[ Ancient Forge ]
Joined: April 15, 2002
Location: Calgary, AB. Canada.
|Posted on Sun Jun 03, 2007 11:32 am
|Link to Post:
| The wings should be easy. The arms should be okay. Legs will be bad. I've struggled with how to turn a tail into an ass into a leg continuing through to a torso for years.
Wings - Depends on what you're doing. If you're mailling the bone structure, just leave a hole and come back to it later, or build it flush and the come back and chop a hole. The specifics will entirely depend on posture. Some postures are easy (square out) others harder (escaping the side of cylinder at a high angle). I've dealt mostly with Euro 4-1, and it's not that tricky, you just have to use all your contractions (horizontal and vertical). Where the wing bones attach to the torso, try visualizing looking down the inside of the bone, end-on (as i looking down the barrel of a gun), arranged at the proper angle, against the torso sheet *already in form*.
Visualize the wing bone being a drill bit puncturing the sheet. Use a dowel or a pen if it helps. Try to see all the rings it will impact, those are the rings to remove/leave as a hole (it's probably easier to weave a complete sheet and selectively chop later). If your posture is square out, this will just be the cross-sectional area of your bone (presumably a circle). If you're rising out of flat plane, I think this deforms to an oval. If you're rising out of a cylinder at a sharp angle, as I recall it becomes somewhat teardrop or shoe-print shaped.
I don't know what type of wing you're building. Few samples:
(Image by Jeff Easley - Used without permission)
If you're building this type, (the Red dragon, ignore the Gold, and call it Wing Type 1) then you are done. This step was much trickier though than other wings types, as the shoulder bone and muscles are beefy, as a birds would have to be. The wing only connects at the one point.
(Image by Keith Parkinson - Used without permission)
If you're building this type, (bat-like, Type 2), you probably have the easiest time. A very small bone stump (hard to see in this one actually, easier in this More Famous Example), and then the fabric of the wings has to pin to the body as well. This requires minimal structuring or planning, you just stitch and go.
(Image by Larry Elmore - Used without permission)
This I think is the toughest of them all (Type 3). You have the front bone, the rear bone, and the fabric that lines up between them. It's only linearly tougher than the either of the other types for the actual seams (you just have to plot another hole for the rear bone), but it is a headache to design. You'll have one hell of a time trying to get the sheet geometry to match up to the bone geometry while simultaneously ending in a smooth geometry to attach the fabric to the torso. (I'm building this type, but I elected to go with leather for the fabric, avoiding this).
(Image by Michael Whelan - Used without permission)
If you want to try some kind of hybrid type like this, muscles, shoulders, and wing fabric attached, god help you.
So, Type 1, you're done. You just connect the large wing bone and that's it. The other types you have to seam the wing fabric to the torso.
To attach the fabric of the wings to the body... just cheat. I presume you're using fully chain sheet wings, not cloth or foil or anything like that? Take whatever pattern you have, and just loop a ring through each ring along the line of intersection. It would help if this is in a straight row line along the back and not at some non-matching angle where you have to slide sideways to other rows (design the back to flow with the wings, put expansions in the back, not on the wing, so you don't have a jagged wing seam).
With very tight E4-1, you can just barely slip a 5th ring rising out into the Z plane, as a 1-1 connection. With other patterns, maybe it makes more sense to loop through 2 (especially if looser). Just play around and pick something that doesn't deform the pattern. For advanced methods, pick something that *does* deform the pattern in a measurable way, giving the skin a stretched/bony look to it, like shoulder blades rising out of the frame.
From that line, just build your wing fabric with whatever pattern you're using. With E4-1, a row of 1-1 rings (like the odd rows of Trinity?) scoops into E4-1 immediately.
Hopefully I'm not breaking too many people's browsers here...
Tail points left, ends on the right at about the neck. This was my first gen dragon. I'm a few abortions past it now.
Sorry, it's a crappy picture. Hopefully you can find the back. See the two stripes? That's just a 4-1 sheet with a 5th ring looped through a single ring below it, in a row. They're rings standing on end, out of the sheet towards you. You can hopefully see they line up just like E4-1. Can't be too hard to do other patterns. (You can also see my first attempt at legs, more on them later).
Here's another example:
I think the upper-right quadrant shows it best. Heading left from the barely-started sheet on the right.. see how it climbs into the 3rd dimension for 2 rows, and the planes back out? I think that's another example of the exact same thing (and a bunch of other things going on, there's actually 2 strips of that and an aborted under-row to brace it into the 3rd dimension, don't look too hard, just showing how it's done).
That's assuming you're building your wings like this:
Left is bone, right is rear wing edge, bottom is spine, top is wing tip.
If you insist on it being sideways.. or if your spine is sideways.. well, I'm sure you can figure it out. You might have to leave off every other ring, or insert an extra ring.
