Stereo3D Toolbox LE 3.0 User Guide

Introduction
Stereo3D Toolbox LE 3.0 is a straightforward, easy-to-use stereoscopic 3D mastering plug-in designed for beginners, hobbyists and stereoscopic enthusiasts alike.

The plug-in allows users to edit with discreet left and right camera files or pre-multiplexed video files in either Side-by-Side format or Interlaced (Left on odd). The user is then able make any necessary convergence adjustments or disparity corrections to their stereoscopic video clips with a comprehensive array of tools. Finally, the user can select their desired output mode in order to view their content on 3D displays, YouTube, BluRay or Anaglyph.

Stereo3D Toolbox LE 3.0 is available through Noise Industries’ FxFactory. The plug-in is supported by Final Cut Pro 6, 7, X, Final Cut Express, Motion 4-5 and After Effects CS3-5.5.

What’s New in LE 3.0

Stereo3D Toolbox LE 3.0 is a major upgrade, introducing features such as automated nesting and more stereo 3D Output Modes:

- Side-by-Side

- Above/Below

- Line-by-Line Interlaced

- Checkerboard

- Anaglyph Outline

- 50/50 Superimpose

- Difference

- Anaglyph Colour

- Anaglyph B&W

- Stereoscopic (Hasbro® my3D for iPhone 3Gs & 4)

- 2D Bypass Corrections

Left & Right Eye Nests
Two new plug-ins included in Stereo3D Toolbox LE 3.0 are the automated nesting plug-ins. These new plug-ins allow the user to efficiently create Side-by-Side files that can then be nested before editing. This replaces the convention Image Well workflow used in the previous version.

When working in FCP X simply select the stereopair, right-click and select “Synchronize Clips” then open the new synchronized clip and apply the Left Eye Nest and Right Eye Nest.

S3D Input Type

Discreet Left and Right Video Files (Two simultaneous streams)

Image Wells

The Image Well method allows the user to mux a stereo pair within the Stereo3D Toolbox LE plug-in.

To use the image well method, place either the left or right clip on the timeline then apply the Stereo3D Toolbox LE plug-in. In the S3D menu select the corresponding option – if the clip is the left, select “Clip = Left & Image Well = Right” and vise versa if the filter is being applied to the right clip. Then click and drag the corresponding clip into the image well directly below the S3D menu.

The sequence that the stereo-pair clip sits in can then be nested into a new sequence to begin editing.

Pre-Muxed Video Files

Side-by-Side

Side-by-Side is a very common 3D format. Since the left and right image exist in a single video stream it allows for easy transferability between a hard drive, the NLE and 3D display. In certain circumstances the user may transcode single SIDE-BY-SIDE files before editing to avoid bottlenecking, a common occurrence when attempting to playback two video streams. The user is able to apply Stereo3D Toolbox LE directly to a native SIDE-BY-SIDE file then set the S3D input type to “Side-by-Side” then immediately begin adjusting the convergence and disparity controls.

Interlaced

Similar to native Side-by-Side, a user may decide to transcode single interlaced files. Not to be confused with native interlaced video, when two video fields combine as one whole video frame. Rather the left image is in the odd lines of resolution and the right image is in the even lines.

The user would simply apply the plug-in to the 3D Interlace clip, set the S3D Input type to Interlaced then begin editing and adjusting the convergence and disparity controls. Then finally, selecting the desired Output Mode.

Input Squeezed

The Input Squeezed option allows users to select if a Side-by-Side is squeezed. Conventionally, a Side-by-Side file contains two 1920×1080 video frames that have been squeezed to 50% of their original width, resulting in each image consuming 960×1080 of each frame, equaling 1920×1080.

In certain circumstances a user may decide to work with “Unsqueezed Side-by-Side” video files. This means that both the left and right eye maintains full 1920×1080 resolution but is contained within a single 3840×1080 frame (4K wide, 2:1 aspect ratio). If this is the case, the user is still able to apply the plug-in and use it just as one would with a squeezed Side-by-Side input. The downfall of an unsqueezed Side-by-Side file is the large frame size and increased processor demand. An average editing system will have difficulty transcoding such a file and playing back within Final Cut Pro. However, it is worth experimenting.

Swap Input

Allows users to swap the left and right input incase of an error earlier in the pipeline.

