Avisynth GUI Stereoscopic Video Script Writer

• Primarily, the 'Avisynth GUI Stereoscopic Video Script Writer', is a simple programming project I have devised to consolidate knowledge as I study computer programming. It is not intended to replace 'on the fly' conversion programs which are by far the best option but also cost the most and as I'm not a rich man I have to make a compromise. But hopefully the script writer will simplify and quicken Avisynth script writing of which I personally use frequently.

• Avisynth is a file server and when used with Virtual Dub can be a powerful tool to correct vertical/horizontal misalignments, resize, convert one 3D format to another, edit video and what's more they're free!
• I have a digital camcorder (Sony TRV120e) and a Firewire card which generates AVI files with the DV Type-1 codec. Virtual Dub is not directly compatible with Type-1 DV whereas Avisynth is, which means if I want to use VirtualDub to edit/manipluate a video I have to load the AVI file via Avisynth which then 'serves' it to VirtualDub. If I need to save the video with the Type-1 DV codec I simply use Link2 which frameserves from VirtualDub to programs such as Ulead's Media Studio Pro or Adobe Premier.
• Avisynth scripts are relatively easy to write but with the aid of a Graphical User Interface (GUI) this program will make the process quicker and simpler.
• Whilst there are programs which convert 'on the fly', such as 3D-Combine, Parallax Player, DepthQ. Their cost is somewhat of a put off for me and judging by the hardware spec recommended by some, leads me to believe they are aimed firmly at the professional / industry market. So I use Avisynth, VirtualDub and the player from Michal Husak, Stereoscopic video player as a compromise between quality, cost and time of conversion.

Avisynth coupled with VirtualDub can create the following 3D formats and convert between them:
	Above/Below (with varying central band width, usually 5% of frame height)
	Side-by-Side
	Interlaced (Field Sequential)
	Above/Below Anaglyphs for use with page flipping video players such as Michal Husak's Stereoscopic video player. This option is good if you want to share your 3D videos with people who don't own shutter glasses and want to maintain high quality anaglyphs without colour bleed/blending
	Anaglyphs of any primary colour combination: Red/Blue, Red/Cyan, Yellow/Blue, Magenta/Green. Including grayscale of one or both channels.

A screen shot from the program

The above example would:

• Open 2 video files located on drive D:\ named: Left.avi and Right.avi . Which had been previously compressed using the Type-1 DV codec, When this box is checked it tells Avisynth to use Windows Directshow.
• Converts:
  • The left channel to grayscale (this helps reduce retinal rivalry in pictures with a high blue/green content)
  • Left image to red, Right image to cyan.
• Then combine the two images into an Above/Below format with a Central Band Width of 0 pixels.
• The Avisynth script (*.avs) is then ready to be Saved
• The Script is then loaded into VirtualDub and saved as an *.AVI file with a suitable compression codec or frameserved to another program such as TMPGenc and saved as an MPEG file.

Pictorial representation

Representation of the conversion stages within Avisynth.	2 separate left/right videos Converted to	'Above/Below anaglyph'. In reality the user only sees this end result

The Horse Shoe Stereoscopic 3D Video

To download the video segment, RIGHT Click on the image and
"Save Target As" (Internet Explorer) or "Save Link As" (Netscape).


Size: 6.9MB, MPEG-1 Above/below 5%, Duration 22 seconds, 25fps

3D Movie Plus

Stereoscopic video player

Nu-View Stereoscopic 3D Video Attachment

The above video was originally shot with the Nu-View Stereoscopic 3D camcorder attachment.

The Nu-View Adapter uses a proprietary, patent pending process which allows two distinct views (left eye and right eye) to enter the single lens (eye) of a camcorder in order to provide a true stereoscopic 3D image to be displayed and/or recorded. A threaded optical adapter is first attached to the camera's filter threads, which provides a stable optically aligned bayonet mount for the Nu-View. After the adapter is attached, the camcorder is used in a normal manner to view and record. To duplicate a true stereoscopic 3D experience the adapter provides the user with a simple single convergence control which allows you to converge on objects near and far.

Look into a Nu-View from the "big" end – where light enters the adapter. You will see two liquid crystal shutters (LCSs), one straight-on, one through the mirror. Think of these as two viewports, one for the left-eye view and the other, the right-eye view. They are positioned 2.2 inches (56 mm) apart, slightly less than the average human interpupillary distance (IPD) of 2.6 inches (66 mm). This distance forms the "stereo base" of our 3D system.

Now look through the "small" end – where Nu-View is connected to the camera. You will see two images, from these two viewports, superimposed on each other. If you now turn the Convergence Adjustment Knob, the two views move horizontally relative to each other. You can make the two views line up for distant objects or for near objects, but never for both at the same time. This is called parallax.

The Nu-View sychronizes with your camcorders field scanning. Standard PAL TV uses 50 (NTSC uses 60) fields per second (2 fields make 1 full frame). The left-eye view is recorded in Field one, followed by the right-eye view in Field 2. This method of recording stereoscopic 3D is called Field Sequential (FS). The image on the right is a slowed down (approx. 10Hz) demonstration of FS working.

Alternating between left-eye/right-eye views at only 50/60 fields per second can lead to perceptable flicker. But this can be eliminated completely if the video is watched on a standard PC monitor whose refrsh rate is more akin to 120Hz. Or the video can be converted to other 3D formats (such as above) which don't rely on the Field sequential method of presentation.

---

e-mail gedburnell@onetel.net.uk

---

Under construction

Using VirtualDub for Stereoscopic Video

Below is a zipped file containing VirtualDub filters I have found useful for 3D Stereoscopic video.

Size: 100KB

Big thank you to the following people; Simon Walters, Valentim Batista, who have kindly written and modified existing VirtualDub filters to accommodate stereoscopic video. Go to their respective website for the latest filter updates.

• http://www.geocities.com/siwalters_uk/fnews.html Simon Walters
• http://www.geocities.com/gc_timsara/vdub.html Valentim Batista

How to use the filters (an overview)

Here is a step-by-step procedure for using the filters from Donald Graft . First ensure that the filter .vdf file is present in a directory called "plugins" that exists in the same directory as the VirtualDub executable. Start up VirtualDub and maximize its window. Select File/Open video file from the pulldown menus and then open your input video file. Select Video/Filters from the pulldown menus. Click on the Add button in the Filters dialog box. Scroll down in the filter list and highlight the desired filter by clicking on it. Click OK. The configuration dialog for the filter appears. Configure the filter as desired. If the filter supports preview, you can interactively explore the effect of different filter settings. When you are happy with the settings, click OK in the filter configuration dialog box. Click OK in the Filters dialog box. Select Video/Compression from the pulldown menus. Highlight the desired compression for the output file and then click OK. Select File/Save As from the pulldown menus and select the desired output file. Processing will begin and when it is completed, the output file will contain the filtered video. (Alternatively, you can simply preview the result by using VirtualDub's timeline player.)

Description of specific filters