1. ## Practical Perspective Question

Hi. I was wondering if somebody could give me some pointers as to how to resolve some perspective problems. I should probably mention that I do know the general stuff, read at least 3-4 perspective books, and all the video tutorials on the subject I could get my hands on, however I keep getting stuck in practical problems like this whenever I don't do a very square urban scene/start with sketching freely, so I would highly appreciate some practical advice. What I'm looking to do is to find out the correct scale of items in the scene, based on the character at the balcony 1). So for the people in the foreground 2) and especially I'm stuck at figuring out what the scale of the statue/plaza in the center is. I know the horizon passes through all pictures at the same high, and a couple of times I thought i had figured out a way to measure things by going to the horizon... however I ended up with results that seem wrong to me, and it seems to me arbitrary as to if I have a figure to which point on the horizon i go to to find an intermediary height which then using that reference point to come back to the desired location. So for example I would take figure 1) and take it towards the vanishing point of the castle wall on the right. Then I would translate it would and project that to the place i want figure 2), and so I ended up with that figure... however if i go to a different vanishing point i get a different height.

I must be missing something obvious and I would really appreciate a point in the right direction. Another question that I would appreciate a hint is something that confuses my intuition: I know that in a 2 point perspective I should be able to move heights up and down, however intuitively i would feel that lets say i could find the height of statue 3) in terms of human heights, and i put a human at it's top, intuitively if i lower him to the base i would think he'd be smaller 'cause he's further away than if he is on the horizon line.

Thank you very much for your patience and hopefully help.

PS: I apologize if I posted this in the wrong category: I looked around and searched but I couldn't find a better place where I could ask such a question.

2. I know this gets old and it must be a bit frustrating to hear, but you've got way more important things to figure out about this painting than some advanced perspective stuff. I'm having trouble making out the forms, overall the piece doesn't read very well.
You say you know the basics about perspective, that's great and should last you a long time until you feel you need more advanced knowledge to pull off an idea. Right now what you need to concentrate on is your skill in depicting three-dimensional form (practice by drawing from life), composition (read e.g. Jack Hamm's 'Drawing Scenery') and then Colour Theory (Gurney; 'Colour and Light').

The reason I jump to this kind of answer right away is that you usually don't ask yourself questions like "how do I measure this precise size of X and transfer it" and so on unless you're doing something highly specialized. You should be thinking in terms of volume, value, composition and so on, not in these technical terms. Spending too much time on these technical things will only hold back your development. At least that's my personal take on this, of course I stand to be corrected by better artists than me (there's plenty of those around here, so you might get you perspective questions answered after all).
Last edited by Benedikt; January 21st, 2013 at 05:39 PM.

3. Thank you very much for the reply and suggestions! Especially the book advice. I didn't know the Gurney book and the Scenery book I had only lightly browsed through it! I wasn't however at this stage worried about volume through light and readability since i was still breaking down and moving things alot, and I've gotten used to adding more of the lighting/shading and adjusting for b&w readability in the later stages as once I put those in to me things seem much more fixed down, still, that might not be the best thing. I saw your works and they're very nice, and I would definitely like to learn to work better all through the process. Thank you so much for your reply. I very much appreciate your suggestions and advice and I can see I have a lot to learn from you!

4. Also when starting out it's best not to take photos or textures and tack them into a piece it's a bad habit people get into with digital. If your an expert and can draw/paint rather well then LIGHT use and knowing how to use a picture for texture or something can be nice. But most people abuse the crap out of it and it shows.

5. Originally Posted by iXaarii
Thank you very much for the reply and suggestions! Especially the book advice. I didn't know the Gurney book and the Scenery book I had only lightly browsed through it! I wasn't however at this stage worried about volume through light and readability since i was still breaking down and moving things alot, and I've gotten used to adding more of the lighting/shading and adjusting for b&w readability in the later stages as once I put those in to me things seem much more fixed down, still, that might not be the best thing. I saw your works and they're very nice, and I would definitely like to learn to work better all through the process. Thank you so much for your reply. I very much appreciate your suggestions and advice and I can see I have a lot to learn from you!
You're welcome.
About Jack Hamm's book; it's quite tempting to go through the pages quickly because at the first glance the concepts seem quite basic. However, there's a lot of very advanced knowledge lurking behind most of the basic concepts. It's compilcated to a point where I can hardly take in more than one or two pages of info in one go before I start tuning out mentally. Get the book, really read it. You'll profit a great deal.

