mesilliac's terrain graphics thread

Contribute art for mainline Wesnoth.

Moderators: Forum Moderators, Developers

Forum rules
Before posting critique in this forum, you must read the following thread:
Post Reply
mesilliac
Inactive Developer
Posts: 79
Joined: November 18th, 2005, 8:09 am
Location: New Zealand

Re: mesilliac's terrain graphics thread

Post by mesilliac »

@Sangel: yeah, that bleeding of the base terrain over the north edge of the cliff is the main problem now. The easiest thing to do will probably just be to redraw the north-facing cliffs to hide it as much as possible. Like JAP suggested, the base will probably be a new tile with some rocks to mask it a bit... but at the angle we're looking down at the north-facing cliff, you wouldn't see fallen rocks at the bottom of the cliff, so again it's not really an ideal solution.

Basically the cliff graphics need to take up more tile-space, without looking absurd.

@JAP: thanks for the useful cliff references :). The one that's closest to my image for how I'd like them to turn out is http://www.mccullagh.org/db9/d30-25/col ... -cliff.jpg . I have some good memories of rafting on the colorado river!

@Jetryl: yum, cookies :)

Re: giant spiders climbing cliffs.... this would be cool 8) but it's probably not worth the effort it would take to do. It's very easy to make the cliffs copy the characteristics of other terrains. As AI says, they will probably just act like a line of lava or chasm in terms of gameplay.

mog
Inactive Developer
Posts: 190
Joined: March 16th, 2006, 2:07 pm
Location: Germany
Contact:

Re: mesilliac's terrain graphics thread

Post by mog »

mesilliac wrote:@Sangel: yeah, that bleeding of the base terrain over the north edge of the cliff is the main problem now. The easiest thing to do will probably just be to redraw the north-facing cliffs to hide it as much as possible. Like JAP suggested, the base will probably be a new tile with some rocks to mask it a bit... but at the angle we're looking down at the north-facing cliff, you wouldn't see fallen rocks at the bottom of the cliff, so again it's not really an ideal solution.

Basically the cliff graphics need to take up more tile-space, without looking absurd.
Easiest way would probably be to move the away-facing tiles a bit up so they cover the bleeding terrain. Though it might then be difficult to distinguish the cliff from a base tile if there is a lot of base terrain showing on the bottom of the hex.

Can you commit the cliff changes? Or post them here? I'd really like to play with the graphics.
Aurë entuluva!

Mithridates
Posts: 14
Joined: July 26th, 2007, 10:44 am

Re: mesilliac's terrain graphics thread

Post by Mithridates »

I don't think that having cliffs as an alias of chasm would be a good idea, as then they become somewhat pointless due to the limitations of the chasm.
However, I am imagining implementing them in a way which would (probably) be horribly complicated. (and they would be a new terrain type)

I was kind of imagining that a cliff-ground transition hex would be made (a cliff piece where part of it has crumbled in the middle?) which would allow units to switch between the level below and the top of the cliff.
I think that units on top of cliffs should be able to see past them (in games with fog etc) while units below cliffs should not, unless they can fly. (this implies that movement can go down cliffs and not up which sounds horrendous- especially since I don't think that units should be able to move down or up them at all (unless they can fly))
I think that ranged attacks should be able to go up and down cliffs but not melee attacks (although due to HAPMA then you could argue that a unit on top or beneath could stay out of range or out of sight).
I think that attacks from one cliff to another (ie both units are on top of cliffs) should behave as normally on the terrain type which the top of the hill shows.

I hope I haven't completely discouraged you.
Mithridates

PS I don't think that I could help with implementing until the summer at the earliest-assuming I can comprehend the WML

mesilliac
Inactive Developer
Posts: 79
Joined: November 18th, 2005, 8:09 am
Location: New Zealand

Re: mesilliac's terrain graphics thread

Post by mesilliac »

Here's the current WML and graphics. It's not in a state to be committed to svn yet, but it's easy to work with.