One of the most common errors in stereoscopic 3D editing and mastering/grading is inverted 3D – the left and right eye in the place of one another. For example, a user may have a Side-by-Side video file and the right frame is in the position of the left and the left frame is in the position of the right. The result is inverted 3D and causes objects to be indented and can cause headaches amongst viewers.

3D Output Mode

For previewing purposes and final output the 3D Output Mode gives the user a wide selection of the most popular 3D formats, including; Side-by-Side, Above/Below, Line by Line Interlaced, Checkerboard, Anaglyph Outline, 50/50 Superimpose, Difference, Anaglyph Colour, Anaglyph B&W, Stereoscope (Hasbro® my3D for iPhone 3Gs & 4), 2D Bypass Corrections.

Once the user has finished their edit and made any necessary convergence and disparity adjustments, they’re able to select the Output Mode. The type of display you’ll be viewing your content on will determine your stereo 3D output type. It is important to check what type of 3D input your display accepts.

For sharing your video on YouTube you’ll need to output as Side-by-Side.

3D Output Mode is also important when grading 3D. Modes such as Superimpose 50/50, Anaglyph Outline and Difference are designed for grading.

Side-by-Side

Side-by-Side has become a standard in modern stereoscopic 3D editing that all current professional and consumer 3D displays accept. The left and right video files are squeezed 50% horizontally (each frame equals 960×1080) and placed beside one another in a single 1920×1080 frame. The disadvantage of this format is the 50% loss of horizontal resolution however the simplicity and popularity of this format negates the loss of resolution for most users.

Above/Below

Similar to Side-by-Side, Above/Below squeezes the left and right video files however it does so vertically by 50%. Each clip is now 1920×540 within a single 1920×1080 clip. Most consumer 3D televisions also accept this format.

Line-by-Line Interlaced

Line-by-Line Interlaced is popular previewing format accepted by some 3D displays. The left clip’s resolution is halved then placed in the odd lines of the video stream. Correspondingly then right clip’s resolution is halved then placed in the even lines of the video stream. When combined the video stream totals 1920×1080.

Checker Board

Checkerboard is a method of combining the left and right eye in both the vertical and horizontal resolution. It achieves the highest diagonal resolution out of all the single-stream 3D formats.

When speaking in terms of one line of horizontal resolution, each odd pixel (1, 3, 5, 7, etc…) contains the left images; each even pixel (2, 4, 6, 8, etc…) contains the right. The next horizontal line below the previous will begin with the right images on the odd lines and the left on the even lines. Thus resulting in a checkerboard pattern throughout the frame.

Anaglyph Outline Mode

New to Stereo3D Toolbox LE 3.0, Anaglyph Outline Mode makes correcting image disparity much easier and intuitive. This mode is strictly designed for mastering functionality by creating a very clear visual representation of the difference between the left and right images. Both left image and right images are treated with an edge effect then the left is coloured red and the right is coloured blue. When an edge from both images overlap (no parallax) in turns white. This effect results in a very clear representation parallax and convergence to the user.

50/50 Superimpose Mode

A standard 3D analysis view, 50/50 Superimpose layers the left and right image at 50% opacity. This mode is useful to users who wish to see a simple, unaffected superimposition of the two images. This format is commonly used when calibrating a stereo-rig. This is another mode strictly used for finding disparities between the two images and is not a delivery format.

Difference Mode

Difference Mode is an analysis view. By layering the left and right image, any area with difference will appear as false colours, any areas with no difference turn thus revealing the convergence point or point of no parallax. This mode is also useful for finding geometric disparities between the two images. Difference Mode is not a delivery format.

Anaglyph Colour

Anaglyph is a classic method of viewing stereoscopic images. It is the easiest method to share stereoscopic content without any special display technology or expensive eyewear. Anaglyph can be viewed on any colour display with inexpensive paper and coloured plastic lenses.

The basic principal of Anaglyph is for the parallax in the left image to appear red and parallax from the right image to appear as cyan. The corresponding lenses in the glasses worn by the viewer filter out the opposite eye’s image only allowing the correct eye to be viewed, resulting in two separate images reaching the viewers eyes independently. The viewer’s visual system then processes and combines the two into a stereoscopic image.

Anaglyph can become extremely uncomfortable after extended periods of time due to the heavy colourization of the images and heavy filtering over the eyes.

