Thread #7915551
HomeIndexCatalogAll ThreadsNew ThreadReply
H
Trying to do this exercise is filtering me.
I think I might be too retarded for constructional drawing and won't even make it to box mannequins.
+Showing all 68 replies.
>>
>>7915551
Post your attempt
>>
>>7915551
what do your attempts look like? Odds are that you wont get it fully right the first few times
>>
>>7915551
Doesn't the example have grids right there showing exactly what he did to get that result? How you could fail to follow that? There's probably a text that accompanied this to explain the exercise too?
Anyway I'm not sure this is something that's actually useful for learning to draw but it's basically do these steps and you get this.
>>
>>7915702
Funnily enough I am trying this too currently and no, there is surprisingly little information.
Most if not all people showing it either don't explain it at all, leave crucial parts out or just do it wrong or inaccurate themselves.
Some don't even follow their own instructions like "draw through".

Even in OPs pic the question that comes to mind: if the box actually rotates, the corners should go outside the bounding box, like this it would basically be zooming out too as it rotates.
>>
>>7915551
drawn grids and then draw on top of them
>>
>>7915551
You haven't unlocked 3d vision yet.
I can do it effortlessly now since I can visualize the grid in 3d.
You're still stuck looking at a 2d canvas
It's actually one of the simplest beginner exercises
>>
>>7915717
> It's actually one of the simplest beginner exercises
>>
File: accuracy.png (576.5 KB)
576.5 KB
576.5 KB PNG
>>7915712
The thing is that this "copy shit accurately" shit just doesn't matter
Look at the shit KJG draws. The guy can barely draw a circle let alone divide it into equal parts. Or maybe he can but just won't. Has he even heard of a ruler? His perspective is all over the place and he keeps hiding it behind the fish eye effect. His art still looks good.
Don't worry about accuracy. It's all about understanding space, three-dimensionality and form.
>>
>>7915588
>>7915674
I will give it another go and post the result
>>
File: 7915551.jpg (30.2 KB)
30.2 KB
30.2 KB JPG
>>7915551
Freehanding a 3-point box. Pay attention to the drawing order.
1. Draw a circle. It doesn't need to be perfect, just make it as wide as it is tall. This serves as a reference point for you, a way to constrain scale.
2. Envision the square you want to draw, then draw the edge that would be closest to you. As the first mark, placement is arbitrary. Things become less arbitrary as we progress.
3. Draw the other two edges that would be logically parallel to this edge. Since this is a 3 point perspective, the lines should be angled such that they all seem to converge on a point somewhere far off from the canvas.
4~5. Repeat the process with the second set of edges. Center one first, all should seem to converge.
6~7. Same with the third set of edges, although placement at this point is not arbitrary. Use the corners formed by the first two sets of edges for placement.
8. Draw the final, pulling back on some of the more extreme angles to reduce distortion. Unless you want it.
2 point and 1 point perspective boxes are drawn much the same way, except in 2 point pers you have only 2 sets of edges that converge, and only 1 set of converging edges in 1 point pers.
>>
>>7915551
how do people have trouble with this? literally just trace it first, then draw it freehand, then keep drawing it freehand while paying attention to the harder squares. The answer is RIGHT THERE just draw the fucking cubes holy shit

do normalfags just not see the cubes? Does it look different to them?
>>
>>7915734
Speaking as someone who is just moving out of /beg/ stage.
The difficulty with this isn't seeing a cube or not.
It's that even the examples and the things experts draw for this are wrong. As another anon points out, having it stay inside a square bounding box is fundamentally wrong that is not how cubes work.
When I looked at some youtubers doing this I traced their lines and their vanishing points did not in fact converge to the same lines as they should. Or didn't converge at all.

The difficulty, as a /beg/, is judging whether your own wrong is good enough to look as intended and aesthetically pleasing, or not. And that is pretty difficult to get a feel for without a feedback loop.
>>
>>7915551
For me, it's drawing the biggest face and projecting corner's off of the corners with perspective skew in mind. not even real perspective skew, just a general idea that they should converge in the distance.

I had a lot of cubes with overlong edges but eventually I got the hang of it and now the other faces besides the biggest face are pretty much the right length


I believe that you can do it too!
>>
>>7915731
thanks for the tip, I will give this method a try too
>>7915785
Oh so you think more about rotating a single plane in 3d first rather than rotating 3 edges?
>>7915588
>>7915674
Here is a fresh attempt but in the style of >>7915712
>>
>>7915840
>>7915840
>Here is a fresh attempt but in the style of >>7915712
Brutal
>>
>>7915871
Yeah, I know, it's grim
>>
>>7915551
i'm still pretty bad at drawing boxes and shapes in general but i feel like it's better to advance with the knowledge i have rather than spend weeks doing drawabox. we'll see what happens at the end of the year
>>
> this is supposed to be easy even for /beg/s
man i'm ngmi
>>
>>7915840
Looks good to me. Just do the op drawing the same way. A ruler helps too
>>
File: uhhhhh.gif (495.3 KB)
495.3 KB
495.3 KB GIF
cubes suck
>>
File: 7915551b.jpg (31.3 KB)
31.3 KB
31.3 KB JPG
>>7915840
To test if your boxes (much less cubes) don't break logic, you can find midpoints and midlines and see if they match. This is easier to do than it seems.
I would also suggest you use a thicker pencil/pen point to start. Thicker lines are more forgiving, and you won't (and shouldn't) have to go over lines multiple times. Make one stroke, make it fast, and draw past where you think the corners would be. Worry about the line's angle rather than where it should begin and end.
>>
>>7915717
How someone is even supposed to learn this 3D vision?
I am drawing for more than 5 years and still I can't rotate figures, or even redoing them from memory...
>>
>>7915717

