[rdgraffix]

puf, puf, sorry I'm late to class ...

I only discovered this forum fairly recently, and just found the 'Paint along with Fred' lessons started in May. Are you still doing these?

I've learnt amazing things, just from that first lesson. I'd like personally thank you for the time and effort that was obviously put into this, and for having the will to help out other artists this way.

Could someone please point me in the way of any other similar lessons, or have they been archived on a site somewhere? I'd be *very* interested in continuing with these, if at all possible.

BTW, here's my attempt at the cube exercise. If you get the chance to comment on it, please tear it to shreds - no mercy.



Thankyou to everyone who has helped with getting the lessons online.

------------------
- rowan dodds
RD Graffix

[spooge demon]

Ok, you might have to read between the lines here...

YOur cubes are reading gray, value too dark.

Facing plane of far cube is in shadow, but would be receiving a lot of reflected light.


Ok, you might have to read between the lines here...

Out of perspective. Are cubes square to each other in plane view? Check construction, VP's not precise enough. Flip image and you will see.

Your cubes are reading gray, value too dark.

Facing plane of far cube is in shadow, but would be receiving a lot of reflected light.

Receding plane of far cube has feels reflective, should be matte. Cast shadow and shaded sides of cube are too similar in value. Cast shadow should be darker.

Fixing both these previous problems will help cubes read as white.

Need accent of dark where shadowed cube sides approach ground plane, shows surfaces coming together by lack of reflected light.

Cast shadow aligning with bottom edge of cube, not helping show form as much as it could

[rdgraffix]

Thanx for replying spooge!

Your comments were great. Now you've pointed it out they do look out of perspective. They were meant to be aligned to different vanishing points, but I think the angles were too similar and cause confusion. Also, I fudged a bit of the perspective, which obviously didn't help. - BTW, I love that trick of flipping it to see it afresh!

The values were totally screwed as well, which I can see after doing a second version.

Thankyou for the tips on reflected light and composition. Below is a second version, trying to take all your insights on board.



What a difference! Thankyou spooge!
- do you have time to comment on this one?

These sort of exercises are fantastic! How many illustrations are ruined by incorrect visualisation of the basic forms?? No amount of detailing or finish can make up for those mistakes. I'm in great debt to you for helping me learn to judge these things.

What next? cones and spheres? or do I still have more to learn with these?

Was there another lesson written after this one?

I just want to say again how much I appreciate you spending your own personal time helping me with this.

------------------
- rowan dodds
RD Graffix

[Sumaleth]

quote:

---

Originally posted by spooge demon:
Need accent of dark where shadowed cube sides approach ground plane, shows surfaces coming together by lack of reflected light.

---

A surprise!

This would be the intuative answer to the question of shading a surface which ajoins a cast shadow, but it seems to be at odds with what Fred found in his practical experiments (from back in the original thread).

From a visual point of view I prefer to have both surfaces darken slightly as they touch so I've continued to paint like this (although the last cubes I did were based on the other approach), but I'd love to know what would be the realistic approach.

Sumaleth

[SushiMaster]

It's probably due to some sort of optical illusion?

Daniel

[spooge demon]

much better!

No, not an optical illusion. The reason there is any light in the shadows at all is 1)reflected light from the surface
2)diffuse light from sky or bounce from light scattering in air or bounce from more distant objects outside the scene.

Think of it this way. Pretend this cube is a building with 50 stories. If you are on the top floor on the shaded side what do you see if you are looking out? A brightly lit plain, with a lot of light coming at you. You are looking down at that plane, so you see quite a bit of it.

If you are on the bottom floor, you see the lit plane at a very low angle, so there is not much light reaching your eye here. So there is less light incident on the lower parts of the cube.

The mitigating effect is the higher up you go, you are farther from the light source, the lit ground, so it goes a little darker. If this were not a cube, but a long box on end, the top would not see as much reflected light as the lower areas.

Think of an arm resting on the torso. As the form rolls together, they mutually shade one another from reflected or ambient light. This is why it goes darker.

[Virotype]

how are you making these cubes? In photoshop or in a 3d program?

