Categories
3DMovement

HLC: Got home, sans phone

This is again a 2.5 D video, the images comes first, then we let the actor play around on it.

Story

Scene 1
eatery @(0,10,0,0,0,0,.6,.6,.6) #1
lady is feeling sleepy @(44,0,-32,225,0,0,.7,.7,.7) #1-#400
legs of lady walks #401-#500
hands of lady is feeling sleepy @(44,0,-28,300,0,0,.7,.7,.7)-@(4,0,-29,305,0,0,.9,.9,.9) #401-#500
Scene 2
stairs @(0,10,0,0,0,0,.7,.7,.7) #1
lady repeatedly walked upstairs @(0,0,-38,0,0,0,2,2,2)-@(0,0,10,0,0,0,.5,.5,.5) #1-#150
Scene 3
bedroom @(0,10,0,0,0,0,.7,.7,.7) #1
lady looking down @(20,0,-35,0,0,0,2,2,2) #1-#130

Ans thats pretty much it! Tell me honestly does it even look like a 1 hour job?

Coming home sleepy, phone lost :'(

And that’s all. This was indeed quite simple. Wanna improve further?

Summary:

  1. This time instead of putting everything at once, we split it into scenes
  2. 3 scenes, each telling the story
Categories
3DMovement

HLC: Mani meets Shani

So, in our last experience, we spent most time getting coordinates together to work well with pictures we had. Hence this time, we decided that this time we will allow for far more human to human interaction rather than depending on Pexels and Pixabay. We choosed 2 models and 4 actions:

Model lady_polka_01 is named Mani in our story and model lady_ts_trouser_02 is named Shani in our story.

20_11 & 21_11: high-five, walk (2 subjects – subject A and B respectively)

20_03 & 21_03: link arms, walk (2 subjects – subject A and B respectively)

environment.egg: from Panda3D model: Panda3D-1.10.6-x64\models

But today we got stuck in a new feature: To be able to perform basic image manipulation on a running video. Hence to keep it short here is story:

  1. Mani and Shani walked towards each other
  2. Mani and Shani high fives (which to our amazement is a verb in SpaCy)
  3. Mani and Shani discussed (I believe topic was why are we not using a nice picture. But as usual, we do not listen to computer simulated complaints)
  4. They walked away hand in hand.
  5. This time we added brightness feature also in the end.

And here is the final video on youtube:

beach #1-#20
camera @(0,0,270,0,0,0,1,1,1)-@(0,0,80,0,0,0,1,1,1) #1-#145
left lady named Mani @(-15,-50,0,180,0,0,1,1,1) #1-#2
another right lady named Shani @(15,50,0,0,0,0,1,1,1) #1-#2
Mani walked repeatedly @(-15,-50,0,160,0,0,1,1,1)-@(-1,-3,0,170,0,0,1,1,1) #1-#108
Shani walked repeatedly @(15,50,0,340,0,0,1,1,1)-@(1,3,0,340,0,0,1,1,1) #1-#108
Mani greeted actively #110-#145
Shani greeted actively #110-#145
camera @(0,0,80,0,0,0,1,1,1)-@(0,90,12,0,0,15,1,1,1) #248-#348
Mani conversed repeatedly @(-5,0,-4,90,0,0,1,1,1) #148-#348
Mani passively conversed #348-#448
Shani @(5,0,-4,270,0,0,1,1,1) #148-#150
Shani conversed repeatedly #248-#448
camera @(0,90,12,0,0,15,1,1,1)-@(0,90,72,0,50,0,1,1,1) #460-#570
Mani repeatedly walked passively @(-2,0,-4,180,0,0,1,1,1)-@(-2,50,-4,180,0,0,1,1,1) #460-#570
Shani repeatedly walked actively @(3,0,-4,180,0,0,1,1,1)-@(3,50,-4,180,0,0,1,1,1) #460-#570 

Actions synonyms used for this challenge

Summary:

  1. Overall, without pictures it is easy to manipulate the models as everything is in 3D.
  2. This time, we tried to reduce the positional markings and add camera movement too
  3. Overall, around 1.1 hours spent, as in again well spent 🙂
Categories
3DMovement

HLC: Walk away to Dance

So, in this challange, we started by looking into few animation files from Subject# 1 to 11. We selected the following:

A20GFLC-A20g01_02: Walking up the stairs

A20GFLC-A20g02_01: Walk normally

A20GFLC-A20g02_04: Jumping

A20GFLC-A20g03_04: Walk on uneven terrain

A20GFLC-A20g05_03, A20GFLC-A20g05_04, A20GFLC-A20g05_07, A20GFLC-A20g05_10, A20GFLC-A20g05_11: Various dance moves

To use them, we downloaded following files:

Timelapse video of clouds and landscape

The sandy seashore by Anete

Cemented stairs by Skitterphoto

Rainbow over bridge by Frans Heerden

Based on this, well I could not come up with a great story, but a simple plot where a person, tired of cement-wood, ran away to nature and danced upto heart’s content

We ran the story in debugging mode, so we can see the story in video itself. But here is summary:

  1. Lady climbs stairs to find a way out of cement rubble
  2. Jumps over to uneven terrain
  3. Finds a way out to rainbow via a narrow bridge, crossing it runnning
  4. Reaches an adobe of nature
  5. Dance (I mean I know it is too Bollywood-ish, but well atleast I desisted from a song)