If you're shrinking from, say, 16G on the torso to 20G on the wing fabric, just ease it down through smaller gauges and IDs, using expansions, easy enough.
So... long answer short, I don't have "A Seam" to suggest using, I can only suggest how to seam.
A simple thing that didn't occur to me until way later, is count the number of rings around your wing tube. Count the number of rings around your torso hole (the torso hole will be larger, of course). Eventually they have to merge, plan accordingly.
The trouble with this method is that your wing tube has one single edge, where your hole will have 4 edges (or 2, if you think about it that way). You can't just "expand out" from the hole, because that pattern isn't compatible with the 3rd dimension. So you have to come up with some way to twist the off-sides into heading in the right direction. A 45 degree seam can work, but is ugly. I suggest just expansions and contractions (which are simply 45 degree seams that live and die on 2-ring pairs at a time). The ugliest, but maybe easiest method would be to just turn the 2 sides that aren't biased properly, into the proper bias with a 90 degree seam (just loop 2 loose rings through each).
You can also create a gap through expansion and contraction, rather than by actually leaving out holes. I think I tried that once, with mixed results. Creates a vagina-shaped (double bell-curve?) hole. Which can look cool if you want more angular bones.. but you almost certainly need to expand sideways into the torso with it, or it deforms the torso (or makes the bone squish flat). Anyway, that's another thing to try if you get stuck.
I can't really offer any more specific advice than that, it depends entirely on your setup.
Arms - These should be just about identical to the interactions with the wings. Visualize through the bone, set at angle of impact, build outwards from impacted area in proper shape, likely using expansions.
Arms are half way between wings and legs. It might help to visualize the tube you're going to be building a certain amount of the way out, to help see what kinds of changes you need between the hole A, and the eventual upper arm B, and how fast those changes need to occur. Then just fill in the gap until you're comfortably down the arm. For wings, since it's just a straight pipe, you can probably just skip this, for legs, it will probably be essential.
Keep in mind that arms are more difficult than the wings (I suggest starting with the wings, then do arms, then do legs last [not fully, just get them started and do them in that order, the learning curves are sequential]). The wings just extended straight out and probably continued along that angle for a long way. Arms you need to do a lot of changes fast. Arms also don't just stick out at an angle, they have shoulders.
Also... I think you should avoid making comparisons between the arms on a chainmaille shirt, and the arm connections for a dragon. Arms on a chainmaille shirt (like humans) either extend straight sideways, or just droop. Arms on a dragon need to extend forward, so the interactions are a bit sneakier. Dragons don't stand with their arms along their bodies at their sides, they stand with them outwards like a person pushing a dumpster.
I think you can get away with leaving a hole, extending that tube a little ways, and then doing a 90 degree turn. The arms can't droop directly from the body, but they can't extend straight out either. They need to clear the body first, then extend forward once clear.
Actually.. maybe it would help you to think of a shirt, just rotate the sleeves forward with the armpits forward on the chest rather than down at the sides. Standard armpit expansions/seams should solve your problems for you there.
Arms are still basically just tubes leaving the torso though.
Legs - Legs are tough. Each thigh is almost as large as the waist, so you don't have a dominating portion you can just cut a hole in and modify. You have to plan it. Oh, and then toss in a tail.
The 4 pieces (Torso, L-Leg, R-Leg, Tail) all need to seam together all in the same area, they're all cylinders, and their zones of interaction all overlap. That is, you spread the transition between Torso and where the Leg has to be out, but this zone overlaps with the zones to transition between Torso and Tail, and Leg and Tail. So, in the middle of where you're plotting how to transition Torso-Leg, the same rings also need to be doing their job to change Torso-tail. If you can visualize all of these at the same time as you build, props to you. I never could.
I tried probably 5 or 6 different systems of building out the legs, you know, the "just start this and go with it, see how it turns out and what you can do with it", and nothing worked. I first tried considering the Torso and Tail to be the same continuous piece, just increasing in size, with Legs attached onto it. Legs didn't work. I tried building the legs as cones and then curling sideways out of the cones with expansions into a hole in the body, no luck. I tried considering the Torso and Legs the same piece and stitching a tail onto them, couldn't even get the Torso-Legs to work. I tried using 45-degree seams to scoop out from the tail to create asscheeks, no dice. I think you really have to be able to keep all 4 pieces (or 3, mirrored) in your head at the same time.
It also depends on the type of dragon you're building. If you look at a human, their stomach disappears into their groin, down to probably 10% of its previous width, the rest swallowed by the legs. The legs also grow past that, on the outside, to make the hips). I think this is the tougher kind, (the kind I'm building, which might be why I had so much trouble with it). Humans also don't have tails. But, if you look at actual lizards, this isn't really the case. Their legs sprout out exactly the same as their arms, hardly (if at all) consuming the torso. So you could just duplicate the arms, but bigger, and call it a day.
