Fast PlayBack Mode

Many users need to utilize geometry to accurately see the effects of lip sync on their models, but on all but the most powerful machines this playback mode doesn't approach real time. To provide real time playback with geometry mode, a new playback mode has been provided that effectively "playblasts" the geometry frames.

To the right of the playback button, a new button that has the playback symbol and a small "S" has been added. In addition, the reference playback window has a pull-down menu that selects between "Reference", "Playback", and "Off".

When the new playback button is pressed, Soundsync will play forward, one frame at a time, on the current selected region. If no region is selected, it will play back the entire width of the window. It will grab frames from the "Mixed" keyset window and prepare them for playback in the Reference window. This process is NOT INTERRUPTIBLE. Once the frames are grabbed, the frames in the Reference window can be replayed in real time.

Example of a Typical Workflow

1. Select a region to work in. This region shouldn't exceed a few hundred frames (for memory reasons).
2. Position the viewpoint in the "Mixed" window to your liking. Turn the "Keyset" window to "Off".
3. Press the "Play Forward & Save" button, and wait while the images are grabbed.
4. Now select "Playback" in the reference pull-down menu and turn the "Mixed" window off.
5. Using the regular playback button (NOT the Playback & save), view the sequence.

Example: Creating a Head to Use in SoundSync

1. Create a head in Alias with the following (simplified) dag node hierarchy:

   

   The keysets required are as follows:

   Mood	containing Happy and Sad.
   Speak	containing Lips tightly closed, Mouth half open, Mouth fully open, and Mouth in O shape.
   Eyes	containing Look Left, and Look Right.

2. Create a neutral position at frame 0. The neutral position is very important as it is the basis for the changes used to animate all of the keysets. Include the neutral position in all of the keysets.
3. Create the Mood keyset by manipulating the cheek-cluster.
4. Create the Speak keyset by manipulating the lip clusters and the jaw.
5. Create the Eyes keyset by manipulating the left and right eyes.
6. Lay out the frames as follows:

   Frame	Keyset/Keyshape	Dags Affected
   0	Neutral	left-eye, right-eye, lip-upper-cluster, lip-lower-cluster, jaw, cheek-cluster
   10	Eyes/Look Left	left-eye, right-eye
   20	Eyes/Look Right	left-eye, right-eye
   30	Speak/Lips tightly closed	lip-upper-cluster, lip-lower-cluster, jaw
   40	Speak/mouth 1/2 open	lip-upper-cluster, lip-lower-cluster, jaw
   50	Speak/mouth full open	lip-upper-cluster, lip-lower-cluster, jaw
   60	Speak/mouth O shape	lip-upper-cluster, lip-lower-cluster, jaw
   70	Mood/Happy	cheek-cluster
   80	Mood/Sad	cheek-cluster

7. Create the keyshapes with increments of 10 frames between them so that you can run an animation of the wire file showing some of the transitions between the keyshapes.

   It is not necessary to create the keyshapes with increments of 10 frames between them, the frames can be adjacent. This is done for the example.

   The information in the above chart is important; you will use it to construct a keyset file later to get your data into SoundSync.

Using Shapeshifter with SoundSync

SoundSync can also work with keysets created using Shapeshifter nodes in Alias. Shapeshifter controls a set of clusters on one or more surfaces through a single control value. Ashapeset (rather than a keyset) in SoundSync allows you to manipulate that control value.

When you are constructing a model, do not put any keyframes on the control node and make certain that its static value is set to the minimum or zero value. You can use as many Shapeshifter controls as you like. You need to know the name of the control node, the channel being used, and the range of the control value when you construct your shapeset.

Overview

Keyset functionality within SoundSync references keyframes that you have set in your wire file. Shapesets allow you to directly manipulate control sliders to affect geometry or attributes through the use of expressions. Shapesets also allow the direct manipulation of slider geometry associated with Shapeshifter.

Shapeset Workflow

There are several differences between keysets and shapesets in SoundSync. The most important difference is that in keysets, soundsync looks at all animated dag nodes at the frame you specify within a keyset block after each "key" entry. When using shapesets, there is no keyframe information to reference, but by using the "key" entry, you specify a value for control slider geometry.

You can set keyframes for sliders for use in SoundSync, but this would be referenced within the keyset file as a keyset block instead of a shapeset block.

Again, shapesets do not reference keyframe information, but instead apply a value to the control dag node.

The Shapeset Block

Below is a sample keyset file with two shapeset blocks...

The following sample contains all of the required elements of a shapeset.

begin shapeset cheeks cheek 0 0 1

key relaxed 0

key smile 0.75

key puffy 1.1

end shapeset cheeks

begin shapeset	Defines a new shapeset block.
cheeks	The shapeset name. This appears as a separate entry in the Keyset menu in SoundSync.
cheek	Control geometry name. This is the dag node to be translated by SoundSync. This control dag can be from shapeshifter or by your own creation.
001	Values for channel (see note, below), min, max. The first number is the channel number*, the second is the min xform value, the third is the max xform value. In this sample, these values have the following meaning: 0 - Apply Xform to the X_Translate Channel 0 - Specifies 0 (Zero) as the minimum xform value for the control dag 1 - Specifies 1 (one) as the maximum xform value for the control dag
key	Begins a new key entry. Each key will have a separate key button in SoundSync (same as in a keyset block.)
relaxed smile puffy	Key name. This will be name of the key within SoundSync. Each key has a unique name and the button for this key will be labeled with this name.
0 .75 1.1	Key value. When a keyframe is set within SoundSync, this value will be applied to the slider when the "relaxed", "smile", or "puffy" key is selected.
end shapeset	Defines the end of the shapeset block.

where:

The channel is entered in the form of a value that corresponds to the following table: Channel number Channel affected 0 X_Translate 1 Y_Translate 2 Z_Translate 3 X_Rotate 4 Y_Rotate 5 Z_Rotate 6 X_Scale 7 Y_Scale 8 Z_Scale

In the above example, an image could be specified after the value for a particular key. The use of images is a keyset or shapeset is totally optional; however, this can improve the appearance and usability within the SoundSync user interface.

To reference a rendered image, change the key entry as follows:

key relaxed 0 cheek.relax

Where "cheek.relax" is an Alias pix image that represents the relaxed cheeks.

If an image name is not supplied, default images will be generated when the keyset file is retrieved in SoundSync.

More on Keyset Files

ShapeShifter provides a simple mechanism for controlling a complex series of interpolations between key shapes with a single control.

A keyset file might look like this (see head.keyset in the clip directory):

begin shapeset Hat hatmorf 0 0 1 key droop 1 key windblown -0.5 key normal 0 end

The syntax is as follows:

begin shapeset <keyset-name> <dag-name> <channel> <min> <max> 

key <key-name> <value> ... end

Alternatively, you can leave the <min> and <max> parameters off and the software will assume 0.0 and 1.0:

begin shapeset <keyset-name> <dag-name> <channel>  key <key-name> 
<value>  ... end

The <channel> parameter is a number indicating which dag channel is changed to alter the shapeset; 0 is X_Translate, 1 is Y_Translate, and so on.

Soundsync will tesselate the <min> and <max> positions. By moving the slider, any intermediate position can be achieved; the keys are provided as a convenience.