Categories
ADVANCE & EXPERIMENTAL Compositing

PROJECT NUKE: GARAGE

In this project we were given a video of garage in which we have to carry out various VFX techniques such as rotoscopy, color grading, color correction, composition etc. Along with this we have to add a 3D objects in the scene which would be located inside the room that we can see in the video below.

In the beginning, as per the video, we were given the distorted grid, which was exactly according to the plate. Thanks to this grid we were able to take its reference and undistort the plate itself. Undistorting the plate is extremely crucial as it flatten ups the plate resulting into a better working output. With the help of this undistort node we can add the exact distortion back to the video after all the editing and composition is done so that it is back to its original way.

DISTORTION GRID
UNDISTORTED GRID

After undistorting the plate, I started by 3D tracking the video. 3D tracking is the key to the videos in which we have to do an addition of 3D Geometries. With the help of 3D track we get these track points (refer to the image below) on which we can set up the object. For doing so we have to select the point and create a card/geometry which would be aligned in the 3D space. In these 3D space we can add as many as 3D objects/images on the plate. All of these geometries as well as cards are then combined together into a scene node. These scene nodes collect all the data of 3D space and combine it together to get the merged result.

CAMERA TRACK POINTS
WORK FLOW OF TRACK POINTS AS WELL AS ALIGNING OF 3D OBJECT
ADDED 3D GEOMETRIES

After 3D the entire sequence, the next process of rotoscoping and cleanup. Thanks to the 3D trackers the process reduces the time it usually takes up for a normal/manual rotoscope. The main process begins by adding a frame hold. These frame hold should illustrate the clearest vision of the cleanup place, as the cleanup can look blurred if taken on the wrong frame. By selecting the point where we want to add a card (specifically on the area where we want to do the cleanup process), we can create a card. On this card, we do our desired cleanup. I have specifically used roto paint, as it gave me much more flexibility for doing the cleanups. After doing the desired cleanup we copy these onto the card which we created prior. Project3D is the node that we use to copy these cleanup roto paint onto the card. After this, we add all the cleanups to the scene so that we can get the final look at our cleanup by combining them.

PRE-CLEANUP
AFTER-CLEANUP
CLEANUP WORKFLOW

As we have already tracked the video, we can carry forward with rotoscoping without worrying about cleanup. For rotoscoping, I used a similar workflow to that of cleanup.

In this technique, we add project3D, card, and scanline render prior to the roto which I added later on in the case of cleanup. Even though I had proper and well-defined tracking points, I had to make slight changes to the roto, so that it matched the borders of the walls throughout the video. These slight changes were done a few frames apart, having approximately a 15-35 frame gap in between. This was a necessary step so that the roto sticks to the wall and it doesn’t slip off.

The difficult part of rotoing this particular video was the lamp handle. Due to the poor lighting at the back, the handle of the lamp disappears at some point which was difficult to see. For this, I increased the exposure of the screen and drew an approximation of the future position of the handle.

And after adding all the roto to a scene I got the result given below. (please refer to the video below)

ROTO WORKFLOW
ROTOSCOPING
Workflow of Manipulation of the Color Passes
Color Grading with the help of different Passes
Merging of the Machine with the Plate

A powerful computer program called Nuke is utilized in the film as well as broadcast industries for compositing and visual effects. Making 3D shadows is just one of its many features, which gives an object a feeling of depth and realism. These shadows can be created through a variety of methods, including shadow maps and ray tracing. In order to produce precise and lifelike shadows calculates the path of light rays from a virtual light source to the object in the picture. Shadow maps imitate the effect of shadows using a 2D texture map, which leads to quicker rendering times but possibly less accurate results. With Nuke’s 3D shadow tools, artists can modify the shadows’ intensity, color, and softness as well as their location and size.

Over here, I Created a scene and added 3D elements to track points with the help of axis which determined the Track pints in the “3D Space”. For getting these 3D objects, the color according to the scene, I Created a spot light, according to that of the scene.

After the step of lighting the scene, Comes the main part, i.e. Casting shadows. Basically, for getting this Shadow, we need to create the shadow pass. For this, the step was to do a setup of various cards/grids on which the shadow would be cast. these grids are aligned onto the track points, according to the position of walls and ground of the plate. Following this, I created a new “Shadow Mask” which would make a new “Shadow Pass” for all the 3D objects that I imported and aligned to the scene.

With the help of this step, I could extract the “Shadows” of all the elements (3D objects), and merge it with the actual scene, to get more defined output.