------------------
- Virotype
http://www.phantommx.com

[SushiMaster]

Spooge: Thank you for the very clear explanation. I'll remember that :-)

Daniel

[rdgraffix]

Thanks spooge, that explanation really makes sense.

Virotype: these are drawn straight in photoshop, over basic guides to get the VP and perspective right first. Airbrushing them like this means I have to work out all the values and tonal variations myself, rather than let the computer do it for me - which is basically what I'm after for this exercise.

It's actually refreshing to go back to the basics like this, concentrating on one aspect. It makes a change from the full illustrations I'm used to, where you have to take into account composition, colour balance, technique, detail, emotion, and everything else as well as just convincing construction and lighting.

------------------
- rowan dodds
RD Graffix

[rdgraffix]

Spooge, I noticed in the old messages that you said the next step would be doing one white cube and one black one. I've tried that below, working over my last file to concentrate on the differences. I assumed that the lack of reflected light from the black cube would have the effect of darkening the black cube's shadow slightly, esp. towards the cube's base, as well as slightly darkening the facing, shadowed plane of the white cube. Is that basically right?



I know I haven't perfected the two white cubes yet. For instance I should have altered the tone behind the front cube to help the left-facing plane stand out from the BG.

I'd appreciate any help from you, Fred, Francis, or anyone else. Tear it to pieces, so there's nothing left. I'm finding this really helpfull, thinking on a more realistic level, rather than just fudging whatever I think might look good at the time, and wondering why it doesn't look right.

I know I'm incredibly late in doing these, and I'm really thankfull to anyone who takes the time to critique this.

------------------
- rowan dodds
RD Graffix

[Sumaleth]

On the black cube I think you'd expect to see quite a bit of reflected light )from the ground) on the lower part of the side. Possibly a little on the back too, on the left side where the cast shadow isn't blocking so much of the bounce light.

Sumaleth

[Danny]

Hmm..

Interesting way of explaining the darkening effect Spooge. I've always been aware of this phenomenon, but never theorized it like this. I've always had a more blunt explanation. "The reflected lightrays having gradually more problems reaching the surfaces near the 'intersection' the closer to the 'intersection' they get." Somehow this doesn't sound quite right now, or overly complex.

Virotype, doing these cubes in a 3D program kinda defeats the purpose of the whole excersize. Look back into the history of the forum to find the original thread to see what this really is all about. You'll see the light..

Danny


------------------
[email protected]

Trust in Trance

[rdgraffix]

Sumaleth (at last I've find another Rowan not spelt with a fuckin' 'h' ), thanks for answering. According to spooge's explanation, wouldn't the area recieving the most reflected light be towards the top of the black cube?? ... unless of course, the black cube was a bit reflective.

... any help with this one?

------------------
- rowan dodds
RD Graffix

[Sumaleth]

Ahh, thats an interesting point. I was planning to redo my old cube picture with the proper shading but maybe I'll wait to see what Spooge says about the sides.

My thinking was this;

A white surface is white because all (/most) of the light that is hitting it is being bounced. So I thought that the side of the cube would be getting all of the direct light from the environment, plus light that is reflected off the bright surface (with the majority of the effect being close to the surface).

This thinking came from trying to understand Fred's practical results. Back in the other thread I had a hard time understanding why one surface would brighten towards the join rather than both surfaces having a core shadow.

The way I rationalized it was that the vertical surfaces were getting the direct light from the environment, plus the bounced light from the ground plain, and since the cast shadows weren't totally black they would still be bouncing a little bit of light. And because light degrades with distance, it seemed to vaguely explain how the lighting was to work.

I never could work out why only the cube would get this though - the same argument could be given for the cast shadow itself yet it did have a core shadow.

With Spooge reaffirming that you do get two core shadows I thought that meant just where the surface is away from the light source and is also joining a cast shadow on the ground. I thought that the sides, which do no join cast shadows, would still get some strong bounce light from the ground.

But now that doesn't seem totally consistant. I'm almost wondering whether there would be a faint darkening on the -ground- along the side of the cube.

I'm very interested to hear what Spooge thinks about this.

(As I said in the BAD HABBITS thread, I tend to over-analyze..)

