PRODUCTION STEPS

    The following is my own interpretation of how they might have proceeded with producing the series based on some resources from the internet :

a. Motion Capture

b. Pre-visualization

c. Motion editing

d. Character modeling

e. Environment Design

f. Miscellaneous Effects

g. Rigging

h. Key-frame Animation

i. Compositing

Motion Capture :

    Equipment  - Vicon MX40

    Location - Pinewood Studios

    Actors act their parts ( in some cases one motion capture artist acts  as two or more different CG characters ) directed by the episode director. Optical markers on tight motion capture suits are used to capture movements. In some cases even a group of 4 or 5 actors in a scene do the capture together. I am not sure if they had facial motion capture every time. But, some parts were really well done (for example Capt. Black has a really good facial animation going for him) for which they might have used facial motion capture.

 

Pre-Visualization :

    They may have used Vicon’s own software package (iQ) for the pre-visualization stage since there is a brief mention in the Vicon webpage about the software being used in the New Captain Scarlet. Pre-Visualization can help improve the quality of the motion captured by on the spot editing and smoothing of data.

 

Motion Editing :

    There surely would have had the need to edit and clean up the data to match their needs for the scene and voice-acting. As one of the actors describes in his website, they were capturing motion for pre-recorded voice. This could have been done in the pre-vis or as a separate process to save time during motion capture.

 

Characters :

    The character modeling artists would have been working simultaneously or would have already done their parts, since there were previous episodes. However, they might have worked on special costumes, bloodied hands or whatever that particular episode demanded.

 

Environment Design :

    Environment artists would design the landscape, space-ship interiors and other locations for the CG actors to inhabit.

 

Miscellaneous Effects :

    A special effects team would have been responsible for the particle systems, fog, lighting & transform, shading etc.

 

Rigging :

    The step after motion editing where there are people who design user interfaces for key-frame animators usually by writing plug-ins in 3D modeling tools. This enables faster animation and gives much more control to the artist.

 

Key-frame Animators :

    Artists who add secondary motion to the rigged models that already have motion capture data associated with them. This is nearly the final stage in animation itself and by far the toughest where minute details need to be paid attention.

Compositing :

    The compositors bring the animation, environment, effects, voice and the timing together to form the final picture.

    * Ammunition Warehouse (16 seconds : 10:02 - 10:18)

    * Environment sets the mood and is pretty good

    * The lighting is also done well.

    * Lt. Yellow holds the heavy gun in a single hand

    * Eyes so wide open and blink very infrequently, slower than they should

    * Facial animation looks a bit impassive

    * Wrinkles on the forehead look a bit overdone

      fig 1. wrinkles on the forehead / Destiny’s eyes stay like this for 2 seconds

    Separate motion capture data for the two characters. One with the ‘startling turn movement’ for Lt. Yellow and one with Destiny laying a consoling hand on Lt.Yellow. Another possibility is a single data file with both the motion capture data with two actors interacting, since the system is capable. Facial animation looks good, but again there is a doubt of whether facial markers were used to track actors in close up shots. Even, if they were used the final animation seems impassive. The reasons might be that there were not enough markers or the key-frame editing / motion editing step had to do away with some of the captured data. If pure key-frame animation was done on the face, it would seem difficult to do without intelligently rigged morph targets or facial bones interacting with appropriate application of forward or inverse kinematics, but, it is possible. My conclusion is that facial tracking was not used, if it was used not enough parts were tracked.

fig 2. good facial expression on Destiny’s face (cheek deforming as she smiles)

    Environment and lighting looked much better than the motion capture. Quite realistic reflection on the armor based on environmental lights and some fog in the background of the warehouse make this scene look good.

    General environmental models in this scene do the required job. The models of the characters themselves look a bit patchy. Inanimate hair, neck-body connection has some artifacts(eg. there is a ridge of some sort that stays with the head model as the characters move their head, it does not look seamless), wrinkles on one of the characters does not look or act right and wide eyes that do not blink at the normal rate make them both look bad in most of the frames.

 

    *Facial motion capture could have improved realistic expressions.

    *Using more facial muscles / bones for rigging & animation would help.

    *Eyes, Cheek, Forehead and Jaws need better keyframe movement.

    *Pre-visualization during capture to remove joint stiffness will surely help.

    *Increasing the number of trackers will both increase the moves possible

     and give more creative modeling control to the animators.

Models :

    *Small effects like hair, sweat etc. possible through shaders will make

    a lot of difference.

    *Secondary motion and smooth skinning of the suit material that the

    characters are wearing will improve visual appearance.

    *Even if a bit stiffness is desired based on the material, sound of leather

     stretching or metal clanging might improve CG effectiveness.

    *The weight of a gun needs to affect the movement of the CG character. A

      probable solution is to use an object heavier than a toy gun while doing

      the motion capture.

 

 

General :

    *A general step by step method for all the motion capture data should be

     followed to avoid inconsistencies between animations & shots. (Here it

    looked like different methods were used in each scene).

    *Care should be taken when a smooth transition between two different

    motion capture data is needed.

    *The background fits perfectly in this scenario, but, I thought that it was

     too static to be interesting.

    *Fog movement, environmental object interaction can improve the scene

     as a whole.

 

    *Synchronized walk back to the team (4 seconds: 10:18 - 10:22)

    *Single motion capture data (i.e same actor doing the same movement) is

     used in this part of the scene.

    *There is no animation in the face of both the characters.

    *They walk synchronized to their actor’s motion capture, even the body

     models look similar (no yellow in Lt.Yellow’s suit).