Importing and aligning of 3D objects.
Work flow of 3D elements and shadows
Newly created Shadow Pass
Final Output of the Project
Categories
ADVANCE & EXPERIMENTAL GROUP

GROUP PROJECT: LOW-FI FINAL OUTPUT

This is the final output of our group project.

Categories
ADVANCE & EXPERIMENTAL Compositing MAYA

TRACKING IN 3D EQUALIZER

3D Equalizer is by no doubt the worlds leading 3D tracking solution, for merging live action footage with digital visual effects. This software specializes in 3D tracking and footage and gather points in order to create a geometry which would be used for Match Move purposes.

In the beginning after importing the image sequence and applying proper settings that are needed, we start by setting up points to track the footage. For this Tracking purpose, we need a will defined image where we can set the point which wont result us by giving jitters. For this reason we start by changing the color settings and getting a sharper perspective, which can be used for holding points at one place through out the video.

Later on, after the step of enhancing the color, we start applying tracking points. For this, we have to ctrl + left click on the desired location,and hit the track button. Sometimes, which there is a jerk or the desired point vanishes, we have to manually end the point or manually track the footage frame by frame. It is a bit burden some times, but we get to good results at such times.

COLOR CONTROL AND POINT DATA
INDIVIDUAL TRACK POINT
TRACKING POINTS

After the step of tracking, we have to solve the points. This step is crucial as this helps us in getting a proper camera animation. Hence, having a smooter line with out anr jitter is crucial. With the help of Deviation browser we get the information through which we can analyse and pop out the track points which would be giving us the errors or a high value. A maximum value of 0.5-1 is needed to get a smoother camera flow.

After the step, with the help of these track points, we can build basic shapes, which would resemble the grid and its positions. This can be obtained by meshing the points using legacy. And after this step we can export the file to maya, where we can utilize this grids into making the match move needed for production purposes.

matchmove obtained after modeling and places basic grids on to the points.
Categories
ADVANCE & EXPERIMENTAL Compositing PERSONAL

PERSONAL PROJECT : METEOR SHOWER FINAL OUTPUT

This is the final output of my personal project i.e., Meteor Shower

Categories
ADVANCE & EXPERIMENTAL PERSONAL

PERSONAL PROJECT PART 4: COMPOSITING

To make it simpler for me to colour-grade each layer individually while compositing, I rendered each layer separately for compositing purposes. So, after reformatting the plate, I color graded the Pyro. Later, I gave the meteor and pyro some glow so that it appeared to be glowing fire, and I combined it with the debris.

However, after importing the clouds, I gave them the appropriate transformations and colour graded them to match the sky plate. I came upon this picture of the sky plate online. Three to four flocks of birds were flying in the frame and there were numerous anomalies in this sky plate. For this reason, So for this reason I photoshopped the the image while making it a clean plate. 

Following this, I rendered it out after adding and adjusting the blur to the meteor and clouds.

workflow
3D CLOUDS (HOUDINI)
MERGING OF BOTH
COLOR CORRECTION
FINAL SCENE

Categories
ADVANCE & EXPERIMENTAL PERSONAL

PERSONAL PROJECT PART 3: RENDERING

In order to simulate fill light and rim light for the scene during rendering, I first took two distinct distant lights and set them up. In addition to this, I used an ambient light to help me gather light from all directions. Another way to use these lights is to take an HDRI image that you like and produce lights with different colours. But for this instance, I refrained from the use of HDRI.

Lights and Camera (ANIMATED)

Since I am more accustomed to the Mantra renderer, I used it for my scene. In the beginning, I started by putting up the correct cameras and rendering path for the image. Later, I changed the settings and increased both the samplings and the pixel samples, which improved the render quality.

The main section, or forcing objects, follows this. In this section, we must manually inform the renderers of the objects they should render. As you can see, for the force object section of the renderings of debris (sparks), I indicated the path of the sparks, however for the forced matte, I used the Pyro path. I would receive a layer of sparks as a result of this procedure, resulting into the sparks that aren’t visible not to be rendered. 

