reversed square.

7.1.5 Reuse of the hypervideo file

The textual format of the hypervideo file fulfills two main goals : 1. a set of data expected to be reused must be in readable and understandable form. A proprietary format can't match that. 2. you can use a simple text editor, open a couple hypervideo files, and start operating copy/pastes to either split or merge the needed subsets of sequences.

We wanted the hypervideo data to be open and reusable since we know everyone will beg for control over it.

7.2 List of keyboard shortcuts

7.3 Toolbar reference

	Open an hypervideo file, save and close the previous one if necessary
	Open a video file
	Save the hypervideo data file, with default filename
	Close the hypervideo file

	Create and select a sequence
	Go to previous sequence
	Go to next sequence
	Watch and modify the sequence properties
	Delete the current sequence

	Set the begin mark
	Go to the begin mark
	Play / Pause
	Start / Stop Record
	Go to the end mark
	Set the end mark

	Enable / Disable loop mode
	Show / Hide the hotspot
	Add a bookmark at the current video frame
	Automatic keyframer (current sequence)
	Watch video properties
	Disable / Enable the sound
	Encode the hypervideo data into a G2 file format
	Switch to fullscreen

7.4 Seek controls

As there are a lot of video seek controls, and we dispatched and focused on them all over the documentation, we wanted to add this section to clear the situation.

Main slider :

• Drag the slider
• Click somewhere in the slider to perform a page seek
• When you have the focus on the control, use left/right arrows to perform go back/go forward

Sequence slider :

• Drag the slider
• Click somewhere in the slider to perform a page seek
• When you have the focus on the control, use left/right arrows to perform go back/go forward

Go back button :

• Click to go n frames back(n being maintained by the frame step)

Go forward button :

• Click to go n frames forward (n being maintained by the frame step)

Frame counter :

• Enter a frame number, then lose the focus (press the TAB key for example) : this will make a seek to that frame

Bookmarks :

• Choose one your manual bookmarks to perform a specific seek. When you are on a frame with a bookmark, the bookmark appears with a check-box.

Shots :

• Double-click on one shot to go at the beginning of this shot

Mark-IN :

• Press the PAGE-UP key to go to the mark-in of the current sequence

Mark-OUT :

• Press the PAGE-DOWN key to go to the mark-out of the current sequence