    *The background does not match the foreground characters.

           Notice that they both have the same body mapped with the motion data

    *none

    Motion capture data is the same on both the models. The only difference is in the way Lt.Yellow holds the gun. The walk is similar and they even stop at the exact time frame. The data is a simple walk cycle, with some key frame animation for 1 eye blinking animation for 2 seconds and the gun wielding pose on Lt. Yellow.

    The background does not go with the CG characters mainly because it is nearly the same color. Even, high class compositing will not be able to get rid of the awkward separation between the foreground characters and the wall panels. The open doorway helps a bit to improve the appearance.

    Both the models for the bodies are similar except for the armor. This is not noticeable unless looked at closely since this animation sequence is just around 2 seconds in the scene.

 

    *Two different actors walking together would have introduced the random

     motion in the CG actors’ movements.

    *Again there is a case where data is reused and mapped to different

     characters. If this is going to be done, some random offsets should be

     introduced by the animator to differentiate movement.

    *The background panel’s color combination can be changed to fit the

     foreground characters in the final composite.

    *Textures on the panels of the control center make the background more

     like a 2D scene. Bump mapping, environment mapping techniques on the

     walls would definitely help improving the appearance of the scene.

    *Reflection and light based calculations for the background objects will

     surely improve the look.

 

    *Same improvements as in the previous scene would be useful.

    *Debriefing (6 seconds : 10:22 - 10:28)

    *Bad textures for the walls surrounding the control center

    *Guy near Captain Scarlet needs to be leaning on the control panel, but

      pose looks wrong.

    *Stiff movement of the arms by Capt. Scarlet

    *Guns again seem weightless in Capt. Blue’s hands.

    *Expression on the guy next to Capt. Scarlet( Griggs ) looks right.

    I am not sure if motion capture was done on this scene. Captain scarlet, Capt. Blue et al stand in a pose which is conveniently achievable by key-frame animation. But, if there was motion capture done, it was not done lively enough. Even if facial motion was not tracked generally in the series, the facial animation of the guy next to Capt. Scarlet (Mr.Griggs) is believable but looks easy to do using key-frame animation.

 

    The control panel seems to retain a retro look but it looks out of place due to the pose of Griggs who is leaning on the panel. If his posture was altered to fit into the 3D model or made to look as though resting on the panel, it might work. For example Destiny is resting her arm in the railing and that works well. The wall around the two characters is circular but the texture seems to lack depth and photorealism.

 

    The environmental model of the control panel does not match the quality of the character models. It is poorly texture mapped, while the suits of the characters have shininess and seem to reflect the lights in the room.

 

    Capturing Griggs’ pose as he is sitting on something would have made the posture more realistic as he leans on the control panel. As far as capt. Scarlet, I feel that the stiffness of the motion is because there is an attempt to prevent collision between his two hands as he brings them together after pointing at Griggs. The motion seemed like an approximation of what was intended to prevent collision. My guess is this could have been solved if appropriate data was recollected during the motion capture stage itself. Again, heavier objects need to be captured with Capt. Blue and Lt. Yellow’s guns. Even Destiny’s pose, where she is resting her hand on the railing seems a bit off, and I guess it was key-frame animated. If motion capture for that pose had also been done, that would have been more realistic.

 

 

 

    The texturing on both the walls and the control panels would need to be improved to match the texturing quality of the CG characters. Photorealistic representation of the environment will also increase the believability of the scene.

    The control panel model needs to be revamped completely to match the 3D characters. Even though it looks 3D, the panel just does not look right with the characters. Even the retro look can be kept in, but just the polygon count and quality of the model can be increased.

    *Close up on Destiny (4 seconds : 10:28 - 10:32)

    *Background looks separate again in this scene.

    *Facial animation is robotic and looks like purely key-frame animation.

    *Again the eyes don’t look realistic enough and look like a doll’s.

    *none

    I believe the animation was just using keyframes. Destiny’s eyes just look doll like and her face looks prosthetic. And, the eyes blink once in 60 frames or so, which seems a bit long. The environment again does not match the character.

 

    In close up shots like this, facial tracking would increase believability by leaps and bounds. Eyes should be more life-like and more lively. In some cases the characters roll their eyes, pretty well, but, this should be used effectively enough. The background should definitely change. Maybe, some Depth of Field would improve the image, with some glow in the background.

 

 

In General, the series seems to have used simpler/approximated animation methods  mainly due to time constraints. Even with technology and talent, the time factor has forced them to make pretty decent animation, but not high quality animation. Given time, they could have refined the methods used to generate the final composites and added the latest cool visual effects to the series. Gerry Anderson himself has said in an interview in Vfxworld that they had problems initially with building skeletons in one package and building faces and bodies with another package which led to inaccuracies. They also could not do water at the start.

Some cases they captured 10 people at a time and 16 cameras capturing the data. Gap-filling packages were used to improve performance in heavily-occlusive situations. At first they used some crew member to do the motion capture, but it looked amateurish and so they had to hire motion capture artists. The results were better as we saw, not high quality, but, quite good.

 

From the pictures of the motion capture artists, we can see the number of trackers that were possibly used for each character. One or two of the motion capture artists were acrobats / martial artists. Counting from the image of the artists, I found about 24 trackers, maybe a bit more but not more than 40.

 

 

I feel that the main reason for the animations being discontinuous is because not all the scenes use the same techniques for the animation. At some places they use motion capture, some places there is key-framing (especially where objects like suitcases, guns etc. are involved) and some places they use a bit of both.