Sumaleth (aka The Other Rowan)

[spooge demon]

Rowan,

I was with you up to the idea of a vertical surface being lighter due to the added source of the bounce from the surface, but beyond that I am not sure I followed you.

Generally core shadows happen on curved surfaces, but I like a broader and more accurate interpretation of a core shadow as the absence of reflected light. So technically you could get a "core shadow on a cube and cast shadow setup. Sounds strange to my ears, but it makes sense.

If you set up some cubes and observe closely, you will see a whole bunch of effects that fall outside the relatively simple rules that we are following in these studies. The rules are simplified because we want to eventually be able to know how to render a simple object well enough that we can assemble them together in a complex way and still be able to reason out how to show the form through values. it is a solid reason for simplifying things. And the truth is that no object is perfectly matte or reflective, surfaces become reflective as the viewing angle becomes lower, there is no such thing as a perfect point source, etc.

Look at Fred's figure studies. You see light, shade, and the division between or core shadow, very clearly. It is the simplification that makes it read, and eventually, the simplification that will lead you to designing your shapes better. This will make your art stronger, rather than emulating esoteric effects of light.

The whole point of this is to be able to design an object, understand the basic shapes, assemble them, and then shade them convincingly. It will also help you to interpret both what you observe in nature and in photos a whole heap better.

Oh, Rowan #2

The black cube image looks good. You do need arrange of value in the facing side of the black cube. Also, your cast shadows and shaded vertical sides could use a more distinct separation in value; cast shadows go down a bit. Perspective, composition, craft, look very good. Proceed to one cube, one cylinder, cast shadow falling on one object from another.

In general, follow the idea of lighting along the diagonals of a shape. On the black cube, if you draw a line from the closest bottom corner to the top farthest corner you will hit the light source, roughly speaking. Since you have control over the direction of light, you were a good designer and did this, right ? The general rule is the farther from the source, the darker a surface becomes. As I think you were discussing, this does not always happen and can be contradicted by a whole host of possibilities, but it will give the most easily understandable read. also things will be easier to reason through when you start building complex forms. As I said, the grad goes along the diagonals, not straight up and down, generally. One that face of the black cube, you could go from a 3.4 to a 3.6. Any more and the surface would start bending. When you get into color, you go from warmer to cooler along the same idea.

[Sumaleth]

Spooge;

Yes, I read through my posting later and wondered how much sense it made. Part of the problem was that as I was writing it I could see the rationale I was working to didn't really make sense. I guess I went along with it because it 'fit the practical model'.

But the question is still there and this time I've got pics to help ask it!

BTW, my use of the term 'core shadow' probably isn't all that accurate. I assumed it described the way that a cast shadow gets slightly darker towards the object/cube but I like the description that it's "the absence of reflected light".

The four diagrams here will hopefully help describe my question;



It's just a really quick cube with cast shadow. The surfaces are all unshaded/ungraded so that the additions are more easily seen. The smaller versions with the arrors indicate the area of interest (I didn't put the arrors on the main cubes because I thought they might make the changes more difficult to see).

Image (A) shows the basic flat shaded cube.

Image (B) adds a simple core shadow (CS). The CS on the ground plane (in the cast shadow) was never in question - it's definitely there - but Fred's experimenting with real shapes suggested that rather than having a similar CS effect on the back face of the cube (as I've drawn here), it actually got -lighter- towards the base. This is where I got confused (it didn't seem intuative) and it's how the theory I discuss below developed.

A possible explanation hit me earlier today (an obvious one too) - in the real world all surfaces do reflect to some extent, there's no such thing as pure matte. So perhaps the reflections on the practical cubes were confusing the results somehow? Still doesn't fully explain things though. The big question is; which is correct?

On one hand, the tight corner where the edge and ground meet, even though there's no cast shadow, could produce an 'absence of reflected light' effect causing both surfaces to darken slightly along the join indicated. Image (C) illustrates this idea. It seems the most logical - it's consistant with the back of the cube.

On the other hand, the absence of a cast shadow on the side means that the ground surface is now bouncing a lot of light off it. On image (D) I've painted a faint highlighting across the join - both the cube side and the ground plane get this extra light.

