[Frost]

(if no one bothers to answer my posts you will)

[AliasMoze]

Frost, I saw your post. It was interesting. I haven't gotten a crit from Fred or Spooge either. Luckily I have friends around who can spot bugs (and Sumaleth). Anyway, I hope you stick around.

Sumaleth, I saw your site. Are you in the US?

------------------
AliasMoze
:) :) :) :)
"That activates my hilarity unit."

[eetu]

only the obsessed ones left heeh

eetu.

[Frost]

AliasMoze; I know that Fred and Spooge are pretty much out of it for a while, I'm just dissapointed that no one offered any response, gave any acknowledgement or any idea on my last 3 posts on which I awaited some news (especially the non-linear light degradation rules). Anyway, guess I'm taking it all wrong.

[Fred Flick Stone]

Where to begin...Thank you all for being so patient...I have my work cut out for me...uh, take a number and wait in line...

Aliasmoze-1st up to bat...I am looking at the image from page 2 to begin this list. I will start with the obvious...The cone is too dark. It looks a little metallic, not quite white...You have different horizon lines for these objects. The cast shadow on the left box has an extra edge where there should actually be a corner...The cubes are not quite cubes yet...I won't even go into the explanation here, but the cast shadow across the cone is incorrect...reflective light in the shadow sides of the cubes are a bit off. You actually get more bounce or reflective lighting toward the top of the cube, not at the base...the modelling of the cone is a bit muddied, that is, there is no definite core, and the direct light is not lighting the cone up correctly. Again, this could be due to the fact that the cone looks metallic, or grey and not white...finally, the cast shadows are all radiating out rather than working parallel with each other. It is though you lit these three objects with a pen flashlight, not a big lightsource...thanks again for waiting a bit, I was a little bummed that I couldn't read spooges comment on your image. I would like to have seen what he said about it...

Frost, thanks for the little tidbit onlight in a 3d rendered world, very informative...

Nex-a shadow tip: Core shadows are present in rounded shapes, or shapes with no corners...as for what you have going on in this image, both the cubes are slightly tipped upward from the middle of the image. It looks like the left cube is sitting on a slight slope tipping downward back to the left, and just the opposite for the right cube. It is really important to use the guidelines in plotting out the shapes you will be drawing before finallizing them. I know working with a true horizon line might be difficult, but all drawing and painting has it, heck, life has it. Whereever your eyes look, is the horizon line you are accute to. On to the shadows...the darks on the objects you have casting shadows should never be as dark as the cast shadows. Right now your right cube has the darkest value in the image, the cast shadows from the cubes should be darker...the left cube looks like the way my styrofoam cubes I bought look when I back light them. The cube looks rather cloudy and transparent, vs. solid and opaque. The right face of the shadow side should be lighter than the left side, because it is getting a lot of bounce light back on it from the right cube. I will post an image for you that will explain this in the next set of crits...Your cubes look as though they are lit from two different light sources. The shadows aren't parallel with each other, indicating different sources of light...Cast shadows...when a shadow is cast from an object, the edge of the shadow will be sharper, when it is closer to the object. The left cube, the bottom corner sitting against the surface...the cast shadow from this point should be very crisp...as the shadow moves away from the object casting it, the shadows edge will slightly diffuse, or soften. How much it softens all depends upon how far your light source is from the object, and how bright, or intense the light is also. An incandescent light at about 50 feet away is going to produce soft edges on your cast, while the sun, as freakin bright and hot as it is, is going to produce sharp shadows. The other variable that affects how soft the shadow edges get is how tall the object is, or how far the object is from the surfaces the shadow is casting on. The only wacky way your edges will be very soft, aside from all these other ways is if you have some sort of softening screen covering your light source. This will most definitely soften your shadows...


joachim-this is based on your first attempt one of the graphite pushers...really nice modelling, but the first box is not a cube. You also have a slight issue with the vanishing points on the two boxes. They each share a slightly different horizon line. Look at how much volume of the top plane is showing between the cubes. There is more volume between edges on the left cube, less on the right cube. Horizon problems...The cast shadow from the left cube is the same value as the dark side of the same cube, the cast should actually be darker in value. With the right cube, the dark plane looks like it is being lit from another light source coming from behind us. And if you look at the cast shadows of both boxes, they look like they are each lit from a different light. The cast shadows on the ground plane are not parallel with each other, indicating different lights...all in all, not a bad job. If you attempt this same exercise again, which you should anyway, repetition is the key to understanding in art, then make sure you draw the exact same image again, so we can work out the issues that are a problem with this first image. If you keep drawing new images each time, you aren't resolving the problems you have already begun to make.


