Thanks, I'll keep those things in mind...though the open/close and firing animations make up for the barrel's shortcomings imo :)Argh wrote:1. Spend more time on the geometry. The gun is just a repeating series of barrel shapes, and that's the most interesting part of the whole model. While the texture can do serious heavy lifting and add a great deal of life to the piece, you've gotta have somthing more interesting to start with as the basic form.
The big areas that stick out, for me at least, is the turret's base, which would look cooler even if you just made it slope slightly from top to bottom (or the inverse), and the "popup box", which should get some greebles around those corners, or something, to give it a little more oomph. I like the cannon itself, although you really need to use the vertex-normals stuff that I described in the post above this one to make it look fully round in Spring.
Random WIP 2006-2011
Moderators: MR.D, Moderators
-
Guessmyname
- Posts: 3301
- Joined: 28 Apr 2005, 21:07
For when it rends opposing players' simbases limb from limb in fury 
Er, I mean, um, when it stops and shoots, it needs to plant these before it can fire. It's a siege mortar. Once planted, it will sit still for awhile- you don't use these up close to the battle in progress, but rather when you're besieging somebody's OMFGBBQ simbase with like HUGE DEFENZ!1 From a slightly shorter distance than their longest-range regular artillery can reply, you can demolish their base with these.
Slow, but steady means to defeat a defense- pretty much useless against attackers, and you'll probably want to put them on Hold Fire (hrmm, isn't that possible to set as a default, in COB or FBI? hmm) and target things manually with them, because the ballistic arc will take quite awhile to complete, and they don't fire very rapidly.
Er, I mean, um, when it stops and shoots, it needs to plant these before it can fire. It's a siege mortar. Once planted, it will sit still for awhile- you don't use these up close to the battle in progress, but rather when you're besieging somebody's OMFGBBQ simbase with like HUGE DEFENZ!1 From a slightly shorter distance than their longest-range regular artillery can reply, you can demolish their base with these.
Slow, but steady means to defeat a defense- pretty much useless against attackers, and you'll probably want to put them on Hold Fire (hrmm, isn't that possible to set as a default, in COB or FBI? hmm) and target things manually with them, because the ballistic arc will take quite awhile to complete, and they don't fire very rapidly.
@Smoth:
You mean polycount, correct? Well, that kind've depends, on a case-by-case basis, frankly. I'm not going all Caydr here, if that's what you want to know, and I'm trying to use minimal polygons to do the job.
The mine is about 1400 tris, the tank is 777 (haha!) and the fusion plant, which I have in-game, is a mere 550. I'd like to keep everything under 1500, except for the occasional thing that really requires more detail, but I've decided to be flexible about polycount, as opposed to my normal level of anal-retentiveness, in the interest of modeling speed. I'm modeling these over less than an hour, on average, and it's taking about 2 hours to uvmap and paint them, which is respectable speed for this kind of thing.
You mean polycount, correct? Well, that kind've depends, on a case-by-case basis, frankly. I'm not going all Caydr here, if that's what you want to know, and I'm trying to use minimal polygons to do the job.
The mine is about 1400 tris, the tank is 777 (haha!) and the fusion plant, which I have in-game, is a mere 550. I'd like to keep everything under 1500, except for the occasional thing that really requires more detail, but I've decided to be flexible about polycount, as opposed to my normal level of anal-retentiveness, in the interest of modeling speed. I'm modeling these over less than an hour, on average, and it's taking about 2 hours to uvmap and paint them, which is respectable speed for this kind of thing.
Ah, thanks, I'll have to download BA and get that code integrated, then, or maybe I should just make this flat-out commandfire, dunno yet. I think players will be very annoyed with this unit if they aren't calling the shoot, though, if it fires at a mobile front that your own units pass through, then you're likely to lose a big chunk of your force!There's a Lua widget that already does that with some BA units.
Keep practicing, read tutorials on how to save time with (insert name of paint proggy here). It's the only way. There are no fancy tricks that can magically make you faster, other than knowing your tools and what you want things to look like when you're done.
I'm basically doing these skins in Photoshop, using nothing more complicated than bevels, fills using various filters, and a lot of Pattern Stamp to do the dirt, combined with judicious use of the Dodge tool, to "bleach" areas gently and simulate wear a little more effectively. I use a surprisingly small number of textures on a regular basis- it actually helps a lot, because it keeps color values consistent in final touchup, so that I don't have pieces that look "off". I have a huge library of textures I've downloaded, but frankly, they're just there for when I'm out of ideas or need a really specific mechanical greeble in a hurry, because they always require repainting to get the light angles right, which is a pain.
