Mtg 16/23: Wed-06-Nov-2024

Outline for Today

Lighting and Shading

Administration

Today

Reminder that our next meeting is Wednesday, November 13 (Monday is Remembrance Day)
Opportunities for Feedback (open until 13 November at 1 minute to midnight)
- Anonymous – formative feedback for Dr. Hepting
- Not anonymous to Dr. Hepting – for participation – rate the midterm
I neglected to mention that I will be using gradescope.com for marking the exam – you likely received a strange email from them – it has a lot of nice features that protect you from having to read my handwritten comments. When I am finished marking next week, you will be able to ask questions about the marks you received using the platform
shadedCube
shadedCube2

For Next Meeting

finish Chapter 6 and start on Chapter 7
Submit your response to this meeting before noon tomorrow
Take the quiz before the start of our next meeting

Wiki

Link to the UR Courses wiki page for this meeting

Media

Transcript

Audio Transcript

So what kind of shading have we used so far? What's the
way to describe it?
Would You like a vowel?
Wow, yes. I
So what other information Can we Add?
Models and
light. Yeah, We can add Lights. We're
when can we add to describe how the light is going interact with
a cube? Let's say I
so we can add
material descriptions, or we have the model of the surface
so we can get different effects when we Apply the
when that Surface is lit. I
so We can think about
describing The
so we have a
plate that comes back from interacting with the surface. We
can describe in terms of ambient, diffuse and specular
components and
so shading, a proper shading, gives us a sense of three
dimensional quality jumping out of the computer screen. So we're
talk about the types of lights we can use and interaction with
materials, and
then we'll talk about how that can be implemented on graphics
hardware.
So the first one looks like a circle. Does the second one look
like a sphere? I
It's not terribly convincing to me. I
so we agree with this assessment that the light material
interactions on each point have a different color or shade, so
we don't see a flat color. That's the variation of the
surface.
So in addition to light sources and material properties, we can
also take into account the location of the viewer and the
regulation of the surface so
We can think about scattering light in the scene and
light strike surface A. Sun is scattered. Sun is absorbed. Some
of that scattered light strikes B, some scattered Sun absorbed,
some of the scattered light strikes a and
So that's keeping track of A lot of interactions and
so we
we have a description of scattering and absorption, but
it doesn't have a general solution, and we can ray trace
using ray tracing With perfectly reflective services and again,
an accurate model.
So you can see there's a lot to keep track of. If we're
going to do an accurate rendering, we're it.
So in this picture of global effects, we see the light
source,
Shadow, multiple reflections, and here's A translucent
surface. So in that case, some length reflects and sun passes
through the object The
so global illumination isn't suited to graphics hardware. So
to the graphics hardware pipeline.
But we're happy if we can make it look reasonably good. So
there are some techniques to be used to simulate global
illumination effects. Light strikes an object is partially
absorbed, length of strikes an object is partially absorbed,
partially scattered, reflected. The amount reflected terms of
color and brightness of the object. The surface appears red
under white light, because the red component of the light is
reflected and the rest is absorbed. The
reflected light is scattering back, and depends on The smooth,
smoothness and orientation of The surface You
So what kinds of legs are available for computer graphics?
I
Are you talking about like directional and positional
or Yeah, so I
so
if we specify This a position of a light in the scene and
so it's, we have the leg Coming from a point, so it doesn't have
any I need. Geometry is just a point, and light is emitted in
every direction, if
we specify direction
it's like when we move
object off to infinity, whereas if we're looking at something
close,
the directions are quite different from the eye to
different parts of the scene. If you move that off to infinity,
then the projectors become parallel and
so when We think of a directional light, the light has
a
so what's a directional light source that we see most days?
Yeah,
so if we take a white light source and restrict the
if we restrict the directions in which it is emitting, If
so then we have a Spot like I
so we talked. We've talked about These ambient Light. I
so we can Cover up. We can we make stuff look good or look
better by adding some ambient light to account for i
to account for
terms that are too hard to compute in your graphics
pipeline And
so
if we have a smooth surface and
the more it reflects like A mirror. So we get the light
bouncing off of
if The surface is rough, we get more scattering. If
so we have diffuse, the specular, the ambient
components, And we have four vectors.
So L is the direction of is the location of the light source to
the viewer. That's V then we have a normal to the surface,
and we have the perfect reflector. So this is that's the
direction in which light would be. If the surface was a perfect
reflector, the light would bounce in that direction. I
So the normal is defined determined by local orientation
of the model. The angle of incidence equals the angle of
reflection. The three vectors must be co planar. So we have L
to the light source,
and the perfect reflector is the angle of incidence, which we
find by looking at the angle between l and n, and
then the reflection is that same angle. And
it so we can think about another type of surface, inversion
surface, perfectly diffused reflector, so it scatters. It
doesn't have
it's the opposite of the perfect reflector that we saw.
None of light reflect is proportional to the vertical
component of the incoming light. The Yeah.
So we're taking the cosine between the light vector and the
normal vector.
So most of the surfaces are neither ideal diffusers or ideal
reflectors. Smooth surfaces show specular highlights due to
incoming light being reflected in directions concentrated close
