Pro Mistakes

[DCillusion]

Here's something I found that may give people hope in this forum if they think their art/shading/perspective is bad, just know that everyone makes these mistakes

It's a screenshot from "The Colonel's Bequest" at the time, it was touted for it's INCREDIBLE graphics.


Check out the SHADOWS!!!

1 side bounces off the ground while the other side cuts through the floor like a knife!!

[Mr_Frisby]

I think that you'll find that there are two moons  .

[HillBilly]

Oh my god!

Kill the beast!

[DragonRose]

Okay, I feel really stupid but I don't see anything wrong with this. Then again, I'm horrible with lighting. What should the artist have done instead?

[HillBilly]

I made this to illustrate the error. Took me 3 hours in photoshop.

[Babar]

I still don't get it. If the light source is somewhere above the top right corner of the room, the falling light should be angled even more downwards that it already is, not upward. Or not?

[HillBilly]

One of the shadows shouldn't be there at all, unless there's two light sources. And I assume the original light source is the moon, like Mr_Frisby said. Atleast that's my guess.

And if the moon IS in the upright corner, I still don't think the light looks right. 

[Snarky]

Sierra's 16-color SCI games have notoriously poor perspective. It's been causing AGDI some difficulty when painting over the QFG2 backgrounds.

When you have shadows of parallel lines projected on a flat surface by the sun or the moon, the shadows are going to be parallel as well. In perspective drawing, that means they'll have a common vanishing point.

So it should look something like this:



Come to think of it, I'm pretty sure the vanishing point is the position of the sun/moon.

[Oneway]

You are correct DCIllusion. Light that reaches earth from large astral bodies is presumed to be all going in the same direction. The bodies are of such a large size that the light that reaches us is virtually parralel.
Light from the moon, being a LOT smaller than the sun, may be less parallel, but not to an extent that we could notice it by the naked eye.

Parallel lights mean parallel shadows and thus will they share a vanishing point.

q.e.d.

[Pod]

if it IS the moon, shouldn't the vanishing point be moved back a lot further?

[Oneway]

hehe, you're confusing 2d with 3d.Ã,  In 3d, parallel lines stay parallel and don't have a vanishing point.
The vanishing point is a technique that helps make 2d images seem 3d.

So in this 2d picture there is no 'back'. Only up, down, left and right.

Does that make sense?

[kjetting]

Well I, for one, would like to see some more pro mistakes if anyone has screenshots.

[DragonRose]

Thanks for drawing that, Snarky.  Quite helpful.

[Oz]

Snarky's corrected image is... correct.

Just draw each line from the light source and make them pass through the corners of the baselines of the windows, and there you go!

[Xadhoom]

That's a good trick, Snarky. Didn't know that one. Thanks.

[Blackthorne]

I can't believe this.

I feel so betrayed by Sierra.

Now, I can never play their games again.  In fact, I'm flying out to Coarsegold tommorrow to piss on the ashes of their dead company.

Bastards......

Bt

[loominous]

As far as I can tell, they re not far off the mark:



Snarkys positioning of the lightsource wouldn't really cast those kind of shadows; it needs to be higher up.

The right windows (our POV) are a bit off though.

[Snarky]

I don't know what the hell you're doing, but I know it's not right.Ã, 

Could you explain the reasoning for your sketch? It looks like you're drawing lines from points on the objects to the corresponding points on their shadows, and finding that they all intersect in two (wtf?) points. That's not the way shadows are cast, though.

Edit: OK, I see what you're doing. The lower point is the vanishing point in the plane for the parallel lines, and the higher point is the position of the moon. Yeah, that seems reasonable. I'd say the shadows on the right are quite far off, though.

Edit 2: Actually, all of the shadows are completely off. I just traced them, and no three intersect in the same point. What I'll certainly grant you is that the vanishing point is not the position of the (infinitely distant) light source, but that the source is somewhere along the normal to the plane from the vanishing point. I can't upload images from the computer I'm at, but I'll make a diagram that shows how to construct shadows (from sun/moon) correctly.

My sketch is still correct, though. The point where the lines intersect isn't the position of the lightsource, but that's not what I claimed.

[loominous]

This is the technique I m using:

http://www.saveloomis.org/fun/110.htm

Edit:

I now realise you weren't placing the lightsource but just showed the perspective. Thought you meant the moon was supposed to be placed that low which would result in much longer shadows.

[Snarky]

Hey, neat link!

Yeah, the point I was using was just the perspective for the parallel lines in the shadows. When I had done it it occured to me that it had something to do with the position of the light source, but I didn't realize I had to draw the normal to the plane at that point. If the moon was actually at that point it would be exactly at the horizon, and shadows would be infinitely long. 

I still can't make the lines in the original picture's shadows meet up neatly in a point the way you indicate. I tried out the method by doing basically what I assume you did: using some of the lines to decide on the vanishing point and the position of the light source, and projecting the shadows from there. It came out virtually identical to my first sketch.