Team:Cornell/Action

Revision as of 04:22, 29 September 2011 by Cpaduano (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Protein Pathway Scene

Original  Description:

• Substrate  represented  by  different  geometric  shapes  flow  through  enzyme  pathway
• As  substrate  flows  through  each  stage,  it  is  processed  by  the  corresponding  enzyme
• Substrate  starts  as  a  sphere,  becomes  an  octagon,  becomes  a  dodecahedron,  and  ends  as  a  stellated  icosahedron.

Interpreted  Components:

   1.  Proteins

                                         i.  VioA  ,  VioB,  VioE  –  attractive,  low  resolution

   2.  Substrates

                                         i.  Easily  distinguishable  shapes.  

                                       ii.  Organic  textures  

   3.  Protein-­‐substrate  interaction  animation

                                         i.  Morphmaps  that  look  like  the  enzymes  are  ‘nomming’  the  substrate.

                                       ii.  Proper  timing  for  substrate  entering  enzyme  bonding  site  and  start  of  processing

4.  Side  by  side  animations

                                         i.  Learn  Final Cut Pro

      Components  Development  Summary

         1.  Proteins  

1. Original  attempts  at  creating  proteins  used  mMaya  to  create  point  clouds  from  xyz  data  found  in  PDB  databases  online.  Simple  meshes  were  replicated  into  the  point  cloud  to  create  convincing  proteins.  
2. File  formatting  prohibited  the  use  of  these  proteins  in  our  transition  to  modo.  Modo  also  lacks  a  streamlined  method  to  produce  proteins.  Subdivided  meshes were layered  onto  the  Maya  point  clouds  to  create  proteins  usable  in  modo,  but  the results  were  lacking  aesthetically.    
3. Finetuning  variables  allowed creation of  smooth and  and  high-­‐resolution  proteins  that   usable  in modo.  These  were  also  able  to  be  ported into  other  scenes.  

 

   2.  Attractive  yet  low-­‐resolution  proteins

1. Standardization of the  proteins  was across  our  scenes  to  make  the  animations  feel  integrated. Some  scenes  used  many  proteins  and  yielded  geometric  caches  that  were  unwieldy  to  work  with.  The need  for creating  low-­‐resolution  proteins was thus born.      
2.  Python  scripts  paired  with  the  software  Blender  created  reasonably  realistic  proteins  from  PDB  files  that still  had  low  polygon  counts.      

         3.  Substrates  –  Easily  identifiable  shapes    

                                               i.  Creating  the  platonic  solids  using  modeling  software  was  redolent  of  geometry  classes  and  took  a  surprising  amount  of  time  to  make  look  both  attractive  and  recognizable.  Classical geometric construction techniques were used to make each of the desired polyhedrons

         4.  Substrate  textures  

                                             i.  The  textures  for  the  solids  were  created  with  a  diffusion  layer  set  to  noise  that  produces  a  gradient  of  magenta  colors.  Adjustment of the  frequency  parameters  of  noise,  the  layer  locator,  and  color  parameters yielded an organic looking  texture .    

         5.  Protein-­‐substrate  interaction  animation  

                                           i.  As  it  was  necessary  for  modeled  proteins  to  vary  from  an  ‘open’  to  ‘closed’  position (see   animation),  morph deformers  became an  essential  tool.  Keyframing  the  opacity  of  the  morphmap would  change  severity  with  which the  morph  is  applied.  This  creates  a  very  smooth  and  controllable  transition  between  the  two  modeled states.  

         6.  Realistic  morphmaps  

                                           i.  Original  attempts  at  ‘closed’  protein  states  looked  very  rigid  and  mechanical.  This  is  because  the  components  in  each  group  selected  moved  by  a  similar  amount,  a  very  naïve  approach  

                                         ii.  Future  attempts  utilized  falloffs  to  make  the  opening  and  closing  of  the  protein  look  organic  and  elastic.    

                                       iii.  Getting  realistic  timing  for  the  substrate  movement  and  toggling  of  the  morphmap  was  simply  an  exercise in  patience  and  persistence.    

    

7.  Creating  side  by  side  protein  animation

                               i.  Each  protein  was  individually  animated  processing  its  respective  substrate.  This  made  the  animations  of each  protein  loopable  and  easier  to  animate/render.    

                               ii.  Final  Cut  Pro  was  learned  and  utilized  to  edit  the  three  final  video  clips  to  combine  them  into  a  single  clip  

Summary  of  Scene  Progression

This  scene  was  done  last  as  it  depended  on  the  final  proteins  being  used  for  the  rest  of  the  animations.  Much  of  the work  involved  modeling  proteins.  

1.    Protein  modeling.  (06/27/2011  -­‐  08/10/2011)

                                                 Created  mMaya  versions  of  VioA,  VioB,  VioE,  and  streptavidin  

                                                 Highly  triangulated  ports  of  these  proteins  were  made  to  modo  

                                                 High  quality  proteins  were  made  for  modo  

                                                 Reduced  polygon  counts  using  Maya  tools  reduced  geometric  memory  use  in  modo  by  half  

                                                 Blender  scripts  utilized  to  create  even  lower  polygon  count  proteins  from  pdb  files  

   3.  1st  draft  –  Crude hinge morph maps  (08/14/2011  –  08/15/2011)

                                                 Creating  very  primitive  morphmaps,  an  animation  using  the  same  protein  colored  three  different  colors  

                                                                           was  made  to  illustrate  the  concepts  that  would  need  to  be  utilized  for  the  final  scene.  Keyframes  were  

                                                                           eyeballed  and  positions  were  manually  placed.    

                                                 Despite  rudimentary  techniques,  reception  from  the  team  was  highly  positive.    

3.  2nd  draft  –  Replicated  proteins    (08/17/2011  –  08/20/2011)

                                                 Pleased  with  initial  success,  proteins  were  replicated  and  given  time  delays  to  create  an  entire  

                                                                           screen  of  proteins  processing  substrate.  

                                                  Reception  was  still  positive,  but  this  didn’t  have  the  same  visual  appeal  as  the  initial  single  protein  

                                                                           animation.  It  was  advised  to  go  back  to  a  simple  single  protein  model.

   4.  3rd  Draft  –  Full  pathway,  three  proteins.  (07/11/2011  -­‐  07/15/2011)

                                                       Scene  was  rebuilt  using  the  feedback  from  above.  The  single  protein  perstage  simplified  the appearance  of  the  animation,  making  more  clear  what  was  occurring.  The  morphmaps  were  created with  move  and  falloff  commands  until  they  looked  reasonable.  Each  of  the  three  stages  was  animated individually  and  combined  using  Final  Cut  Pro.  

Reception  and  Review:  Positive.  The  current  animation  illustrates  the  process  well,  and  is  visually  pleasing.  

Future  versions  may  include  a  remake  using  our  very  high  resolution  proteins.