Minor axis points to the Vanishing point when ellipses are in a vertical position (in a 2pt perspective)
What about when ellipses are in a horizontal position like my picture? Specificaly in a 2pt perspective, they are perpenticular to the horizontal line.
> by the way this is 45/45 perspective with 60 COV (you can see the COV on the side of the picture)
>>2296396
OP here. Wtf part of my message got deleted??
IGNORE FIRST MESSAGE AND READ THIS ONE :
In a ONE point perspective drawing, when you sit an ellipse on the flat surface (horizontal) the minor axis is always perpenticular to the horizontal line.
In 2 points perspective, minor axis, when ellipse are in vertical position, always point to the vanishing points.
MY QUESTION : Where does the minor axis points of an ellipse points out, in a 2pt perspective picture, when it is sitting on the flat surface (horizontal surface?)
Downwards vertically. If there is distortion, to the 3rd VP.
>>2296686
This, in one and two point any ellipse lying on a horizontal plane parallel to the floor has a perfectly vertical minor axis.
So you're image is wrong OP, go fix it.
>>2297030
>you are refering to a 1 pt perspective
No I'm not.
When the ellipse is on a horizontal plane, its minor axis is perpendicular to the horizon line in BOTH 1pt and 2pt, just like perfect verticals are still vertical.
Cone of vision and what grid you use are completely irrelevant and don't change a thing in that regard.
>>2297030
>it's a 2 pt perspective, so where is the 3rd VP?
Where the minor axes converge.
The minor axis appears perfectly vertical when there is no distortion. As you go farther away from the center of the cone of vision, distortion is introduced whether you like it or not. This becomes more apparent the nearer the station point is to the horizon line.
>>2297030
>it's a 2 pt perspective, so where is the 3rd VP?
He's saying IF it were 3pt perspective (i.e. 2pp plus an additional vp straight up or straight down), then the minor axis for a circle on the ground plane would point to that 3rd vp, instead of being perfectly vertical.
>>2297030
>you are refering to a 1 pt perspective where the minor axis is always perpenticular to the horizon line. My image is a square grid 45/45 inside a 60 COV using 2pt perspective.
Go re-read Perspective Made Easy. 1pp and 2pp exist under the same parameters.
If you have a cube seen straight on but rotated (let's start at 45 degrees, so we're looking at the edge). In this scenario you have 2pp. As you keep your vantage point the same, but rotate the cube along the z-axis so that it stays in contact with the ground plane, your 2 vanishing points move along the horizon. One vanishing point moves towards the center of vision at a certain rate, and the other vanishing point moves away from the center of vision at an accelerated rate. This continues so that once you have the first vanishing point at the center of vision, the second vanishing point is at infinity. This scenario IS 1pt perspective: One set of parallel lines for the cube point to the vanishing point at the center of vision, the other set of parallel lines point to the vanishing point at infinity - meaning they're perfectly horizontal.
The minor axis is ALWAYS orthogonal to the plane containing the circle in perspective. It's always functionally pointing to a vanishing point; if that vanishing point is at infinity, then all minor axes for all ellipses drawn side by side on the same plane will be parallel to each other.
Rotating the square used to draw the ellipse along it's orthogonal axis does not change the ellipse. This means the ellipse is the same regardless if you're dealing with 1 or 2pt perspective for circles on the ground plane. Pic related
>>2297039
When you go outside of the COV in a 2pp, you will get distortion so what you said is wrong.
>>2297041
> Where the minor axes converge.
So basicaly it can be anywhere, like I presumed. And I agree with what you said, sadly there is not a reference point you gotta guess it.
>>2297051
I didnt read your book but I can tell you that 1pp and 2pp are not the same at all. You should draw and find out by youself, when you get out of the SP in a 2pp there is distortion so you end up having minor axis that go towards the SP (but not perfectly at the same spot thought)
Here is a picture so you guys can understand. There is not a single point where all the minor axis on a flat plane would converge in a 2pp. Also as you can see, not perpenticular to the horizon line.
(and we are clearly in the COV in this picture, so that's before the distortion of picture)
>>2297079
>When you go outside of the COV in a 2pp, you will get distortion so what you said is wrong.
>Cone of vision and what grid you use are completely irrelevant and don't change a thing IN THAT REGARD
[emphasis mine]
>I didnt read your book but I can tell you that 1pp and 2pp are not the same at all.
1pp is a type of 2pp where one of your vanishing points is at the center of vision, and the other is at infinity. Likewise, 2pp is a type of 3pp where the third vanishing point (up high or down low) is at infinity
Don't even know what I'm responding to old b8 anyway.
>inb4 Robertson doesn't know how to draw an ellipse correctly
>>2297086
Here is another 2pp perspective
You can clearly see the minor axis converge to different points.
Holy shit some of you guys in here are retarded.
