Mtg 22/26: Thu-27-Mar-2025

Outline for Today

Reflection Models, Texture Mapping, Area Lights

Administration

Today

For Next Meeting

Wiki

Link to the UR Courses wiki page for this meeting

Media

Transcript

Audio Transcript

  • Struggling with it, and then I looked at it.
  • Okay, so we're at meeting 2226
  • so
  • reflection models I'm
  • textures and materials and
  • so this is chapter 10.
  • This is in chapter 12.
  • I want to mention these other things, because
  • the assignment that's due and
  • thinking about what we're going how we're going to finish up The
  • class. I'm
  • so meeting
  • 324, 2526 and
  • so are we shooting to get to chapter 12 by the end of the
  • course is that this is indicated.
  • Yeah, I'm wondering if, if we Touch on chapter 11 as Well. I
  • Bless you. You're
  • too quick for Me. You
  • Yeah, I think we'll dabble in chapter 11 as well.
  • So I suppose Are we just lightly touching on those three
  • subjects? Do you have another three kind of lectures? Yeah,
  • 11 looks a little bit shorter, Though I
  • CS, so I would invite your input for questions that would be
  • final exam questions that would Be a fair assessment of Your
  • learning this semester. Applause,
  • so I'm sorry for missing last day's meeting.
  • Let's go over the quest, The quiz questions From that you
  • so what are the three major classes of behaviors that come
  • out of The electromagnetic theory of light.
  • Semiconductors Do
  • conductors,
  • and I guess, insulators. I
  • The answer you have on the quiz is dielectric, conductor and
  • semiconductor just has two of them listed. I know I'm just
  • looking at it right now, since I got that one wrong, but yeah, me
  • too.
  • Yeah, I have dialectics, conductors and semiconductors,
  • because I put semiconductor is one word and you put it as two
  • words.
  • It might be because of the order, oh no, I
  • put mine in the same word.
  • I put commas in between them, and I have semi and conductors
  • one word,
  • okay, I will check my Answer and regrade them. Oh, yeah,
  • that makes sense. I
  • I also didn't pluralize or I did pluralize conductors, so
  • arguably, I have none of those parts.
  • Yeah, I realized all night.
  • Okay, let's look at the question. Since there are
  • questions about the question, I Oh, I
  • didn't put wild card At the end Here. I
  • I will check
  • it right now.
  • I will check other issues as well.
  • If you have other quizzes that you think are Right, that got
  • marked wrong. Let me Know If
  • what impact does the difference between indices of refraction of
  • neighboring materials have on reflection? I
  • Are you asking for? Like, the like mathematical impact, I
  • wasn't quite sure about what you're exactly asking for on
  • this one, what I wrote was the differences between indices of
  • refraction is scaled by the square root of the two indices
  • of refraction. This is reflection, which I don't know
  • if that's what you were looking for, necessarily.
  • I also I had that, plus I had, intuitively, this could be
  • understood those results, the radiance is differential, solid
  • angle being compressed or expanded due to transmission.
  • I was thinking about if there's a greater difference between in
  • the season refraction neighboring media
  • to reflect this reflection will be more intense. So diamonds are
  • more reflective than glass, because there's a greater
  • difference.
  • Diamonds are more reflective than glass. I guess you don't
  • often see a nice diamond mirror like that's not something you
  • just commonly come across. Diamonds are so small it's hard
  • to see. I imagine it would be really hard to make a diamond
  • mirror. I don't think it's impossible, though, with
  • synthetic diamonds, i i
  • because otherwise that's really expensive Mirror, mirror on the
  • wall, who's the fairest of the
  • Wall? Could use it in like a telescope or something.
  • You said, the greater the difference, the greater
  • difference More intense function. I that we Were saying
  • yes,
  • what is Snells window?
  • If you're at like, the bottom of a pool, and you look up, there's
  • gonna be an area that's clear to you, and then the rest of it
  • will be, I guess, a little bit occluded. You basically have a
  • nice little pothole of like vision that is clear that that
  • actual like frame, that window is Snells window, right? Yeah, I
  • don't know if that's the most articulate way of saying it, but
  • it made me want access to a pool.
  • Can't say I've ever opened
  • my eyes. Don't you like chlorine in your eyes? Do
  • where the pain goes away in a few days. I also
  • don't stand too well.
  • Less Salt i
  • So cells window is
  • when you're underwater, you'd look up, there's a circle.