Nori-your cool cubes have actually started melting...seriously though, the construction of the cubes is a little bit iffy. Not totally solid "cubes", but anyway...All the values you have used in this image are flat. The image looks posterized or very graphic because there is no value transition in any of the planes represented here...And check out Francis' description of casting light. As of right now, like most of the other images I have crit. so far, both your cubes are being lit from different lights. Look at the angle that both shadows cast from the boxes. Are they parallel? If not, then they are lit from different sources...


Gimbal, the ole eyeball will never fail you. If you see that both cubes are not on the same ground plane, then draw out the horizon line, and really project all the vanishing points out. There is no other cheat for this. Your eyeball alone is the greatest tool for spotting errors, you will unfortunately have to resort to basic construction rules to get back what has been lost. If you do this stuff enough to where it becomes fairly intuitive, you'll be able to place everything into your drawing plane close to correct, then it's just a matter of fine tuning. Look at the book list on page one of this tutorial that I posted. There is an excellent Dover book, only $12.95 that will easily explain every little tidbit an artist needs to ever know about perspective...


TranceR-nice little flash animation. Very informative and to the point. I may email you sometime soon here I would like to build a couple tutorials this way...very effective...now...as for the image, the cubes are not cubes yet. Seems to be a common problem most everyone s having. Both boxes have these huge white outlines separating the planes from each other, unfortunately this also breaks up the whole image. Bump those edges right up to each other. Solid forms, no outlines. All the faces of the cubes, and the cast shadows are all constructed with even tones, this is incorrect. There is value transition happening on every surface of everything, even in the shadows. This image is all flats, posterized much like Nori's image. Need those transitional values. The cast shadows on these cubes are too fuzzy. Where the cast eminates, it should be firm, solid, not fuzzy. You are also missing the cast shadow of the one box onto the little box in the front. This really distorts the image alot, and it loses it's three dimensionality. When you go back and fix these errors, do the same image again, don't start a whole new one. Let's resolve the problems you are having here first before we go on to another image.


This is all I have time for right now...I think I am going to group up certain images that all have similar problems and crit them all at once. I see that there are many people that are having similar troubles, I think this will help speed up the crit process.

Synj-I know the cube thing is getting old for ya, but there are a lot of people who still need to get through this step properly. I will post some of step two by the end of the week. But I don't want to get too far ahead, if everyone else is still confused with what has already been described. I will also be starting the head tutorial that I was mentioning to yall this weekend. I just need to find a bit more time to get everything done. I need four more hands, that might help. Someone want to draw me up a couple pair of arms, I'd really appreciate it... Till tomorrow, yall have a great night...signing off now...

[Fred Flick Stone]

frost-I haven't had time to read through all the posts yet...what exactly were your questions concerning light. If you can post this, first thing in the morning I will try and answer them for you. Like I said, it has just been a monster trying to find time to do anything, and when I am looking at this tutorial forum, I am just glazing over the images, I haven't read too much yet. But every once in a while I will glance through and see the name Fred or Spooge stand out, and I just have to read it. This is why I logged back on really quick. I should be gone right now, but your plea caught my eye...Post the questions again after this post, so I can find them, and like I said, I will answer you first thing in the morning.

I am outta here...

[Gimbal]

Thanks Fred! I'll pick that book up soon.

My problem right now is trying to get the nice smooth gradiant on the the faces of the cube in photoshop. I know how I want it to look but I can't figure out the technique to do it. I just end up with bands or brushstrokes showing through. Maybe I'm not working in a high enough res and sizing it down? Is there some way of adjusting the gradiant tool so that it isn't so visably banded? Should I be setting up a mask layer and using very low pressure airbrushing? If I can figure that out I'll be well on my way to getting a more realistic cube.