Categories
ADVANCE & EXPERIMENTAL PERSONAL

PERSONAL PROJECT: PART 2: METEOR, PYRO AND SPARKS

I started by building a simple model of a meteor. I took a sphere, sized it appropriately, and then added a mountain node as a result. This mountain node introduces noise to the surface by adding deformities to a geometry.

meteor

After this, I added an attribute named “attribute randomise” while scattering various points throughout the sphere’s surface. Every single point is given a random scale or characteristic thanks to this node. Later, I added spheres to these places and a boolean to make the surface of the model void.

work flow for meteor

After the modelling phase, I began the project’s primary duty, which is the creation of FX. I started by animating the meteor as necessary for this. I then used the point Velocity node to determine the velocity and add irregularities to it. Volume have the features of density, temperature, velocity, and in certain cases burn, which are the essential characteristics required for pyro emission.

velocity

After the modelling phase, I began the project’s primary duty, which is the creation of FX. I started by animating the meteor as necessary for this. I then used the point Velocity node to determine the velocity and add irregularities to it. Volume have the features of density, temperature, velocity, and in certain cases burn, which are the essential characteristics required for pyro emission.

paritcles for volume
volume with includes density, temperature and velocity
workflow for volume

The voxel size of a pyro stimulation is one of its most important components. The size of the voxels is a crucial aspect of stimulation quality. Slice thickness and pixel size both influence voxel size. The effect of stimulation performs more effectively the lower its value is.

setup for pyro

I attached each of these volumes to a Pyro source node after this step. The input geometry of the Pyro Source SOP is transformed into points that can be used as sources for pyro and smoke simulations. The generated points can be rasterized and imported into the desired DOP fields by the Volume Source node after being added with the specified characteristics by this SOP.

By changing the values of density, temperature, and velocity, I began pyrotechnics. This phase is essential since it is the centre of attention. This phase allows us to distinguish between the pyro of a candle and the pyro of an eruption of a volcano.

The process of manipulating the pyro to increase its potency begins after this phase. Due to this, I changed the properties of Turbulence and Disturbance (which refers to a fluid’s departure from a smooth (laminar) flow), which means that we added noise or, disturbance to the flow. Vorticity, which is an irregular eddying motion, is what defines turbulence. Adding shredding to this project was a must as Shredding is the main method for adding the kind of high-frequency details to the stimulation.

Additionally, I enhanced this stimulation by including a gas vortex confinement node. This gas confinement node enhanced the visual appeal and further shredding the pyro effects by adding even more small features.

Workflow for caching

I eliminated all the attributes that I would no longer require after Pyro solver. Temperature and density would be the necessary factors. The texturing of Pyro requires these properties. Additionally, I volume compressed this stimulation while I was conducting it, which helped to shorten the stimulation’s duration by removing unnecessary voxels and characteristics. File caching comes after this stimulation.

Later with the help of Pyro bake volume I added the texture to the pyro.

Final Pyro
Addition of texture

The secondary components of stimulation—sparks—come next. For this, I created a straightforward pop system and added pop force to give it variety in particle emission and velocity. Following that, I added a trail to these particles, turned it into a line, and then textured it appropriately.

sparks
Final
Categories
ADVANCE & EXPERIMENTAL GROUP

Low-Fi: Project: WRENCH

In the task of modeling a second item, I pondered over my options and ultimately settled on a wrench or spanner. These tools are particularly useful in providing a secure grip and mechanical advantage when it comes to applying torque and turning objects, such as nuts and bolts. Furthermore, they serve the purpose of preventing these objects from turning, thereby ensuring that they remain securely in place.The Second model that I had to model was any choice of “tool”. Hence for this reason I started going with a wrench or a spanner. A wrench or spanner is a tool used to provide grip and mechanical advantage in applying torque to turn objects—usually rotary fasteners, such as nuts and bolts—or keep them from turning.

wrench or a spanner

wrench or a spanner

It never ceases to amaze me how a seemingly ordinary block can be transformed into a true work of art. I recently embarked on a project that involved utilizing a variety of basic shapes, including boxes and cylinders, to create an intricate spanner. To begin, I took a rectangular box and employed the smooth select feature to bend the upper portion of the spanner. This was accomplished by selecting any edge, vertex, or face and pressing the key ‘B’. From there, I used the extrude tool to create the upper section of the spanner.

Upper part of spanner

Using the same technique, I went on to create the second head of the spanner. However, to ensure that the upper portion would fit seamlessly into the void, I had to extrude some faces and merge the edges to create a single vertex. Initially, I had considered using the Boolean tool, but ultimately decided against it due to the potential for “N-Gons” to occur. These are faces with more than four sides, and can lead to unusual rendering results if not handled properly.

For the handle, I started with a cylinder and skillfully beveled the corners to create a beautiful capsule-like effect. With the extrude tool, I was able to create the final, simple handle, which perfectly complemented the intricate design of the spanner. All in all, it was a challenging but incredibly rewarding project, and I’m thrilled with the results.I am amazed how a single piece of block can be turned into an art work. In this project I have manipulated all the basic shapes such as box or a cylinder.

