Mtg 6/26: Thu-23-Jan-2025

Outline for Today

Geometry and Transformations

Administration

Today

For Next Meeting

Wiki

Link to the UR Courses wiki page for this meeting

Media

Transcript

Audio Transcript

  • It's so the computing resources at the
  • University of Regina are not happy with me lately.
  • Does anyone else feel like they're being singled out for
  • rough Treatment?
  • Let's see if I can find a screen i
  • Oh, welcome.
  • You said, I mean suspicious activity.
  • Got so the one thing I can the edge room I can
  • deal with with my cell phones, hot spot. I'm waiting on a
  • ticket for that, but the other one last night and today, my
  • password manager, so I'm using one password. Anyone familiar
  • with that one? I i have a 19 character password, but last
  • night, I decided to not accept my credentials. So then I called
  • it service, and I have somebody had somebody who was helpful,
  • but a little bit condescending.
  • And so when it happened anyway? So I reset my password, and it
  • seemed to work. And then today, when I logged into your courses,
  • I had the same problem, and I called it support, and I got the
  • same person and said, if you're confused, I said, Listen, I'm
  • not confused. I have anyway. This is an IT person at the U of
  • R, Yeah, you sound shocked. I
  • didn't know that we had an IT.
  • Help. Help Desk,
  • but it's the service it's now it's the Service Desk. IT help
  • desk. I can't
  • believe we even paid for one, but we have one. I worked there,
  • but there being an IT help desk. It was like one person does kind
  • of check out? Well, there's lots of people in that office. I
  • think it'd be funnier that's just like, one day, just sitting
  • there, like, all the time, just like, oh, going through set my
  • passwords, like, Oh, it's you again. It's the whole
  • person that's just like mumbling to back. Like.
  • So what I learned today is if I copied the
  • password out of the Password Manager and click Login, it
  • worked. So I'm I'm not sure if there's a strange character
  • somewhere in the process, or this particular person with whom
  • I spoke was not interested in my problems with a passive password
  • manager. He was convinced that the problems with me,
  • the other apartments that are like that.
  • Yeah, anyway,
  • for sure, it's not limited to my car Statistics.
  • Not CS, so we'll say onward and upward. You.
  • So I was talking about the comes
  • from a bow and arrow.
  • Yeah, and
  • a mistranslation along the way, apparently.
  • Anyway, this gives you a few resources to look at if you're
  • interested in etymology. Etymology is study of the
  • history of words. Entomology is a study of bugs, so there's one
  • letter difference, I believe I
  • see if my black marker that I brought is going To be
  • any better. So
  • So do you agree with that statement? Chapter Three is too
  • big to say, just read chapter three.
  • Okay, so I will endeavor to break things down so
  • we have a bit more depressed. I compressed so we have more focus
  • on What we'll be talking About the Next day. I
  • so we don't finish in chapter three today, we'll cover next
  • day, and then in future, I'll break down chapters by sections.
  • So I'm pleased to report somebody came to my office hour
  • and somebody posted to the class discussion, and I replied right
  • away to encourage that behavior so everyone's got the software
  • sorted. Now. Is that true?
  • Should we go? Does anyone not
  • is anyone not willing to put their hand up in
  • class? Who's not willing to put their hand up in class?
  • I guess I should ask who's willing to put their hand up in
  • class. Okay, I think That's what everybody saw. I
  • so keep you abreast of developments, whether we get
  • pbrt installed in the labs. I
  • so let's talk about the quiz questions and
  • it's so appealing to use This mouse, but it's not
  • connected to my computer, So
  • okay
  • quiz, so?
  • So where are the two things we have to specify.
  • We're dealing with array,
  • Origin and Direction. Yeah,
  • and to cover the matching question, what kind of
  • thing is the origin and what kind of thing is the direction,
  • point and vector. Yeah, point and vector.
  • So how would you? How do you see the quizzes so far? Are they
  • for those who did 315 last semester? Are they better than
  • what I did in 315 okay, I
  • so generally, the quizzes are tied to the readings. So when
  • you're dealing with questions that come up, you can look at
  • the readings to find the answers. So it's a way to
  • incentivize you to look At the things I've asked you to look