In the icon on (D) I've got three 'O' arrows indicating 'light from the environment', plus two arrows that indicate bounced light from the ground. One of these two bounce arrows hits the ground some distance from the cube whereas the other one hits right next to it - the result being that the light has different distances to travel after bouncing from the ground before hitting the cube, and this I thought might have accounted for the practical results (ie. shorter bounce means lighter surface value).

This logic seems like a bit of a stretch now. But the interesting thing is that this version of the cube looks more realistic than (C). To my eye at least. I'm thinking that it's probably "reflection" that would cause this effect which might explain why it reads "right".

The cubes you drew early in the 'Painting along with Fred' thread seem to have features that could follow both of these approaches - artistic choice? In retrospect, if you were to do that image again what would you do differently, and would you follow (C) or (D) in general?

Sumaleth (Rowan #1)

[rdgraffix]

Thanks for you help spooge. I've tried what you sugested. It took quite a few guides to make sure I got the bend of the shadow falling on the cylinder right, but hopefully it's pretty accurate.



I've tried to improve the cast shadows on this one, and I've tried to take into account what you were saying about the lights direction and the grading of the planes. The diagram below is what I've taken you to mean by lighting on the objects diagonal. Is this right?



Would it be best to vary the light source slightly from the diagonal to add some variation and interest to the shading and shadows, or just keep it simple?

If you have the time, I'd appreciate any comments you might have on this one, esp. the tones on the cylinder. Hold nothing back, so I can hopefuly learn a thing or two for next time.

Thanks again for helping me with this. I can't believe I've never come across these exercises before - they're really usefull, and I'm already starting to see an improvement in my other work, just from learning to think along the right path.

- Rowan 2 (the EEeeevil one )

------------------
- rowan dodds
RD Graffix

[spooge demon]

One thing that I have tried to stress is the separation between light and shadow. The key light takes precedence. I don�t care for the darkening toward the bottom as pronounced as you have it in B. It looks like a change in local color or a turn of surface.

Yes, all surfaces reflect, or they would be black. a matte surface reflects light out in a random fashion, the light comes in and is scattered all over the place. A reflective surface bounces the light in the reciprocal direction that it came in on. If I remember my jargon this is angle of incidence=angle of reflection. The degree to which this organization exists determines the degree of reflectivity or matteness. A bright white matte surface reflects just about all the light incident upon it. That is why we get very strong reflected light in these studies. If we made the ground plane black, well, you know.

But keep in mind the brighter the key light, the brighter the reflected light. The darkening you are talking about in the lit side blurs the difference between light and dark. For god�s sake, don�t do it it will cause you heart ache later on, and it does not buy you anything

For my money, �d� is correct. Keep your halftones on the light side dependent on the key light for it�s variation, not reflected light.

So like I said before, you can set this up and because you have many �impure variables� you can get a lot of artifacts that fall outside this simple idea of light, shadow, core shadow. The light can be somewhat diffuse; the cubes could be somewhat reflective, etc.

Another aspect to include on these studies is the idea of �there is no such thing as a sharp corner� All surfaces have a radius. A thin dark line would indicate this where the objects sit on the ground. A cast shadow, yes! The same thing in the shaded side. It might seem esoteric, but the things we include here will help us later on.

Rowan #2,

You are on the right track. Your ellipses are too closed. If that is a cube next to the cylinder, a circle must fit on the top surface. Do that and compare that ellipse to the cylinders. Shadow plotting looks good. I like the subtlety of the halftone falloff on the lit side of the cube. The portion of the cylinder that is in cast shadow that faces the viewer should go a little darker. The reflected light would be blocked somewhat by the cube. You can go a little darker still with your cast shadows. Try to get the separation between shaded side in reflected light and cast shadow with none.

[Sumaleth]

Spooge;

When it comes to C, I'm guessing that in reality it probably does darken in there, for the same reason that it does on the shadowed side, but the overall brighteness on the side (and subsequently the strong bouncing of light), along with the effects of mild reflection tend to negate it.

I won't do it anyway .

As for the darkening on B, it was just a quick brush mark to indicate what I was talking about, it wasn't fine turned.