You know, once I get cubes, spheres and cones down and I'm able to control if they are matte or have shine, I'll be able to present you with a realistic rendition of the award you deserve.

I hope everything is going well with finding a new dwelling and getting settled in.

[AliasMoze]

Fred, thanks! I agree, my first pic's modeling is totatlly f'd, and the shading sucks too. I'm still working on more cubes. You have no idea how much this assignment has helped me. The more complex stuff that I rough in is coming easier and easier because I'm more focused on values (in a logical way).

I actually have the book "Creative Perspective For Artists and Illustrators" by Dover. It's really good (and cheap). I got mine from Amazon for I think $7.95.

Again, thanks for the crit.

------------------
AliasMoze
:) :) :) :)
"That activates my hilarity unit."

[dines]

Sumaleth,
can you tell me where you find this option (not to display images), it sounds very interesting

thanks


dines. http://www.nexen.net/koubis

[iska]

Maybe we could ease the critics process up a little bit? Simple things like putting your ownname and version number of the picture in the image.

For example I would put name "iska" and "3" on my next cube image

I guess this would help the grouping Fred was thinking about...

-iska ;D

[Joachim]

Hey, thx a lot for the feedback Fred.

I didn't know that it was preferable to make the same cubes all over. I will do them again
But, even so, I think I learned a lot about vanishing points and all that by doing the new one, since it made me start to think of the logical reasons for all the stuff, even though I did some of it wrong. And also, since I didn't get much feedback on the first one before I started on the other, I wouldn't know how to improve it too much. But, after your feedback, I'm now really inspired to do it again.

Thx again !



------------------
Joachim
web: http://home.sol.no/~jbarrum/

[Joachim]

Hey Frost !

Don't be so sad and down. I would have answered your posts, if I could have helped you out. I don't want to talk about stuff I can't answer correctly

Make another cube as spooge said, without two lightsources and enviroment maps! I'm sure you will blow our minds..true

Btw,try to be on ICQ whenever you are awake and on the computer...I'm online whenever I'm sitting at the computer,which is pretty much all the time ...- I will cheer you up

------------------
Joachim
web: http://home.sol.no/~jbarrum/

[Gimbal]

Hey Samdragon. I'm relly in no position to critique your work but I'm going to give it a try anyway. I think it will help me look at my work with the same objective eye. Basically this is a disclaimer that I probably don't know what I'm talking about so take it with a grain of salt.

The boxes don't look as though they are solidly planted on the floor. I think this is due to the shadows. If you look at the cube on the right, the shadow doesn't crisply start from the corners of the cube but look to start from around the corners. This makes it look as if it is slightly floating above the surface. If you could make the edges of the shadow near the base of the cube come exactly from the corners and have a more defined, crisp edge it should help give the cubes a sense on weight as they rest firmly on the substrate. The edges of the shadow should only get slightly fuzzy as it gets further away.

Now when it comes to assessing the values used on the cubes, well, I'm just shooting of the hip with this one since I need to learn more about it myself. I think the tops of the cubes look good. The sides seem to have a very fast change in value halfway down the face of the cube. I think you went to dark near the bottoms of them. You also may need to lighten the shadow a bit to avoid everything looking as though it is grey instead of white. umm...oh, and maybe a relatively sharp dark line under the edges of the boxes where the shadow meets up with the cube? Not certain if that would be needed but since I don't really know what I'm talking about I might as well go for the gusto.

Bounced light? Haven't tackled that too much myself but spooges image shows that aspect of it pretty well. I'll have to reverse engineer that over the weekend before I can even pretend to know what I'm talking about. Going to paint a couple little wooden blocks I have with flat white spray paint and stare at them for a while.

Hope something I said was helpful. Wonder if someone can critique my critique

[samdragon]

yep, those shadows are too soft. I agree with what you said (Gimbal). That sharp fall off was my attempt at bounch light..yuk! All in the name of learning
I'll do some more after I finish what I'm working on now.Thanks for the info.