  • at. I
  • that's not the word I was looking for.
  • I that's not quite the question was, I guess
  • I could read the question, what is the unfortunate turn of
  • terminology with normals not
  • necessarily normal?
  • Yes, they're not necessarily normalized. Yes.
  • I try and do more alliteration. I do
  • normally. I
  • so what is a stereo steroid in
  • I don't understand what it is. I'm trying to figure out it
  • seems to be a unit index, but that's as far as I figured out.
  • It's the way of like measurement of sphere specifically, it's
  • kind of similar to how, when you're measuring like a circle,
  • you have like radians, like a PI is like something pi. Radians
  • for, like a single, like a whole circle is measured in radians,
  • yeah. So if you apply that to a sphere, since you can, like,
  • integrate the circle into a sphere, you can turn it into
  • stir radians, which is just basically a solid radian. So the
  • amount of area that if you project a ray into, like a
  • closed loop, makes, like, that amount of steradians.
  • Thank you.
  • Makes good sense, actually.
  • So when we're talking about the in 2d we can think about an
  • object, the angle, the arc length of the projected object,
  • and we'll maybe see those
  • look at a couple figures. So in ve,
  • instead of the planar angle, we have the solid
  • angle, which measures The amount of the sphere that's covered by
  • the
  • by the object that's projected. So steradian is a way to measure
  • angles within A sphere and
  • maybe I fill that in a bit more.
  • Okay, so before we get into
  • the Text stuff I
  • promised A gold sphere I
  • looks like a Christmas ornament?
  • Does it look like a gold Christmas ornament?
  • Though? Yeah, I think so too. So let's
  • there are lots of parameters in this description that we're not
  • going to deal With today, but I
  • so see the
  • physically based and
  • the parameters at work here.
  • So we can see there's a reflection of the
  • checkerboard pattern here that's clear enough, but there's about
  • 3% of the pixels here that are out of gamut or clamp
  • zero to one.
  • So let me I me
  • show you what the file looks like.
  • I just changed the sphere that was the dielectric, or the glass
  • sphere, and I made it into the gold sphere.
  • I will separate these a little bit better and
  • put them on to make them available in the Dropbox. I
  • it so it's the same as a
  • nose of audio is connected to the HDMI. I
  • so I took this material from an example of a gold plated
  • kangaroo. We can look at that file in a minute as well.
  • So the end, these are spectrum data. So there are files here
  • that are referenced for Gold.
  • Au is gold and
  • So let's try Changing i
  • Four. I'm
  • so we made the gold rougher so it looks more like an ornament,
  • doing it the other way. I
  • it so it's a much smoother surface. Does that make sense?
  • Yeah, it's really reflecting the sky so heavily.
  • So let's see if, what if we
  • Increase the samples here. Sorry, Sorry. I
  • so this has more out of gamut pixels, so you can see
  • here the errors or
  • the pixels that are out of gamut and
  • so would you want to, like, clamp them to keep them in
  • range?
  • Well, they do get clamped here. So the report is
  • three. That's a little bit less than two or 3%
  • but the pixels are
  • have been clapped, and
  • so that's where we see these specs
  • come up. So I
  • part of what we'll explore is how to set the
  • parameters so we we get
  • better images,
  • so We don't end up with
  • Without of gamma pixels. We
  • Okay, so we had Killer root coated Gold. Yes, I
  • can you See I should
  • change this. I
  • Is it first? Visible enough? I
  • so what I've done for for the gold, The simple,
  • I've taken the
  • coded conductor description i
  • from here.
  • So this is applying the code of the
  • conductor to a killer
  • room. So Let's I
  • up. Let's run This. We're
  • it. So once we get through chapter three, we'll
  • look at the user guide for the pbrt and understand what's being
  • specified in these images, in these files, and how they're set
  • up, and then use that For our first assignment.
  • I Let's look at the gold one.
  • And it look
  • like a solid gold compared to the other one I
  • yeah, there I was only COVID and
  • gold. Yeah, I'm gonna just look At the file now. I
  • so this is
  • conductor material, as opposed to coded conductor I
  • it Okay. What do you think about that?
  • Can I see the other one back and forth?
  • Should be just down a little bit. I see it there. Says one
  • beside it,