As to the process you're describing, of course everyone works differently, but in general the rule of thumb is to work from big decisions and forms to smaller ones. Try doing thumbnails first. Get your composition nailed before you move to technical details like spot-on perspective and anatomy etc.
You'll find that you will start to think in a more 'holistic' manner- right now you're stuck with specific questions about one specific aspect (perspective). The more you use thumbnails and the more you work from general to specific, big to small, the more control over the outcome you'll get. That's why with pros, the picture is basically finished after 20 minutes (tops) of drawing/painting. All the additional hours are just elbow grease and detail work. Focus on the first 20 minutes.
Last edited by Benedikt; January 22nd, 2013 at 09:36 AM.

6. Originally Posted by iXaarii
So for example I would take figure 1) and take it towards the vanishing point of the castle wall on the right. Then I would translate it would and project that to the place i want figure 2), and so I ended up with that figure... however if i go to a different vanishing point i get a different height.
That's the problem, you can't just use an arbitrary vanishing point. The correct vanishing point will depend on how much deeper into the scene line 1 is meant to be compared to line 2, as this controls how high up on line 1 point X is (the point that is physically level with the base of the figure at 2). [For example, if lines 1 and 2 were both the same distance into the scene, x would be so low that the line through x would be horizontal, and there would be no vanishing point]. So place point x at the point on line 1 that you mean to be about level with the base of the figure at 2, and find your vanishing point accordingly.

Originally Posted by iXaarii
I know that in a 2 point perspective I should be able to move heights up and down, however intuitively i would feel that lets say i could find the height of statue 3) in terms of human heights, and i put a human at it's top, intuitively if i lower him to the base i would think he'd be smaller 'cause he's further away than if he is on the horizon line.
.
You'd think so, but strict linear perspective doesn't work that way!

7. Originally Posted by Benedikt
You're welcome.
About Jack Hamm's book; it's quite tempting to go through the pages quickly because at the first glance the concepts seem quite basic. However, there's a lot of very advanced knowledge lurking behind most of the basic concepts. It's compilcated to a point where I can hardly take in more than one or two pages of info in one go before I start tuning out mentally. Get the book, really read it. You'll profit a great deal.
Thank you! I'll get back to the book right now!

Originally Posted by Benedikt
As to the process you're describing, of course everyone works differently, but in general the rule of thumb is to work from big decisions and forms to smaller ones. Try doing thumbnails first. Get your composition nailed before you move to technical details like spot-on perspective and anatomy etc.
You'll find that you will start to think in a more 'holistic' manner- right now you're stuck with specific questions about one specific aspect (perspective). The more you use thumbnails and the more you work from general to specific, big to small, the more control over the outcome you'll get. That's why with pros, the picture is basically finished after 20 minutes (tops) of drawing/painting. All the additional hours are just elbow grease and detail work. Focus on the first 20 minutes.
I've tried working on the first parts, and sometimes that comes out okay, however my bottleneck is still at the stage where I have to turn that thumb/sketch into something more final. What often happens to me is that in the sketch I made some very wrong perspective assumptions, so I need some ways to figure out scale even at that stage, otherwise the images I have in my head just can't be worked out on paper. For example I might imagine a mountain and a castle and at the top of the castle some figures, but regardless if I can work out a composition or not it all falls to bits at this stage when i realize there's no way I could see the figures on the castle at the same time as some other item in the scene.

Originally Posted by briggsy@ashtons
That's the problem, you can't just use an arbitrary vanishing point. The correct vanishing point will depend on how much deeper into the scene line 1 is meant to be compared to line 2, as this controls how high up on line 1 point X is (the point that is physically level with the base of the figure at 2). [For example, if lines 1 and 2 were both the same distance into the scene, x would be so low that the line through x would be horizontal, and there would be no vanishing point]. So place point x at the point on line 1 that you mean to be about level with the base of the figure at 2, and find your vanishing point accordingly.
You'd think so, but strict linear perspective doesn't work that way!
Thank you sooooooooo so very much! That was priceless! That X point i now see is very meaningful. And you've implicitly answered another big question I always had in the back of my mind: so it is a lot up to how you choose things, in the sense that you have to decide the depths of things on the picture, it's not enough just to pick points in teh 2d space and they're automatically determined. I still have to figure out what the difference is however, because this seems to happen whenever i try to create something, but whenever i get back to a perspective book, and they've got streets and stuff there it all seems so clear, it's all perpendicular streets and everything is determined by just moving along the perspective projected perpendiculars.