"thinwalls.cfg" has the macro the cliffs use. And the macro is deceptivly named "path" ;)

I left the old never-get-on-top-of cliff in there.
cliffs.tar.gz
(313.29 KiB) Downloaded 441 times
@Mithridates: in their current incarnation, these cliffs don't technically have a top, or a bottom. It's all smoke & mirrors!

I hadn't thought about sight though. It would make more sense for units to be able to see from the "top" of cliffs. Still, IMO implementing this would be more effort than it's worth.

mog
Inactive Developer
Posts: 190
Joined: March 16th, 2006, 2:07 pm
Location: Germany
Contact:

Re: mesilliac's terrain graphics thread

Post by mog »

Most of what you suggest is not really possible without changing the way wesnoth works, but that part:
Mithridates wrote: I was kind of imagining that a cliff-ground transition hex would be made (a cliff piece where part of it has crumbled in the middle?) which would allow units to switch between the level below and the top of the cliff.
is a good idea (would be aliased to mountain).

@mesilliac:
Thanks.
Aurë entuluva!

User avatar
Espreon
Developer
Posts: 630
Joined: June 9th, 2007, 4:08 am

Re: mesilliac's terrain graphics thread

Post by Espreon »

mesilliac wrote:Here's the current WML and graphics. It's not in a state to be committed to svn yet, but it's easy to work with.

"thinwalls.cfg" has the macro the cliffs use. And the macro is deceptivly named "path" ;)

I left the old never-get-on-top-of cliff in there.
cliffs.tar.gz
@Mithridates: in their current incarnation, these cliffs don't technically have a top, or a bottom. It's all smoke & mirrors!

I hadn't thought about sight though. It would make more sense for units to be able to see from the "top" of cliffs. Still, IMO implementing this would be more effort than it's worth.
Lemme take a look at these, I will implement them in my campaign. Which editor group do these belong to?

Ebob
Posts: 38
Joined: October 19th, 2007, 2:52 am
Location: classified

Re: mesilliac's terrain graphics thread

Post by Ebob »

It might be a little past this in the forum, but why don't you make walls (house walls), take up a whole hex, make their movement cost 99? Then you can add wooden floor to make houses! (RPG heaven :geek: )
You should see my staff!

User avatar
Neoskel
Art Contributor
Posts: 724
Joined: November 27th, 2007, 5:05 am

Re: mesilliac's terrain graphics thread

Post by Neoskel »

On a related note, whatever happened to the water/chasm and deep water/chasm transitions someone was working on?

User avatar
Espreon
Developer
Posts: 630
Joined: June 9th, 2007, 4:08 am

~

Post by Espreon »

Neoskel wrote:On a related note, whatever happened to the water/chasm and deep water/chasm transitions someone was working on?
Yes! The fact that it has not been mainlined yet is driving me crazy!

mesilliac
Inactive Developer
Posts: 79
Joined: November 18th, 2005, 8:09 am
Location: New Zealand

Re: ~

Post by mesilliac »

Espreon wrote:
Neoskel wrote:On a related note, whatever happened to the water/chasm and deep water/chasm transitions someone was working on?
Yes! The fact that it has not been mainlined yet is driving me crazy!
You mean like... waterfalls? :o does anyone have a link?
Ebob wrote:It might be a little past this in the forum, but why don't you make walls (house walls), take up a whole hex, make their movement cost 99? Then you can add wooden floor to make houses! (RPG heaven :geek: )
Indeed... I am working on walls... which will eventually act like this :D

In fact I just used the wall WML that I already had for the walls to draw these cliffs... basically all the walls in page 1 need is a suitable texture and they would probably be perfect for small RPG houses :). They make nice corners and can join in T's and X's.

The current stone walls look a little "fat" if you use them this way (but it's not what they were designed for anyway).

Arr, I started too many projects at once.

User avatar
zookeeper
WML Wizard
Posts: 9742
Joined: September 11th, 2004, 10:40 pm
Location: Finland

Re: ~

Post by zookeeper »

mesilliac wrote:
Espreon wrote: Yes! The fact that it has not been mainlined yet is driving me crazy!
You mean like... waterfalls? :o does anyone have a link?
Try here: http://forum.wesnoth.org/viewtopic.php?f=9&t=18010.