Anaglyph Black & White

Anaglyph Black & White works within the same principal as Anaglyph Colour however the saturation from the images is completely removed with the exception of areas with parallax.

The advantage of Anaglyph Black & White is the absence of competing chrominance of the images themselves and the colour effect created by the Anaglyph method. For example; an image that contains dominant areas of red or orange chrominance (such as red clothing, skin tone and essentially any red objects in the scene) can compete with the red effect in the left image, resulting in an image that can be difficult for a viewer to fuse resulting in a lose of the stereoscopic effect. This same result can occur with images containing dominant areas of cyan or blue.

Stereoscope

New to Stereo3D Toolbox LE 3.0, Stereoscope output mode allows users to master their project for viewing on the Hasbro® my3D stereoscope for the iPhone 3G, 3Gs and iPhone 4.

2D Bypass Corrections

2D Bypass Corrections output mode only displays the left image and the clip becomes unaffected by any disparity of convergence adjustments the user has made. This is useful when a project has been completed and the user has created a Side-by-Side version, an anaglyph version and wishes to have a 2D version as well. Instead of selecting each clip and removing the plug-in, the user is able to select 2D Bypass Corrections on a single clip and the entire sequence will only display the left image without removing any convergence or disparity adjustments.

Use Global Output Mode

Use Global Output Mode is a setting that allows fast switching between S3D Output Modes and its use is highly recommend. In order to gain full advantage of this feature the user must select the checkbox in every instance of the plug-in, in other words, whenever the plug-in is applied, check the box.

If a user decides to switch from Side-by-Side output to Anaglyph Colour output simply select Anaglyph Colour and check “Set New Global Output Mode”. This will change the output mode of every instance of the plug-in. Global Output Mode is a system-wide setting. If you select this in Final Cut Pro, the settings in Motion and After Effects will also change.

Set New Global Output Mode

Set Global Output Mode works in unison with “Use Global Output Mode”. If the user decides to switch output mode, they can open a single instance of the plug-in in the Viewer window (in FCP), select the new Output Mode then check “Set New Global Output Mode”. This will change the output mode on every instance of the plug-in to the chosen output mode.

Convergence and Disparity Controls

The convergence adjustment slider allows the user to adjust where the parallax budget (depth bracket) sits on the z-axis in relation to the screen plane. The disparity controls allow the user to adjust for geometric disparities between the left and right images. These can be the result of a misaligned camera rig or a bump while shooting. Each slider can be adjusted until the desired position is achieved.

The point of zero parallax or the convergence point is on the screen plane. Objects in a scene that appear behind this point appear to be “inside” the screen, this is called positive parallax and objects that appear in front of this point appear to “come out” (they exist in theatre space) this is called negative parallax.

When shooting, the inter-axial distance will determine the parallax budget (depth bracket) of the scene, or in other words, how “deep” the scene appears to the viewer. Adjusting the convergence point will slide the parallax budget or “depth” either in front or behind the screen plane.

When shooting 3D it is extremely important to be taking measurements of the further object, the convergence point and closest object within a scene and to be making calculations to ensure the proper inter-axial distance is set for the desired convergence point. It can also be described as fitting the depth budget of a scene into the appropriate parallax budget or depth bracket for the desired screen width. If these steps are ignored during shooting the footage may have excessive positive parallax, causing the viewers eyes to diverge or excessive negative parallax, causing extended periods of hyper-convergence, resulting in headaches. There is also the potential for objects in negative parallax to break the stereo-window, also resulting in uncomfortable viewing.

Stereo3D Toolbox LE 3.0 also features Vertical, Rotational and Zoom Disparity controls.

Convergence and Disparity Controls are not keyframeable in Stereo3D Toolbox LE 3.0.

Autoscale

The Autoscale feature automatically scales the frame to remove any black edges introduced as a result of convergence or disparity adjustments.

Swap Output

Swap Output is the same as Swap Input however, as the name describes, the images are swapped during the output. This is useful if the user discovers inverted 3D and would like to quickly swap the two images.

3D Preview Mode

*Not available in FCP X or Motion 5

The 3D Preview Mode is a section window to help analyze the stereoscopic effect. It essentially acts as a secondary display that appears in the Filters tab. The user is then able to preview in the desired output method in the canvas window and analyze with a different output method.

Mix

Standard to most effects, the mix slider adjusts the amount of the effect you see in the final image. Always leave this at 100.