For illustration, back above, the Red dragon from Wing Type 1, this is a humanoid dragon. It doesn't show it, but I think it has a narrowed groin. That is, the belly plates must start wide at the torso, narrow to the groin, then get wider again on the the tail. An hourglass cutout. Any dragon that looks like it could stand on its hind legs without looking spread-eagled, is this type. (Most flying Easley dragons are like this).
See how the legs stick out sideways? It would scurry low to the ground, never able to rise up with its legs under it, like, say, a cat. (Here's a bigger image, of A Crocodile that also might help illustrate).
So, if you build it like an actual lizard, it might not be so bad (though, I tried this too, intending to just strip parts away later, and I couldn't even figure out how to get the legs to join then, either, let alone trimming). The problem is, just about every picture of a dragon I've seen, they don't scurry along the ground or have those kinds of legs. They are like humans with tails and long necks. They're like lizards with more bird-like legs for arial combat, that can point forwards from the torso and also help them land and absorb shock, which legs out to the side cannot. If I ever meet a dragon, I will be sure to question him on his evolutionary particulars.
Most dragon art I've seen is a mix of the above. They have legs that point forward and can crouch, but very little hourglassed groin (most just narrowing down from torso to tail, smooth). This means they must either have an extremely narrow thigh, or a *huge* ass.
(Image by Keith Parkinson - Used without permission)
If you look at that dragon, as the groin doesn't hourglass at all, but the ass is narrow, the picture hides it pretty well, but take a look at it and estimate how little thigh there must be at the back. Barely bone-thick I'd says. The leg:torso connection must be frail. Same as the More Famous Example I linked above. They have these massive feet, and even the front of the thighs look fine, but towards the back there can be almost no leg at all. (Most Parkinson dragons feature this).
Comparatively, on Wing Type 3 above, the Green dragon....
(Image by Larry Elmore - Used without permission)
... it might hourglass a bit, but mostly it doesn't. And you can definitely see the back of the legs is not weak. The picture again hides it well, but that dragon has a gigantic ass. Because the thighs can't steal away from torso width, the hips must be entirely beside the body. Image that thing on all 4's. It's ass must be like, 1.5 times as wide as the shoulders, (where a healthy human has it the other way around, making dragon asses of this type more than twice as wide as a human's). (Most Elmore dragons feature this style, or a hybrid style with slight hourglassing, which is what I like best).
Anyway, those are your design choices for legs.
As to how to join them.. well, with scurrying lizard legs, same as the arms.
With the other types, prepare to lose some sleep. It wouldn't be so bad if you only had to worry about shape, you could probably make do, but you also have to worry about pattern. You don't want your dragon to look ugly, or have scales running diagonally or mixmatched patches all over the legs. So, the challenge is, how to do it smoothly.
If the legs just came straight out of the torso and then went to the ground, I think it would be doable. But dragon legs are almost always in a crouch, with the leg brought up towards the stomach. Since maille doesn't stretch, that's really hard to do.
Back to my first gen:
Take a peak at the leg. I gave it almost no groin (you can see the big leg hole goes right up to where the tail will curl under). The leg rises out okay, and that leg line is actually a continuous row I think.. but the angle is wrong. It drops okay, but I wanted it to go forward towards the belly. Can't do it. The best I could come up with is that I needed 2 45-degree seams around 1/4 of the leg each, so the opening is forward, not straight out.
I also hadn't figured out vertical contractions (dropping columns) yet, which I'm almost positive you need to do.
I think an armpit seam might do some good here.
One visualization trick that might work, is to consider the torso a pipe. Then consider that big pipe to end at some point, and have 3 smaller pipes emerge from it all side-by-side, straight (in the same direction as the big one was, not off to the side at all). Center pipe becomes tail, the sides the legs. Build it out for a bit, maybe half-way to the knees, then do some vicious cutting and contraction on the front of the thighs to twist it forward.
Another trick I've found useful sometimes, is, even though you're dealing with cylinders (or cones), just think of them as boxes. If you can plan it in square terms, it can mostly take care of itself in roundness.
So, hopefully didn't drown you in infos there. If it's too much, just read less of it. I kinda rambled there for a while, thinking about you making yours has me thinking a lot about what I went through and what I'm going to do to fix it and figured I might as well write out my own reminders as much as for you. I'm pretty sure I have my specific problems worked out, I think I've finally got the legs, but it's more of a "Think I have a feel for it now" thing than anything I can explain.
I'm eager to see what you come up with, particularly with the various weaves you're using. I chickened out even considering anything more than E4-1. If you're stumped on anything specific, feel free to bug me. I'm of the opinion I've made just about every error possible, I know where they all lead