If those transitions were ported to your cliffs, we could really have proper waterfalls. ;) Since it'd be potentially annoying to have to draw graphics for water pouring over the edge both when there's water on the other side (like with a waterfall) and when there isn't, it would probably be ok to simply limit it to only working with the former case.

mesilliac
Inactive Developer
Posts: 79
Joined: November 18th, 2005, 8:09 am
Location: New Zealand

Re: mesilliac's terrain graphics thread

Post by mesilliac »

So I think I worked out the transitions for interior floor terrains.

They do take up a lot of space around the tiles, it is true. But, it allows me to make a thin wall variant (like in the shot here), and have the floor come right up to them. Also a "door" should be fairly easy to do.

Thus, this is the pretty-much-final floor template and .cfg:
floor-template.xcf
template for interior floors
(99.96 KiB) Downloaded 425 times
floor-template.cfg
just has one extra macro for nice iso corners when needed
(1.4 KiB) Downloaded 388 times
The screenshot isn't so impressive, but here it is anyway. The official Wesnoth "graph paper" terrain!
(yes these are the template images)
notimpressive.jpg

Sangel
Moderator Emeritus
Posts: 2232
Joined: March 26th, 2004, 10:58 pm
Location: New York, New York

Re: mesilliac's terrain graphics thread

Post by Sangel »

You've rendered square-based geometry... in an isometric plane... on a hex-based field... with tessellating properties... :shock:
"Pure logic is the ruin of the spirit." - Antoine de Saint-Exupéry

mesilliac
Inactive Developer
Posts: 79
Joined: November 18th, 2005, 8:09 am
Location: New Zealand

Re: mesilliac's terrain graphics thread

Post by mesilliac »

The hard part was finding a spiritual medium to contact Escher who spoke Dutch :augh:

mesilliac
Inactive Developer
Posts: 79
Joined: November 18th, 2005, 8:09 am
Location: New Zealand

An essay on terrain perspective Part 1

Post by mesilliac »

So, umm, I'm gonna have to eat my words a bit. I said elsewhere that drawing something from top-down then squishing it vertically to 50% of its height puts it in the correct perspective to be lying on the ground. I was wrong (sorry Turin).

To explain, here's an essay on terrain perspective in wesnoth ;). To summarise (because tbh the rest will be a headache to read):

For terrain only:
Something drawn from top-down should be squished vertically by 1/3 to 66.666...% of its original height.
Something drawn from side on should be squished by 1/4 to 75% of its original height.


Unit perspective is not the same. It looks better this way.

Now for the essay.

You might already know that a regular hexagonal grid has a kind of inherent perspective. The hexagons align at 30° angles, implying an Isometric Projection.

Wesnoth tiles aren't regular hexagons, but they happen to have some very nice properties anyway. Either high-school trig or playing around with the measuring tool in GIMP will tell you that the inherent angle that wesnoth hexes are aligned on is 33.69°. This is close to isometric. As close as most "isometric" games, which usually use a 26.565° angle. The reason it is some funny angle is that this is the angle you get when you move 3 pixels horizontally and 2 pixels vertically (26.565° comes from moving 2 pixels horizontally and 1 pixel vertically).
wesometric
wesometric
geomertry1.png (9.51 KiB) Viewed 6135 times
For the purposes of this post, I'm gonna call this perspective "wesometric" ;).

To find out how much this view was squished from the top-down view, just expand it until the wesometric square is a true square. A little math or playing around will give you a vertical expansion to 150% of the original height. This shows you the "true" top-down view of a wesnoth hex, as implied by this perspective.
implied top-down view of wesnoth hex
implied top-down view of wesnoth hex
geomertry2.png (11.93 KiB) Viewed 6126 times
In corollary, if you have drawn something from top-down and want to scale it to fit in with this perspective, scale the image to 2/3 of the original height

Part 2 to follow...

Post Reply