Thank you so much for taking the time to explain that and for the drawing. Now I may be wrong but i feel somehow under the surface your answer also contains the answer to another question that has been tormenting me for months: in all the perspective tutorials I haven't yet seen anybody explicitly explaining or giving an intuitive correlation between the vanishing points/horizon lines and the real world experience of cameras/3d software where you can change the lens, basically changing the rate of diminution. So intuitively I know I "could" for example sketch a really fast rate of diminution, whereby chracters get really small really fast as you go deeper into the image, thus allowing in the same picture to show both a figure and something as big as a canyon/fortress... HOWEVER, I don't have any understanding as to how I'd model that in building my perspective grid. Where in the construction of the perspective grid does one nail down the lens/rate of diminution? Because I'm guessing it MUST be somewhere nailed down, as it would be physically impossible in reality to have in the same imagine two pairs of near&far characters, where one pair has a very strong/fast perspective shrinking, and the second a very slow one, meaning they're almost the same size despite depth, almost orographic projection. I guess what I'm asking is that I'm assuming once i drew the relative scale of the first pair of figures I'd expect the lens depth/dimution rate is fixed so I should be able to translate this to the second sate of figures. Is this in any way related to what you were saying? Any light on the subject would be appreciated.

Thank you very much for all your patience and help.

8. Originally Posted by iXaarii
in all the perspective tutorials I haven't yet seen anybody explicitly explaining or giving an intuitive correlation between the vanishing points/horizon lines and the real world experience of cameras/3d software where you can change the lens, basically changing the rate of diminution.
The distance between VPs or the size of the circle of view (in relation to a fixed pictorial area) between VPs determines the effective focal length of a lens as seen through a camera. The larger your circle of view (cone of camera vision in this case) the closer together the VPs and the wider the effective angle-of-view the lens would be. Reducing the circle of view (effectively zooming into the scene) will increase the effective distance of the VPs, representing a lens with a longer focal length (e.g. a telephoto lens if the circle/cone becomes small enough). The smaller the circle of view, the more compressed the perspective will become and the rate of diminution will decrease accordingly.

*Every photographic image expresses every possible focal length as you zoom into the scene--*starting with the actual focal length of the lens used and increasing from there.
Last edited by bill618; January 24th, 2013 at 10:54 AM.

9. Originally Posted by bill618
The distance between VPs or the size of the circle of view (in relation to a fixed pictorial area) between VPs determines the effective focal length of a lens as seen through a camera. The larger your circle of view (cone of camera vision in this case) the closer together the VPs and the wider the effective angle-of-view the lens would be. Reducing the circle of view (effectively zooming into the scene) will increase the effective distance of the VPs, representing a lens with a longer focal length (e.g. a telephoto lens if the circle/cone becomes small enough). The smaller the circle of view, the more compressed the perspective will become and the rate of diminution will decrease accordingly.

*Every photographic image expresses every possible focal length as you zoom into the scene--*starting with the actual focal length of the lens used and increasing from there.
Hmmm, now that is very densely packed information! I must've read those passages like 5 times already (and will get back) and I'm still not sure i've gotten everything... What I think i'm understanding is that the distance between the vanishing points of say a rectangular building determines the rate of diminution, if they're both on screen it's gonna decrease really fast, and if they're sort of infinitely spaced apart would that mean ortographic projection? wait, no... i'm guessing something wrong, aren't I.
That passage with the fact that every photo contains all focal lenses inside... wooow! Mindblowing ! I never thought of it that way. Thank youuu!!!

10. Originally Posted by iXaarii
Hmmm, now that is very densely packed information! I must've read those passages like 5 times already (and will get back) and I'm still not sure i've gotten everything... What I think i'm understanding is that the distance between the vanishing points…determines the rate of diminution, if they're both on screen it's gonna decrease really fast, and if they're sort of infinitely spaced apart would that mean ortographic projection?
Sorry for not being clearer, you're basically there. The further apart your VPs (or the smaller the object you are drawing between them) the more parallel the ray lines from the VPs become. As a result the diminution decelerates. The greater the distance between the VPs the flatter (more compressed) the perspective becomes, approaching ‘parallel projection’. Parallel projection is basically perspective projection seen through a lens of infinite focal length (or VPs infinitely spaced apart).

Here’s an animated example of the effect of expanding distance between VPs out to infinity:

http://en.wikipedia.org/wiki/Perspec..._(photography)

This illustrates the last point you quoted:

Last edited by bill618; January 26th, 2013 at 03:53 PM.

11. thank you very much! that was very useful! And the wikipedia link too! thank you so much for taking the time to explain this! This is so awesome! Thaaank youuu!

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•