i like to draw a flat square on the ground, at an angle, before i attempt drawing anything in the image. the blinding white void is a major hurdle to creativity.
>>
>>7916374
Train yourself to visualize a cube in your head and rotate it, I don't know what to tell you .
I don't know bro, I just imagine it and rotate in my head. Guess you might have aphantasia...
>>
>>7916374
Figures are much more complex . The way the Koreans do it is they built it up from simple forms. You learn to rotate cubes, then different body parts. Then combined body parts until you memorize the human body
>>
>>7916548

you can see shit in your head if you think about it? are you insane or something? is it like dreaming?
>>
>>7916552
Yes, I can visualize entire scenes, figures etc... in my head. I find it absurd that other people can't lol. At school people always said I was daydreaming but of course I was I can live out an entire seperate life inside my head
I go to sleep every night visualizing Sydney Sweeney giving me a foot job
And yes it is like dreaming but less intense.
But it has its drawbacks I get nasty intrusive thoughts that are quite vivid.
Like maggots eating my insides and stuff but it doesn't disturb me as much as when I was a child
>>
>>7916044
>>7916252
thanks brehs
>>
>>7916374
>>7916552
Just wanted to chime in as someone who initially thought/feared he had aphantasia
Turns out I didn't.
Yes Aphantasia is of course a thing. And it could of course be that you have it.
But mental visualization is something that is very much learned and can be trained. Humans even have to learn basic real life perspective.
Mental images are just a common thing to develop via certain games as children etc.

And you can get extremely good at observational drawing while still remaining unable to draw from imagination. My comparison point would be music and instruments: being able to play from sheet music vs. improvisation and composing. Some highly skilled classical musicians are very much unable to improvise in the slightest.
>>
>>7916548
You talk as if you were always able to do it
>>
Blender may help.
>>
>>7918070
Not if you're in orthographic, friend.
>>
>>7918102
>As you wish, friend.
>>
Drawing gets way harder than this lmao
>>
>>7916160
Naw,cubes can be fun
>>
>>7918102
>>
>>7918102
>As you wish, unknown anon friend.
>>
>>7918240
>>7918241
I try 6 times to post the orthographic view, the 1st time I forgot the pic, it's not funny, I try to help.
>>
>>
>>7918263
Rotating left and right seems clear, we move the left and right vanishing points, same as in 2 point perspective.

But what happens if I want to rotate this box as it is there towards us so that the top faces us, would I only move the bottom vanishing point up on the center line?
Or the bottom and right vanishing points up?
>>
File: file.png (718.2 KB)
718.2 KB
718.2 KB PNG
Is there a cheatsheet/blueprint/guide or whatever you wanna call it like this one that has dots I can practice with, but without the already added in scribbles?
>>
>>7918236
I'll contribute to the thread since no one has posted The Secret™.

Just pick the 3 lines closest to the viewer and then rotate them left or right, build the box around it. The only other piece is just keeping your proportions in check which requires some practice. I'd also suggest doing the cylinder version of this since those are way more useful for actually drawing foreshortening.
>>
>>7918399
The "secret" was posted here >>7915712 and if you're still confused you can follow https://www.youtube.com/watch?v=Fh9QhuwtqIU&t any other method of teaching a rotated cube is a psyop
>>
>>7918403
if you need a grid youre ngmi tbqh. Try my method its the correct one.
>>
>>7918406
Life's too short to play games. I can freehand a box without a grid but if you force me to mathematically put boxes next to each other and rotate them that's a retarded waste of time and I will use the grid and you will seethe.
>>
>>7918416
Idk isn’t the point of the exercise to build actual foreshortening and construction skill not just to game the test? to each their own ig
>>
>>7915924
what brush
>>
File: file.png (312.4 KB)
312.4 KB
312.4 KB PNG
>>7918403
Even if you're following this method this video is a bad example because it completely neglects perspective on the vertical axis. These are not cubes.
>>
>>7918629