------------------
VISIT SAMDRAGON'S HOME

[Frost]

Fred; What I was waiting a reply on was the following.



Here I tried demonstrating how light is usually calculated in a 3d scene, where a surface is tested for lighting by it's angle from the light-source by casting rays. As shown above, the 50% (45 degree ray impact on the surface) area is lit quite a bit more than a simple 50% intensity of the light source - I was wondering if anyone had any knowledge of the light falloff with regards surface 'perpendicularity'(?) -- some kind of light 'remap' table of some sort.

Of course, this also depends on the type of material is being lit and how 'rough' that surface is (rougher being more 'diffused' as the minuscule bumps on the surface cast shadows over the next, etc.)... I guess that's also something I'm a little fuzzy on -- surface texture... Spooge mentioned making the cubes 'matte' ... but, for some reason, that seems a little vague to me, as I have problems understanding how much environment lighting and reflectiveness a 'matte' surface has...

Thanks for your time Fred.

cheers.
frost.

[Frost]

I just spoke to a programmer friend of mine who explained to me how it works. Without going into deep math (vectors, normals and dot-products), apparently the falloff is a simple cosine wave, ranging from 1 to 0 (for surfaces facing the light source, negatives are surfaces facing the opposite side of the light)sine/cosine wave. It is in fact non-linear.

I guess I should have asked him in the first place...

[This message has been edited by Frost (edited June 01, 2000).]

[eetu]

ahh, here we go again..

most 3d applications use a very simple formula, just a cosine of the angle (so-called lambertian model). so a 45deg hit would yield a 71% illumination.

this is a very fast and simple way to do it, but unfortunately it tends to give the renders a bit of a 'plastic' computery outlook. in the real world matte objects tend to reflect more of the light back to the viewer in low angles. in other words, the illumination of e.g. a ball is more even where brightest and then drops sharply when approaching total shadow.
an extreme real-world example is the moon. it's a matte ball lit by a distant lightsource. we all know what that kind of a situation looks in a renderer, but in real life the moon is lit evenly and then cuts sharply to shadow.

there are a number of 'smarter' models that are beginning to surface in modern renderers, the most common seems to be one called the 'oren-nayar' model. i won't go into specifics as no-one here is probably interested anyway, but here is the original 1994 paper by M. Oren and S. Nayar.

(i've actually written an oren-nayar shader for lightwave, that's why i happen to know all this shit )

and now, back to art...

ps oops frost beat me to it himself.. oh well.

eetu.


[This message has been edited by eetu (edited June 01, 2000).]

[AliasMoze]

Frost, I'm just taking a wild stab, but is it possible that the light on point on the sphere (where the light is perpendicular) is brighter because of reflected light from surrounding areas. On a flat surface, you normal don't have THAT much reflected light, but --with a sphere-- you have an infinite number of angled planes surrounding any point. What do you think?

------------------
AliasMoze
:) :) :) :)
"That activates my hilarity unit."

[Trance-R]

What's all this mumbo jumbo?

Teachers... could we pleeeese move onto the next lesson?~

[samdragon]

my second try at this.

I can't figure out whats wrong with the shadow on the right, any suggestions? I'm still trying to figure out the bounce light too.


------------------
VISIT SAMDRAGON'S HOME

[This message has been edited by samdragon (edited June 01, 2000).]

[AliasMoze]

Trance, I think the purpose of this exercise was mastery. Since many here (including me) haven't mastered it yet, we may stay here for a while. That's fine by me. I'll keep doing these cubes anyway.

------------------
AliasMoze
:) :) :) :)
"That activates my hilarity unit."

[Joachim]

If you find drawing cubes boring, I don't think the next lesson will be much more fun.

I don't mind moving on, but I still think it will be the exact problems and difficulties in lesson 2 as on 1, + a few others. It's still not anyone that has done any 100% correct cubes, maybe Sumas , so it's still a lot to learn.

Though I agree we should try not to get stuck on this either.

