I know this is not generally the most difficult of exercises, but the reason I joined this class, the reason I’ve been returning to perspective exercises, was my inability to rotate and place skulls at different angles. It’s something that would crop up frequently in other courses and I’d return to perspective each time. I would be disheartened again and again at not being able to get ellipse sides or craniums to match the direction I intended, the face, etc. Last night was the first time I’ve been able to do it. (!!!!) They’re referenced and aren’t perfect but it was only using the axis and proportion knowledge from the square/cube exercise that I was finally able to get them to point in the way I was meaning. Used the square to think how it changes during a rotation and applied to the object as a guide. Big milestone for me. Thank you. :3

At about 6:20 Marshall states, "this is not a square in perspective," without any explanation of why this is so. I've been playing around w/ Zolly and squares can get pretty distorted and it was not obvious to me why this is not a square. After looking at it a bit more, I could see that, within a reasonable field of view (say 60 deg), this cannot be a square. Given the way the lines converge on this shape, the longer edges would need to be more foreshortened for this to be a square. But this did not satisfy me. So I took the not-a-square and reverse engineered the R/L vanishing points and the diagonal vanishing point. From those, I derived the station point and field of view. It turns out that, for this shape to be a square in perspective, it would have to be way out side the field of view. So, would it be correct to think about it this way: If you make the shape satisfy the geometric requirements of a square in perspective, it is not in a useful field of view. If you put it in a useful field of view, it cannot satisfy the geometric requirements of a square. Thus, it is not a square in perspective. I know this is kind of a academic point, but I find it helpful to understand what I'm doing on a conceptual level to help guide me on a more practical level.

This is much ado about the simple effect of varying the distance between a stationary picture plane and a moveable station point. Incidentally, viewing a perspective from the station point that created it correctly foreshortens all distortions in the projected image, effectively making them invisible.

Inspired by the image Carlos sent and the 3 one-point examples, I saw this in my mind. I hope it’s not totally wrong.

Marshall, thanks for this video explanation. It helps me very much with the earlier Proximity section of Part 1. The three room thumbnails are great. Thanks

Help…t-the perspective exercises…they’re following me everywhere (no, but jokes aside, this video was illuminating.) If I understand things correctly, it’s all about the station point/viewer. If the cube’s near corner is 3 units away from the viewer because we are super close, and the far side is 6 units away, the far side is twice as far. So you get this huge difference and dramatic perspective. Near stuff is much closer and thus bigger, and far stuff is much further. When the object (e.g a sheet of paper) is very close, its left and right edges are far apart in your vision. Your eye has to look across a wide angle to see the whole sheet. That’s wide lens. It captures a huge angle of the world and squeezes it into the picture frame. VPs squeeze close together, corners jut out, objects near the camera are super big. The rays of the cone are extremely oblique which contributes to that distortion. If the near corner is 300 units away and the far side is 301 units away, that difference is basically nothing. The cube looks flatter. And that’s telephoto? Which captures a narrow angle of the world and puts that narrow slice into the picture frame. Rays from the cone are quite straight and thus perspective is much less dramatic. I tried it by pointing fingers in front of my eyes, when I look at a window from far away my fingers are straight. But the closer I get the more I have to spread them creating an oblique angle/wide cone to capture the same window. And in the example above you made the cube’s edge touch the picture plane so we can steal the size and then build the cube in the correct dimensions…!

Ok. If I understand this correctly, obtuse become right angles (appear as perfect squares) in a close-up (When we are too close) and is the reverse with acutes, they become right angles when we are looking from afar (zoom-in). So the obtuse becomes more acute and vice-versa and the right angle is the middle point, the in-between in that process. Correct? As for the close-up and far away squares, is the same logic for an endless corridor, the space around us seems more open than the space on the distance and as we go forward the space that looked compress opens up to reveal even more compress space at the distance that we couldn't see, and the new area around us becomes a mirror of the space we were before. That's why it never seems to end, is the same effect of been so far away the moon seems to follow you and you can never get close to the horizon line, is just as we perceive distance so we would just end up circling the planet. SO TAKE THAT FLAT EARTHERS!! Did I straight away from the point of the lesson? Am I missing something?