Her sense of 3point is a little off and her boxes at the top are squished because she's drawing trad and or doesn't care either way, because it's still a box...not a "cube" which can easily be corrected digitally. I'd still follow the method knowing I can correct her own mistakes rather than goof around like an idiot with some chinese wu-han instruction on cubes where they (not you) can see magical protractors that aren't there or an Indian telling me I gotta go back with a ruler and "check my perspective" after drawing 200 cubes or telling me to draw through the cube while they're using a grid w/ snap ruler enabled or aren't even doing it themselves. I have a brain, I've been doing this for a while and I've seen the traps that these people do to keep you stuck forever drawing endless drills.
>>
>>7918648
And again, this isn't a practical application anyway. Just open up sketchfab with a cube or hold a physical cube in your had and you can reference that when freehanding boxes if you're stuck. The grid seems to be useful if you want to put an object inside and have a reference of what it would look like turned around, like your own character. "Getting a sense of turns" is a byproduct. Nothing matters until you actually draw a box in production.
>>
>>7918648
>>7918649
If there's one thing to take away from this exercise it should be an intuition for plane convergence at different viewing angles. If you're satisfied with any box at all then you might as well just skip the exercise (this is also somewhere I think drawabox falls short by the way). You don't need protractors and you don't need to draw a perfect cube, but if you want to place an object in a box the important thing is to have some vague idea of how quickly or slowly these edges are going to converge and you should at least be able to freehand somewhere in the ranges of 0-30, 30-60, 60-90 degrees. An effective demonstration of the exercise should do that reasonably well and the result would look more or less like cubes as they do in the OP.

So sure, you can correct it, but if your point is that you don't care about understanding the perspective in the exercise then you're wasting your time whether you use a grid or not. In that case, as you say, just use an actual cube as a reference whenever you need it.
>>
>>7918263
Lmao this brings me flashbacks to erik olson's vids

>30 30 for a total of 60!
>>
>>7918659
>but if your point is that you don't care about understanding the perspective in the exercise

I do care about “range” I don’t care about forcing myself to put boxes together, side by side, and determining the degree and angle. It’s exercises like these that filter beginners from ever moving forward with course material because the instructor makes these starter lessons harder than it actually is. I’d try to find other artists going through the same courses and 90% of the time they drop the course because they get to parts like this that filter them due to poor explanation and unintentional deception.

On the same topic, Krenz has been shared around here since 2017 widely available and I don’t think that many anons leveled up following him…but he’s recommended constantly. It’s odd.
>>
>>7918853
at any rate…this >>7915712 is the same as >>7915551, can you not see the method works?
>>
>>7915551
Never once have I seen a professional, working artist recommend something like this. Maybe if your goal is to draw cityscapes with exaggerated perspective for comic books i could see working on this. Observational sketching and gesture would yield such better results. This is also boring as fuck and discourages one to keep drawing
>>
>>7918853
You don't have to force yourself to do it, I don't think it's for everyone, but if you want to learn to use perspective intuitively this is probably the simplest exercise to develop that skill.

>>7918854
OP has a grid too if you look closely. I'm not the person who was critiquing the grid, I was just saying that if you're presenting a demo of the exercise the bare minimum, critical element is to make sure the cubes read as cubes, since that's exactly what it's supposed to be training. Use whatever approach helps you to get used to that, or don't and just use references for your perspective instead. Completely depends on your goals as an artist.
>>
>>7918853
There's a very annoying trend in this board to think that construction and anatomy will solve all your problems, when neither is crucial for making a good drawing.
>>
>>7918874
What would you suggest to do instead? I have indeed failed to learn to draw the human figure despite all the grind...
>>
>>7918877
You should learn construction and anatomy but Vilppu is right that getting good at things that sell a pose (gesture & acting) take a lifetime of practice. Also draw from memory and sketch people from life as much as you can.
>>
>>7918430
juju line. it’s been deleted so you’ll have to check the brush thread here
>>
https://www.desmos.com/calculator/34138e810f
i made this a long while ago and it helped me, could help you if you play around with it and see how it works, although theres alot more to it too
>>
>>7918938
Damn anon, thanks a lot I was actually for something like that.
Genuinely appreciate it.
>>
>>7918974
im suprised no one made and shared this type of stuff before, there is also way easier ways of explaining perspective than the perspective normies let on, even the very technical formal perspective books, they are showing you only 2d representations of perspective, not the 3d geometry behind the 2d constructions they're doing for 1, 2 and 3 point perspective that wouldve made it so much easier to understand
>>
>>7918985
If these drawings are illustrating your point, that's much more about form than perspective. Perspective is how a 3D object/scene projects onto a 2D surface so the 3D geometry by itself really can't get the main ideas across.
>>
>>7918996
the image was unrelated so my bad
here's what i mean with the 3d geometry, this is related to the first desmos graph
https://www.desmos.com/3d/hrairepdas
>>
If you want more details you can add me on dc
Littlestlive
>>
>>7919002
Okay yeah my mistake, that is much more what you were describing. I'll be honest though, that's pretty involved and I'm not sure that would help clear it up for someone who just wants to draw. Also seems difficult to describe in a book, but a very nice project.
>>
File: image-4.png (23.2 KB)
23.2 KB
23.2 KB PNG
I've been in the weeds so it's hard for me to understand someone else's view on the subject atp so I feel you, this shit consumed me for like 3 years, I'm just trying to actually use the things I understand atp since no one really cares about theory if you still can't draw

Reply to Thread #7915551


Supported: JPG, PNG, GIF, WebP, WebM, MP4, MP3 (max 4MB)