And here is the final video on youtube:

Final Video
stairs @(0,5,0,0,0,0,2,2,2) #1
lady walked upstairs repeatedly @(-28,-10,-30,0,0,0,2,2,2) #1-#140
lady jumped @(9,0,-15,270,0,0,2,2,2)-@(9,0,45,270,0,0,2,2,2) #141-#223
terrain @(0,4,0,0,0,0,2,2,2) #224
stairs @(0,35,0,0,0,0,2,2,2) #301
lady @(9,0,29,0,90,0,2,2,2)-@(9,0,0,0,90,0,2,2,2) #270-#300
lady jumped @(9,0,39,0,90,0,2,2,2) #250-#310
lady unevenly walked @(9,0,0,0,90,0,2,2,2)-@(9,0,-36,0,90,0,2,2,2) #310-#360
orchard @(0,3.9,0,0,0,0,2,2,2) #361
lady carefully walked repeatedly @(0,0,-39,180,0,0,2,2,2)-@(0,0,-10,180,0,0,.5,.5,.5) #361-#480
clouds @(0,3.8,0,0,0,0,1,1,2.1) #481
camera @(0,-120,-23,0,0,-23,1,1,1) #481
lady is repeatedly walking @(-30,0,-55,90,0,0,1,1,1)-@(0,0,-57,0,0,0,1,1,1) #481-#540
head of lady danced @(0,0,-48,0,0,0,1,1,1) #601-#700
head of lady danced @(0,0,-48,0,0,0,1,1,1) #701-#800
hand of lady danced @(0,0,-48,0,0,0,1,1,1) #701-#800
lady danced @(0,0,-57,0,0,0,1,1,1) #801-#1000
lady danced amazingly @(0,0,-57,0,0,0,1,1,1) #1001-#1100
lady danced performantly @(0,0,-57,0,0,0,1,1,1) #1101-#1200
lady danced securely @(0,0,-57,0,0,0,1,1,1) #1201-#1300
head of lady danced stably @(0,0,-48,0,0,0,1,1,1) #1341-#1440
hand of lady danced stably @(0,0,-48,0,0,0,1,1,1) #1341-#1440
hand of lady danced stably @(0,0,-48,0,0,0,1,1,1) #1441-#1540

Actions synonyms used for this challenge

Summary:

  1. We were unable to find any good audio to go with this
  2. Most time and iterations were spent on a good story, honestly this is the learning here, first decide what you want to show then start
  3. Overall, a bit more than 2 hours spent, as in well spent 🙂
Categories
3DMovement

Everything about Coordinates in Meme’er

Meme’er is based on Panda3D and uses the 3D coordinate system. To overcome the inherent difficulty of modelling for 3D as much as possible, by default, the camera is kept at coordinates of (0, -120, 0) looking at (0, 0, 0). This can be changed at any story line, with same format as for any model. This document assumes we have a 3D model with a synonym “lady”.

To load lady at any point we can type

lady @(x, y, z, h, p, r, sx, sy, sz)

This will load lady at points (x, y, x). On load, the 3d orientation for the lady will be (h, p, r). HPR here refers to Head (angular orientation around Z axis), Pitch (angular orientation on XY axis) and Roll (orientation around itself). A far better and complete clarification of HPR can be found at https://discourse.panda3d.org/t/very-confuse-about-hpr/8935.

The last 3 (sx, sy and sz) are the scaling around X, Y and Z axis. Please note scaling around X or Y does mean scaling with 0 orientation.

Camera, which in Meme’er is just another model follow the same priciples. Except 2 things:

  1. Instead of HPR, for camera, they will be used for Looking At
  2. Sx, Sy and Sz is not used and is reserved for future use.

So far so good. Lets take a example.

Required: The model position coords are (0,0,0, 0,0,0, 1,1,1). The camera needs to zoom past her from 50 points to -50 points along X axis directly 10 points above head.

So, we want the camera to move from (-50,0,10) to (50,0,10) looking at point (0,0,0). This can be achieved as:

camera @(-50,0,10, 0,0,0, 1,1,1)- @(50,0,10, 0,0,0, 1,1,1)

This tells that the camera:

  1. moves from (-50,0,10) to (50,0,10)
  2. LookAt (0,0,0) all the time

Now look at somewhat complex issue. Lets say we have a pic, and as part of motion graphics, we need our car to travel a very undefined path to be set manually (something like https://youtu.be/YQ5qXx4KBtY)

For such scenarios, we have the coordinate module. Basically, the steps involved for YT video are as:

  1. Load the picture (or 3D model as applicable) with correct camera position (if not default)
  2. Using replay frame, load one such frame of picture on Canvas
  3. Manually draw the coordinate path.
  4. Click on “Co-ord”, then “Save Coords” and save the coordinates with correct camera position and looking at values.
  5. This coordinate can be used with syntax like “lady @f(coordfile)” except one issue if the aspect ratio is not same, the coordinates will not match as manually entered vs movement as shown in animation. For this we “transform” the coordinates
  6. In “Co-ord”, goto “Transform Coords”, enter the Screen size as you are using (say 1280, 720). This will be used to recalculate the new coordinates and new coordinate file will be saved as (using example above) “T_coordfile
  7. Now, we can inform that the model should follow the coordinate path of T_coordfile.