Upon completion of the UV cleaning stage, I proceeded to import the model into Substance Painter, a software program that was initially unfamiliar to me. As a newcomer to the world of texturing and layering effects, I found the interface to be user-friendly and highly creative, requiring only a basic understanding of texturing to produce visually stunning artwork.

To start, I decided to apply a “rough steel” texture to the spanner, which I then augmented by adding several scratches and paint marks to give it a more rugged, weathered appearance. I found the existing brush effects to be highly effective in streamlining this process, allowing me to achieve the desired outcome with minimal effort. Moving on to the handle, I used a Carbon Fiber texture, adjusting the tiling size to suit my vision and bring out the desired effect. I then used the brush tool to add more roughness to the handle, enhancing its overall aesthetic appeal and creating a more immersive experience for the viewer. Overall, I found the entire texturing process to be both rewarding and creatively fulfilling, and I look forward to exploring more advanced techniques and tools in the future.

Categories
ADVANCE & EXPERIMENTAL GROUP

Low fi: Project: Window

As I set out to undertake the task of designing the window for the Low-Fi project, my mind was filled with a clear and concise vision. The window needed to be both practical and functional, while exuding a sense of antiquity and simplicity.

To achieve this, I began the process by utilizing the boolean function on a cube with bevelled edges. This action allowed me to create a surface intersection, which I then used to outline the window’s border. From there, I took the central faces of the surface and pulled them downwards towards the middle, creating a border that would house glass panes.

iNNER_WINDOW

Next, I used the same booleaned surface to create the exterior of the window. I selected the outer edges and extruded them with a slight offset, which gave the window a sloping appearance. This step was essential in achieving the desired effect of antiquity and simplicity. Finally, I merged the extruded edges to create a single, cohesive unit.

Outer_Part

Initially, my plan was to design a set of three windows. However, after consulting with the group leader, it was decided that a single, unified piece would be more suitable for the project. The resulting design was not only practical and functional but also visually appealing, perfectly embodying the desired sense of antiquity and simplicity.

Initial Concept

Work_Flow1
Work_Flow2

Upon completion of the UV cleaning stage, I proceeded to import the model into Substance PainterTo start, I decided to apply a “rough steel” texture to the window, which I then augmented by adding several scratches to give it a more used, rusted appearance. Overall, I found the entire texturing process to be both rewarding and creatively fulfilling, and I look forward to exploring more advanced techniques and tools in the future.

UVs
Categories
ADVANCE & EXPERIMENTAL Compositing

Week 11 Nuke: Patching & projection

In this week we started by the technique of PATCHING in Nuke. These patches are helpful for removing he objects from the plate itself, where these patches are 3D by itself. We add these patches on the Card taken from the 3D trackers which we did in the last lecture. There are various techniques for doing this.

Having the card that is taken from the specific 3D Track is the most important aspect as all these patch are about to sit in the given card / points.

The rotos are done on a particular frames, where we have to be very cautious about the frame that we select as these frame hold carry on to the whole plate. After cleaning up / rotoing (in this case i have roto painted the plate to get my desired look), these roto is then send on to a frame-hold which carries it to the entire plate with the help of card.

work flow for Patching (Frame hold method)
After cleanup

We also looked on how to do the patching with the help of UV method. In this method, we stabilize the video. This is done by transferring the plate onto the card, and the the plate is changed in to the format of UV. After which we can roto paint in again, and carry the rest to get the patching done.

UV METHOD
After copying the UVs.

PROJECTION WITH HELP OF POINT CLOUD

Pointcloud helps in making the scene 3D through which we can add grids/cards to the points.These point help us to get a 3D visual representation of the plate. With the help of these points which we got after analyzing and tracking the points we can add 3D objects to these points.

Visual representation of Point card

We can also view these points into a 3D geometry, with the help of poisson mesh. Poission mesh converts these points into a typically 3D object.

3D visualization with the help of poission mesh node.

Along with the poission node we can also visualize these points by baking these point created by the point cloud generated. For getting these points we have to bake the points which can be obtained by selecting the vertices of these points and baking them into geometry mesh after grouping the points selected.

Visualization with the help of point bake node.

Model builder is another technique by which we can do a proper visualization. Poission method maybe easier than model builder technique, but its not quite efficient when compared. In Model builder method we have to add cards and grids same like how we used to do the corner pin method.

FINAL MATCHMOVE