  • Inside the circle, you can see the blue sky and the palm trees
  • or whatever else you have around your pool,
  • outside that outside of that window, you don't, you don't get
  • to see out. You see
  • reflection. What is that last word after Water? Slash, air
  • interface.
  • You're welcome. You
  • What are three important geometric effects to consider
  • with Micro facet reflection models,
  • masking, shadowing and inter reflection.
  • I so I
  • just shadowing.
  • I masking happens when?
  • When the light hits the surface, but You catch you See it? Yeah,
  • I
  • How do you like my pictures?
  • Some of the mountains. I don't know what happened to them. I in
  • their example, they use three identical mountain ranges. You
  • got a little bit liberative With your creativity there.
  • Thanks. Creative license. Exactly.
  • So why does pbrt use a specialized BS DF for air and
  • fur? Because
  • otherwise it would be extremely expensive computationally, but
  • it also make important sound like difficult for Monte
  • Carlo, yeah. So there's a lot of parameters that
  • there's a lot of ways to improve the efficiency of the sampling
  • for hair and fur, so they have A special one. That's what I was
  • looking for. I Sorry.
  • Could you say that again you're looking for
  • just could you just repeat what you just said? There? Sorry.
  • So there are
  • using COVID diffuse materials and
  • more generic BS df and
  • we can do the do hair and fur, but it's very computationally
  • expensive, and I
  • so it helps to make,
  • it helps with efficiency, to to create a special BS DF, so it's
  • more efficient
  • and produces
  • better results. And
  • thank you.
  • Well, I know how hard it is to resist the urge to just have one
  • thing work for everything like that, just have one sampling
  • method that works for everything, and you just try to
  • do it as much as you can, but then it messes up other things,
  • and then, like, Oh, yeah. Now hair works really well, but now
  • all of a sudden, leaves are breaking. You just get a bunch
  • of like, weird things, as lame as it is, just to Break Your
  • coat up and create other specialized forms i
  • What is the purpose of a go Neo photo meter or go Neo, reflect,
  • Do meter?
  • So the answer you were looking for was it measures the SDF?
  • Yeah. I talked about how it had two separate, like, two or more
  • arms for separating the camera and the light, so you can
  • measure something from all the different directions. That's
  • what I wrote. But I suppose that's a very indirect way of
  • saying the same thing.
  • I love. So go Neo,
  • the etymology of that word Gonio is angle do.
  • So it's the same the same root, and as we see in polygon, so
  • polygon would be many angles, and
  • so a couple of those sections had the Don't you can skip this
  • on your first reading. So I didn't, I wasn't looking for a
  • great deal of detail. I just thought, Gonio photometer,
  • photometer. Either
  • way, I thought, I thought it was cool, and I understood its
  • intended purpose right away, because there's a lot of value
  • in not having the camera and the light just actually attached to
  • each other, especially if you're trying to measure the way light
  • hits a surface. Well, that's very limiting. And to have all
  • these separate arms I can picture, you could
  • computationally have the arms all move around and like, take
  • it from different angles, and you just record it with record
  • it, just sort of capture that you can get a lot of data out of
  • that from any material or whatever.
  • Yeah,
  • really good for digitizing new materials, like maybe even,
  • like, storing them to using, let's say, a game or something,
  • just storing a texture by capturing lights. That would be
  • a very interesting way of talking about
  • You. So I wanted to show you to start of rendering of the manger
  • sponge was linked from Last Meeting And
  • it'll be a good problem. No,
  • I don't think so.
  • This is
  • a finished version. I
  • so this generates a segmentation fault. So I'm not sure how many
  • millions of triangles are in this. I didn't do the six. I
  • CS, I did the Fifth Level, not the sixth level. I
  • so you can see the starting point looks pretty good already.
  • I
  • The sponge appears. It's just the spaces that need to be
  • cleaned up. And you can see on the histogram,
  • there's a peak
  • at 1.065
  • so the value should as we refine the
  • image that
  • maximum value will go down and get us to the point of,
  • ideally, We'll be in the range of zero to one. I
  • so I want to show that to you because of
  • the relationship with Sir Pinsky tetrahedron, you
  • can See it's going down And
  • so this is