Anyway, with all this in mind I dug out the old shapes PSD and made up version 4;



I feel like I'm getting close now, but I thought that about version 1 too.

Anything still stand out as wrong? Or could be improved?

Thanks,
Sumaleth

[ozenzo]

man this is a tutorial in itself!! Thanks for all the info!!

[This message has been edited by ozenzo (edited August 19, 2000).]

[nori]

you guys made really good ones. I can't seem to get the color of my cube right. What tool do you guys use for coloring your cubes?



[This message has been edited by nori (edited August 19, 2000).]

[rdgraffix]

Thanks, spooge, I think you're right. I must have messed up my construction for the cylinder. I've tried adjusting the tones of both the cast shadow and the shadowed area of the cylinder. Is this better?



Nice pic Sumaleth, I think you're a bit bolder than me at this point, complexity wise. The base of the cone looks a little wierd to me though. Is it based on a full elipse, or just eyeballed?

Nori, I've drawn my cubes straight in Photoshop, over construction lines. To shade them I've basically made each plane a seperate layer with the polygon lasoo, and used that layer as a mask for various layers of airbrushing.

Ozenzo, I'm glad I'm not the only one learning something from this. It's all to easy to think you're beyond the basics, but all to many artists (myself included ), never got these right in the first place. Keep this information in the back of your mind and I bet you'll be amazed at the differance it'll make to your work.

------------------
- rowan dodds
RD Graffix

[Jorge]

I was not subscribed when the past thread was up, I caught up on it by this one, so I'll understand if I get no answer. But before the lighning I got caught at how do you control the perspective on a 2nd cube. That is, if I use 2 vanishing points for a cube, how can I find the other 2 for a 2nd non-parallel one somewhere on the scene. And how do you make the cube look like... er... a cube (ratio between axis). I looked for this in the archives but could not find it. It is a very big archive, though.

------------------
Jorge

[Sumaleth]

Dodds;

All the eliptical surfaces where first positioned with constructions lines (forming squares) and then I created eliptical selections that fit into these construction boxes. To be honest it wasn't totally accurate but I think it would be reasonably close.

As for being a bit bolder, I missread the original assignment and thought we were meant to be doing all 4 shapes in the one image :}. Definitely bit off more than I could chew but I've slowly got a handle on it and now feel that it was worth the effort.

Jorge;

We did bring up that question in the original thread and there didn't seem to be any way simple to work it out directly in in the image.

If you imagine that the horizon line is really a circle which wraps around the viewing position then pairs of VP's, for two cubes at different rotations, would rest on this line at equal distances. Probably. I couldn't find any formulas that support this and didn't spend much time trying it out myself so thats only a guess.

The only way around it was to first draw a plan view of the scene (ie. from directly above), and then project it down into the 3D space. There are diagrams somewhere in (the 6 pages of) the 'Paint Along With Fred' thread.

Sumaleth

[spooge demon]

Here's that trouble with your cylinder.



Sumaleth, There are a few funky things going on, but you are pretty close I feel. It will do you good to try another set up. This one has a lot going against it. Your cast shadows intersect each other so that the shapes of the shadows do not reinforce your forms. Backlighting is always a difficult thing to pull of, if you are trying to be specific about form. It is nice and dramatic, but see how much of your objects are in shadow, and what is in light gets a little blown out. Not much to play with, and with a light BG towards the light, you can't really put contrast in there either.

Ok, now do a cylinder and a cube, one white one black. Same for you Rowan1.

Nori, Go back and look at the old thread, there is a lot of info. It has all been said before, and I can�t type worth beans, so please help me out here.

And about two cubes VP�s, yes the method is called ray intercept, and you can make accurate perspective drawings from plans and elevations. This is how it was done years ago. So you can plot it, I think someone did an example of it on the old thread.

If you are having a large amount of trouble, go back and do some using this method. It will train your eye to what is correct. Eventually, you should get yourself to where you can eyeball basic shapes from any angle or lens or distortion.

[Fred Flick Stone]

Mr. Mullins-howdy. I have a question for you related to this.