to the direction of a perfect reflection and
so we're interested so we can see specular highlights in the
region Around the reflected, the perfect The perfect reflection
reflector you
so fun proposed
a term that makes the highlight stronger when the viewer is at
or near the perfect reflector,
and it's a
so that highlight is going to be much less apparent If the viewer
is not looking in that direction
So we can
try different values To get
the effect we want you.
So to handle ambient light,
we have an intensity term for The ambient light
and the reflection coefficient light from the
so the Light from a point source is proportional to the distance
of the source from the surface, and then we can add A term like
this to attenuate and soften The effects of the light.
So we have in the farm model nine coefficients for each point
source,
nine absorption coefficients and the shininess coefficient.
So for each color component, we add the contributions across the
sources and
so This is where the halfway vector becomes more efficient
and
so here's An array of teapots with different parameters In the
blend faux model.
So we have the light vector and the Uber vector specified in our
scene, so we can compute R from L and N. But how do we determine
n for simple surfaces and
so how do we deal with the
Timing In the
again, suspension, disbelief, that's a cube. I
so On the surface of the queue and that
Normal is
easy to define
what happens when a group,
then we
have a bunch of triangles to make up a surface. So the marble
on the surface of the triangle is clear, but What about where
it meets up with other triangles? So
like, what about that edge or This corner?
So we can interpolate the
so the normal to the sphere. So,
at a point on the sphere, we have a 10, the tangent plane,
and that allows us to do the cross pride, to get the normal
so Here are some details about competing normals For triangles
you
so
we can this is a little bit about calling Back faces. You.
So when we
do the vertex per vertex, shading calculations are done
for each vertex, which makes sense.
So We Could we
so we could use that
attribute approach Changing
the color, or we could do The shading and
in The vertex shader and
so we have the normal for the triangle, but in the case of a
sphere that we can see here, we want different normals at Each
vertex, even though it's not quite correct mathematically, We
so if we set the normals
at each vertex, then we get smooth shading, But we see the
edge of the sphere Model and
so what's called Grow shading involves averaging
The normals of the triangles around The vertex and
so with corrosion, we want to have that normal vector for each
vertex so we can apply The modify the limb form model at
each vertex And the interpolate vertex comes across each
polygon, the
So where can we get Highlights if we're doing garosha,
Because what we're doing is we're doing
the whole calculation for
the vertex so if the light The sprinkler highlights correspond
with the vertices of our wall, then we'll see the highlights
come out, because in the vertex shade, we're computing vertex
color that gets interpolated by the fragment shader.
So
in Grow shading, it all depends on the intensity values at the
vertices, because those values are interpolated.
So Fauci says find the vertex normals and
like we've done before, but interpolate the vertex normals
Across edges, and interpolate edge normals across The polygon
And
so what's the advantage of Fauci compared to grow shade? I
What's An advantage and what's a disadvantage. I
let's say
that's the point where the specular highlight would Be most
pronounced.
So in Grow
shading, we're calculating intensities at the vertices, and
we're interpolating the value so we don't get the highlight and
but in Phong shading, we're interpolating the normals. So in
the fragment shader, we're doing the calculation with an
approximation to the surface normal. So we can get we can
capture these highlights and
Okay, so that's an advantage. What's a disadvantage
is this a little bit more work
or a lot more? I
because instead of doing the four vertices of a quad and or a
triangle strip, we're doing the calculation not four times, but
once, for each fragment, for each pixel. Does That make
sense? I
that's a bit more complicated. I so
it's
not clear where the light is going either. Well, it is
because you can see the surface. The
it. So anyway, in the shaded cube, oh, we're out of time. I
knew that because of the
so we're doing the lighting and the faux calculation in the
vertex shader, and then fragment shader. We're just assigning the
color to the fragment color. Thank you for today. Have a good
rest. Of your Wednesday.
Good weekend, a
good long weekend.

Responses

What important concept or perspective did you encounter today?

I thought the discussion on Phoung and Guroud Shading (Sorry for spelling I couldnt see the board great) Was very interesting along with the different pros and cons of each and how they both look and are calcuated
We discussed the Phong model of lighting in computer graphics. One important perspective is that we are trying to find a balance between realistic lighting and computational efficiency. We want results that look good while still being fast to compute.
Shading and materials - the different properties, specular, metallic, and diffuse.
Diffuse and Ambient, Normals
Gouraud and Phong shading as well as specular surfaces and reflections
We discussed the lighting, types of surface, Gouraud shading, and Phong shading.
Lights, Shading and types of lights, Steps to incorporate lights in a scene. Goroud and Phong Shading
we learned about lighting, and compared the gouraud model to the phong model
Today, I learned the concept of lighting in 3D. The halfway vector is an efficient method for computation. Specular surfaces reflect light in a narrow angle, creating a shiny effect, while diffuse surfaces scatter light in all directions. In Gouraud shading, a normal vector is calculated for each vertex, while Phong shading computes lighting at the pixel level for more detailed results.

Was there anything today that was difficult to understand?

Was there anything today about which you would like to know more?

Site built: 2024-12-12 00:01 -0600 • Give feedback about page