  • an example that I did with a triangle mesh. So it's really
  • one triangle, and I put them, I use them twice. So, so first
  • three values are point one, second, three values are point
  • two, and the third three values Are the third point, so indices
  • 01, and two and
  • so you don't Have to have them. I did two triangles back to back
  • that we could just do 01 and two for The triangle and
  • so here I've combined
  • the model with spheres in the gasket, dot, PBR, key file with
  • the triangle, so The triangle is the same size as the gasket. The
  • it. So if we're doing for generating the tetrahedron with
  • a triangle,
  • well, let's talk about the gasket. If we're doing a
  • triangle, how do we create more detail in the triangle mesh?
  • Does that make sense? I
  • Are you referring to? Like the interior faces, like,
  • if you were to take the triangle and let's say you slap four
  • triangles, one of the bottom and the three side ones you would
  • have, like An empty triangle, is that What you're referring to?
  • I so if we do it in 3d the tetrahedron. So this is made out
  • of four triangles, Then the first step would be to make the
  • and we get four. And then
  • we get I got you,
  • but with a triangle, I
  • We can construct the model by making the triangles so we're
  • shrinking the triangle
  • by a half
  • and bringing it to vertex and
  • so we can start With The
  • so we can Take the triangle then we can define three new
  • triangles and
  • so we have three triangles. So 00102, that's a new triangle.
  • And then 011, and one two, that's another triangle. And
  • then we can do 02122, to make a third triangle. And we can do
  • that recursively.
  • Does that make sense? I
  • What do you observe these two images.
  • The one on the right looks good.
  • I I want to vote the one on
  • the left. Well, the one on the left is trying its best. It's
  • It's definitely having troubles in figuring out what it wants to
  • display it looks like.
  • Yeah. So
  • the one on the right is filtered. So what do we want to
  • filter? So this is a texture. It's the same texture, and it's
  • one sample per pixel.
  • So we want to filter out
  • what wet from
  • our texture And
  • sorry,
  • yeah, it's noise,
  • Maybe even more specific. I
  • so how do we
  • how do we know we have a
  • of enough samples
  • to reconstruct an image
  • when it looks less like The left one or more Like the right one.
  • I two times. Highest frequency,
  • highest frequency. Thank you.
  • I so if we're sampling at a rate that isn't
  • that doesn't encompass all the high fruit, the highest
  • frequencies in the signal we're trying to reconstruct. I
  • so we can think the frequencies above the above the sampling
  • rate will be noise.
  • Does that make sense? What's the word between we and Two one.
  • Thank
  • you. You're welcome. I
  • so I'm going to filter out I'm
  • so
  • done some filtering on The texture, and it looks a lot
  • better.
  • And you said anything about that highest frequency is noise. I
  • Yeah, so the frequency that are above
  • the samples and
  • so If we're sampling at a rate, at
  • rate, r so,
  • so if we sample the radar, then we can reconstruct frequencies.
  • That have a maximum the maximum frequency we can reconstruct is
  • R over two.
  • So anything above r over two So
  • so that makes sense. I
  • I think It would be useful if there were code examples or
  • input file examples to generate these examples.
  • That would be good. That's that seems like a really easy quality
  • of Life thing. Yeah.
  • So I wanted to jump to talk a bit about area light sources.
  • Here is the mythical point light source, Deluxe. Illuminating a
  • dragon like mythical Dragon, mythical Dragon, Yes,
  • so we have Very sharp shadows and
  • here's a small disk area, Lake source. Let's compare it to
  • white light source
  • so we can see some terrible soft the other word,
  • let's look At a large disc.
  • There. You're almost non existent. How
  • there's still glass in here, guys,
  • the correct answer would have been way outside and
  • we're still in Class. Can you guys leave? There's still COVID.
  • All right? You now, I feel like we should go on for 10 minutes.
  • We're not done yet. I like,
  • sorry. So,
  • what do you want us to read for next week? Yeah,
  • probably 1011, I think you said,
  • let's look at What
  • let's look at chapter 10,
  • and we'll talk about purlin noise as a way to generate
  • textures. So if you want to do wood and some other very
  • realistic textures there are. So we'll look at some details of
  • that, okay,
  • thank you for today.
  • Have a good weekend.
  • Sorry. Any questions about the assignments, please post into
  • the class discussions.

Responses

No Responses