Are you doing all this based on a sunlight, or is all this based on theory of how light should react to a surface. Like you said, once in sun, there is all sorts of distortion to take into consideration, but everytime I go into sunlight conditions, a lot of what is being done here looks formulaic in the lab sense. So, are these rules and analogies you speak of just a simple formula to get the idea of form going because we do not have the luxury of working with subject matter all the time?

The reason I ask is two reasons, one, I have been doing these outside and the objects take on more than what is mentioned here. I am also repeating this exercise for my conceptual design class, and the results are more like all that you are mentioning here. I have a controlled Halogen flood lamp, the one we light the model with, and I get these results that are happening here. Outside though, very different. And two, when I am painting from life, I want to have a better understanding of what I am looking at, in terms of value, vs. color, since almost every pocket of dark is influenced by a hue of some sort, which is a variant of value in the dark side.

Thanks again Craig, you really kicked things into high gear with these cube exercises again. I kinda spread myself a bit thin splitting tutorials, between the cubes and the heads�on that note, do you have any analogies for cones and pyramids, like your analogy of being in a skyscraper. That makes things incredibly easy to understand within terms of reflective light on the dark side of the cube. Are there any rules of thumb for these objects, as they tend to be more influenced by atmosphere than anything else. Is there ever any influence on them when next to say, a large cube, almost the same height as the cone,? Etc.

I am actually going to digitally redraw my charcoal renderings from this exercise so they have a similar feel to all these displayed.

Sumaleth, glad you resparked this thread. You were backing me into a corner, not in any harsh way of course, but technically, I couldn�t explain any further the absolutes, because like you, I tend to overanalyze things too, and I am working strictly from outside sunlight, and I have all sorts of problems with some of the variables mentioned here, until I started my conceptual class, with a fixed light, now I see this stuff much easier. But, the sun does tricky things to these objects.

RDGraffix-very nice, I think you are in position for an A on this assignment as well�J.

I will post some boxes soon. And thank you for your help Sir Spooge�J

[spooge demon]

Yes Fred, it is exactly a formula. The idea is to analyze form to draw out of your head eventually.

And after you can do that, you can look to include more interesting and complex phenom.


The main difference between inside and outside is the very strong ambient fill. A very different system. These objects setups and crits are mostly assuming an indoor setup with a fairly dark room. This makes core shadows and reflected light very pronounced and easy to analyze.

The one thing to keep in mind if you are trying to simulate an outdoor scene is that the sky is a strong source of light. There can be a lot of variables with the brightness, uniformity (clouds) and hue.

You do have one advantage, the difference in color temp between sun and sky. This carries through to the shaded side where you can see the reflected light from lit surfaces reflecting a warm light and the blue sky influence remaining cool.

Also keep in mind the very diffuse shadows from the sky source.

[Fred Flick Stone]

Thank you for responding, and so quickly...I am on the right track with this, and not going insane seeing things differently...whew..

[rdgraffix]

Aha! Thanks spooge, I've found what I was doing wrong with the elipses, and why they were coming out squint. What I'd done was create the guides as if another angled box were there, create a layer with a circle inside a box and distort the box to fit inside the desired plane, assuming the elipse would now be in correct perspective. What I was doing wrong was the angle of the defining cube, having it's vanishing points far to the left and right of the viewer. Because an elipse directly in front of the viewer should be basically symetrical, the vanishing point should be in line with the viewers eyes. This will also give a realistic distortion of elipses seen peripherally.

The diagram below shows what I'm talking about. The elipse on the left looks unnatural because that level of distortion would only be seen peripherally, not dead ahead, as it is trying to tell us. However, the ellipse to the right, with the VP in line with the viewer looks more natural.



I hope I've figured this out right an I'm not just gabbering a load of rubbish.

Hopefully, I've fixed this up in this new version, with the black cylinder.



Is this close tone-wise? I wasn't sure how much light the black surface should reflect (from direct as well as bounced light) to get the balance right between defining the shape and risking it looking grey instead of black.

------------------
- rowan dodds
RD Graffix

[AliasMoze]

Hey, Spooge got me thinking. Does the color of shadows in sunlight intensify the stronger the light is? I've noticed that shadows on the sidewalk at high noon can look incredibly blue. Thanks.