  • just above your mouse, there's a highlighted thing just above
  • that with one, okay, okay,
  • yeah, it looks like A golden ring this time, instead of an
  • ornament. I
  • Okay, so I I
  • can see the
  • this coded one doesn't Look as
  • looks different than The Golden One. Anyway, I
  • so I just want to highlight the camera
  • implementations in tbrt, so We're using the differentials.
  • So the differential is,
  • if that's
  • one ray going out, then we have the difference
  • in x and in y.
  • So the rays are going out from the camera one pixel apart. So
  • we have this, this one, and then we're talking about using the
  • differentials to capture the information, more information
  • about The
  • the area surrounding the array of that pixel And
  • so bounding boxes.
  • Have You encountered that idea before? So
  • pen is for pen is horrible.
  • Maybe I should use it this one, maybe that's going to last a bit
  • longer bounding boxes.
  • So I just want to think about the general case,
  • why bounding boxes might be helpful in processing a scene
  • that has billions of triangles in it. We don't have to start
  • with billions. We could say millions or 1000s, even,
  • well, I can see if, like you're trying to decipher
  • array or something, to see if it interacts with that particular
  • bounding box at all. I and then if it interacts with that box,
  • then, due to further calculations, if it doesn't,
  • then you can skip it.
  • Was that your answer?
  • It was similar. Okay, I
  • i have something that could also be added. All the vectors that
  • are going upward will probably not collide with anything,
  • unless you have a roof. So you can have a similar bounding box
  • that's anything that's scattering upward probably won't
  • collide with anything. So you can just stop calculating them
  • at a certain point.
  • You could even, like, clamp anything that shoots up too far
  • from The light source. I
  • You so we can talk about in ray tracing, that
  • we send out the rays from the Camera and we measure their
  • interaction if we have a perfectly reflective sphere,
  • then we bounce a reflection off The sphere. And if there are
  • more spheres, then we create more rays that reflect.
  • If we're if we have a model that's very reflective has a lot
  • of reflectors in it. At some point we could go on forever and
  • so this use of boundary boxes in 3d is discussed later. Another
  • example using them In the Software Is The
  • that's a question criterion to stop, other than
  • depth of weight of race, because, if so number of
  • reflections, you could have it stop after a certain amount,
  • because we talked about like infinite back and forth
  • Reflections, but maybe you don't want more than eight.
  • Alternatively, if it's not necessarily the number of
  • reflections, but if we assume that a ray is basically just
  • like a photon, have a little walk around, there's like a
  • predictable curve. It's like the inverse squares. It gets farther
  • and farther away. It loses energy. But if it reflects off
  • something that's not perfectly reflective, it loses some
  • proportion of energy. And you can clamp it so that after they
  • lower bound, it just stops computing because there's no
  • additional info being added.
  • Yes, i
  • Does that make sense? I
  • so this is the planar angle.
  • So that's the angle from a point p of an
  • object.
  • So from the top to the bottom, that's the angle and S is the R
  • length on the
  • unit circle the
  • so when we have a three dimensional so
  • I'm Going to go from the unit circle to the unit
  • sphere. Then we're projecting the object onto the unit sphere,
  • and then have the total area s of the projection, and that's
  • the solid angle instead of the planar angle. And the solid
  • angles are measured in steradian
  • so we can think of polygons being projected onto a sphere.
  • That becomes spherical polygons and
  • so again, think About edges of the polygon being great
  • circles. I
  • so a great circle is one that includes the center
  • of the sphere. So it's not just off to the side. That's not a
  • very technical definition. And
  • so there are a few different spherical parameterizations used
  • in the software we can do spherical coordinates and
  • so there's two angles and
  • one in the plane, the x y plane, and then another
  • one In
  • the Z axis.
  • It's quarter after two.
  • So thank you for today, and we'll finish the rest
  • of Chapter Three on Tuesday. And then we'll get into using the
  • software in more detail. Okay, thank you for your attention
  • today. Have a great weekend.

Responses

No Responses