1pt - minor axes are either perfectly vertical, hotizontal, or a perfect circle
2pt - minor axes are either perfectly vertical, or they go to a vanishing point.
3pt - minor axes are not perfect anymore...it will converge to one of the 3 vanishing points.
I will draw a picture for you if you still can't wrap your head around this.
>>2297082
1. That's a one point perspective grid.
2. Strictly speaking, you don't need to be outside the COV to have distortion. It's not a yes or no thing. It's a by-degree thing.
3. Extend the minor axes of all the ellipses all the way down. It will converge to a center.
>>2297092
You're an idiot.
That is a 3pt grid.
How do I know?
We're looking DOWN at it!
>>2297095
Ok, if you insist. Draw a 2pp and go outside of the 60 COV and try to make an ellipse sitting on the flat plane with minor axis being perfectly vertical or going to a VP. You won't be able to do it, I garanthee.
>>2297092
I think you understand the word "converge" differently.
>>2297097
Yeah nevermind this message you are right its 1 pt perspective, thats another discussion.
Ok so here is a 2 pp as you can all see the minor axis converge to different points. I hope that now you can all see with your eyes I am not retarded, not trolling, not bating. If this image is wrong then please show me how.
>>2297082
>(and we are clearly in the COV in this picture, so that's before the distortion of picture)
No we're not. I used the information given to find the station point, and from that the 60 degree cone of vision. All this picture shows is that distortion outside of the cone of vision becomes apparent enough that that 1pp and 2pp models no longer produce accurate-looking results.
>>2297105
A lot of your lines don't line up with the minor axis lines already shown. Second, a lot of those minor axis lines are aliased from being a low resolution render. Any extrapolated line drawn from them would be an approximation at best If given a perfect render of the same view, the minor axis lines would converge to your 3rd VP.
___
All perspective tools we have are models to approximate vision within set parameters. Going outside of those parameters gives funky results, which isn't anything ground-breaking for any model.
If methods used by professional artists and art instructors to draw accurate circles in perspective aren't accurate enough for your needs, then stick to 3D renders. Nobody in your prospective audience would notice or care for ellipses that marginally differ from what is taught.
>>2297099
Yeah you're right...I can't
Here look.
Wait a minute...
If SQUARES are distorted when outside of the COV...then ELLIPSES will be distorted too.
>b-but muh undistorted symmetric ellipse!
Technically it IS still symmetrical and "undistorted" as it touches all tangent points while retaining the rules of 2pt... but it visually looks retarded.
>>2297135
Oddly enough...I could probably make it look like it's undistorted...buut it would be stretching the rules
>>2297115
Dude.. wtf are you talking about and what is your point?
- First there is no necessity to trace every single minor axis, you only need a bunch of them to see they don't converge to the same point.
- Second wtf : the quality is not bad all the VP lines match, this is just a bullshit excuse because you don't want to admit that I am right and you are wrong. If the VP lines are good and they are cleary good because they all match to the same point, then you can do an ellipse by yourself and you will have the same result even if you are at the maximum render definition quality etc.
Look, like so many users here, you don't even understand what you are talking about, so please fuck off with your lecture about the professional artists methods, this is completly pointless to the discussion and you absolutly have no autority to determinate who is or isnt ineligible to these techniques. It's not because you read the book of a professional artist that you become one. And yes sometimes professionals make mistakes, it happens.
>>2297135
Glad you can see this problem too. Be carefull on your red ellipse your minor axis is not in the right position, you should be able to split your ellipse into 2 symetrical sides.
>>2297148
You mean the Major axis isn't in the right position.
Yeah, I saw that error as soon as I posted it.
BUT regardless, All verticals remain perfect NO MATTER WHAT in Two Point Perspective.
>>2297161
no i mean the minor axis and whats your point with the verticals? you can clearly see that the minor axis are not pointing all down perfectly. ANYWAY I came here for solution and no one seems to have one, thanks for the help, gotta get drawing now. Mods feel free to delete this shit pile of thread
>>2297165
>hi I'm op. I made a thread seeking answers for a perspective problem, especially regarding two point
>ok op here are the rules of two point : all verticals are perfect...every other line will converge to a vanishing point
>wait anon look at this example, the verticals seem to not be perfect! Look at these minor axes!
>thats because you're doing it wrong op. The minor axis will always be perfectly vertical in two point perspective. Even when outside of the cone.
>You anons are all dumbshits who can't even answer my question
>Op we gave you the answer
>Mods, delete this shit thread
I shouldn't have even tried to convince you...Mayne
>>2296396
Two point perspective is a theoretical simplification of what happens in the real world. Fish eye lenses, or any wide angled camera capture the world around us better. If you are surround by a field of flat elipses as far as the eye can see in every direction, no matter where you turn their minor axes will all point...... at your feet.
This is because there exists a vertical vanishing point too.
Try it yourself with some plater or toilet rolls or whatever.