------------------
Joachim
web: http://home.sol.no/~jbarrum/

[Frost]

quote:

---

Originally posted by AliasMoze:
Frost, I'm just taking a wild stab, but is it possible that the light on point on the sphere (where the light is perpendicular) is brighter because of reflected light from surrounding areas. On a flat surface, you normal don't have THAT much reflected light, but --with a sphere-- you have an infinite number of angled planes surrounding any point. What do you think?

---

Well, I'm obviously not one to teach such matters, but, in my oppinion and from what I have gathered through looking and analysing lit objects, these rules are quite precise.

In the image I made, the falloff from white to black on the sphere is wrong, as I approximated by eye while being conflicted with not knowing WHY it was not lit as I had calculated -- before I learned how it worked. The image does not take into account ANY environment, light diffusion over distance, nor specular from the POV - it's a simple, one-point, ray cast lighting example over an (I would guess) absolutely matte object, projected in simple 3d without perspective. Imagine a simple light in a pitch-black void. The sphere IS an infinite number of angled planes, and I have only pointed out how to calculate rays for the extremeties and the mid point to give an idea of how it actually works.

The plane seems well lit to me (if you disregard the unprecision of it in a general sense)... you'll still get a brighter area where the ray is perpendicular and shorter to the light source. I have just taken into account in this pic the angle of the light to the surface, not the light diffusion/falloff over distance that would slightly influence this shading. If you have a wall lamp around your house somewhere, look at it and see the results... mind you, you'll get twice the falloff because of the angle AND the fact that house lights are very diffused and the falloff is great.

Of course, there are many other parameters you can add to this, such as projectors, colored lights, etc... but this pretty much describes the basic rules of angular falloff over surfaces (without any environmental or external influences), which I think is important to understand.

This makes sense to me, and I am willing to go by these rules unless someone points out flaws that can be somewhat proven. Anyway, I think this is right.

cheers.
frost.

[This message has been edited by Frost (edited June 02, 2000).]

[Zaphod]

Hi, I have a question. I haven't looked throught the last post, so I don't know if anyone else have mentioned this yet. So bare with me

Anyway: Wasn't this exercise supposed to be done with "natrual" lightsource? Cus I talked to my brother (who have done some of this stuff before) and he said that sun rays are always paralell (since the sun is Really big, and Really a long way from the earth). Wich means, no lightsource viewpoint, as everyone (including me) is using. What we're using now is artificial lightsources (lamps and such) wich means you'll have to use a vp.

I would illustrate what I mean, but I'm short of time right now.

------------------
/Zaphod
http://www.designmodule.com.bi

[Bilbo]

a bit late, i know, but here's my first attempt of the cubes assignment:



looking forward to hearing your pointers and advice.

[AliasMoze]

zaphod, you're right about natural light. Actually, I think Craig said that [natural light] is preferred. On my next one, I'm going to use it, instead of a lamp.

------------------
AliasMoze
:) :) :) :)
"That activates my hilarity unit."

[Frost]

Now I'm just splitting hairs, but, the sun is still a point light, and its rays are never parrallel (even though they are DAMN close to being that.=)

You see the convergance when the sun is in your viewport, uh, sight.

[synj]

what the heck guys hehehehe hundereds and hundereds of cubes. i know this could be argued all day, but, i really dont think knowing how to make a cube scientifically perfect is going to help you on that guy stabbing the zombie with a giant sharpened playdoh impaler. yeah, basics are applied to everything but how anal can we possibly get?

------------------
-dan 'synj' paladin
animator/producer
synj industries, inc.
Z.com

[AliasMoze]

Frost, yeah I agree.

Synj, these cube paintings help immensely. The world is made up of primitive shapes, so you have to learn how to paint them.

------------------
AliasMoze
:) :) :) :)
"That activates my hilarity unit."

[sydneyshan]

Very true. I'm completely happy to stay on this topic/ lesson until those who are serious about it have it perfect. I'm on a 5-week break from uni now, which means I've got some time to spend drawing cubes and getting my perspective and value describing skills down-pat.

I wouldn't hurry on to the next lesson until the people who are serious about getting good have mastered this first 'basic' step.

------------------
Shannon Murdoch
www.bounce.to/shan