It's really not terribly hard- what's hard, frankly, is getting the feel right, and that just takes looking at a lot of reference photos of real-world materials, if you're trying to go for realism, or studying the way that line and value is used to convey meaning in toon work. Either way, there's a lot of practice involved. Just keep going, and if you're stuck, always go Google a bit- for a lot of effects, there are tutorials on how to achieve a given look.
Personally, I think most of them are fairly useless, because they're pointed towards 2D people, or the guys who turn out sigs consisting of someone else's work, a beveled font and a lensflare and think that makes them artists
But then again, I'm pretty comfortable with my process. The only stuff that makes me stop for a minute is when I have a really specific greeble I need, and I've found that the easiest way to do that is to make it in Rhino really fast, get the lighting angle right, and render it flat, then import to paint in color. It's really easy, and since it's just a greeble, I don't care about polycount, so I can just go nuts and use all of the high-poly tools I never use otherwise, like lathes and complex lofts.
On the project that I was working on, I had to do a lot of realism again, which was the first time since NB, which was very intentionally not supposed to look real, other than the Knight II. So I'm still in "realism mode", artwise. It's actually rather boring, frankly, and this monochrome palette is driving me bonkers- expect to see more (muted) color on the structures, just to relieve my sense of style, if nothing else...
And that's what it looks like, with the legs retracted. Nothing fancy, just rotates each leg at a slightly-different speed when moving them, to make it all look more complicated than it is, and some FX stuff when they land.
The "sit still and do not move at ALL" script is working well, although I'm a bit unhappy about how inefficient it has to be, in order to work perfectly.
I'm basically doing these skins in Photoshop, using nothing more complicated than bevels, fills using various filters, and a lot of Pattern Stamp to do the dirt, combined with judicious use of the Dodge tool, to "bleach" areas gently and simulate wear a little more effectively. I use a surprisingly small number of textures on a regular basis- it actually helps a lot, because it keeps color values consistent in final touchup, so that I don't have pieces that look "off". I have a huge library of textures I've downloaded, but frankly, they're just there for when I'm out of ideas or need a really specific mechanical greeble in a hurry, because they always require repainting to get the light angles right, which is a pain.
It's really not terribly hard- what's hard, frankly, is getting the feel right, and that just takes looking at a lot of reference photos of real-world materials, if you're trying to go for realism, or studying the way that line and value is used to convey meaning in toon work. Either way, there's a lot of practice involved. Just keep going, and if you're stuck, always go Google a bit- for a lot of effects, there are tutorials on how to achieve a given look.
Personally, I think most of them are fairly useless, because they're pointed towards 2D people, or the guys who turn out sigs consisting of someone else's work, a beveled font and a lensflare and think that makes them artists
But then again, I'm pretty comfortable with my process. The only stuff that makes me stop for a minute is when I have a really specific greeble I need, and I've found that the easiest way to do that is to make it in Rhino really fast, get the lighting angle right, and render it flat, then import to paint in color. It's really easy, and since it's just a greeble, I don't care about polycount, so I can just go nuts and use all of the high-poly tools I never use otherwise, like lathes and complex lofts.On the project that I was working on, I had to do a lot of realism again, which was the first time since NB, which was very intentionally not supposed to look real, other than the Knight II. So I'm still in "realism mode", artwise. It's actually rather boring, frankly, and this monochrome palette is driving me bonkers- expect to see more (muted) color on the structures, just to relieve my sense of style, if nothing else...
And that's what it looks like, with the legs retracted. Nothing fancy, just rotates each leg at a slightly-different speed when moving them, to make it all look more complicated than it is, and some FX stuff when they land.
The "sit still and do not move at ALL" script is working well, although I'm a bit unhappy about how inefficient it has to be, in order to work perfectly.
-
[KnoX]ElementalGizmo
- XTA Developer
- Posts: 266
- Joined: 24 Aug 2006, 01:33
does it "deploy" like the sc tank for terran? the barrel looks v nice with the detail of burnt edges through soot. the cut sheet metal is something im working on, and hope to include in with the texture im doing for Peets model.
as u can see i have plans for "sheeted metal" lolz but i havent thought about the shadows of each panel so to speak as ur texture presents, which i do like!
But yeah i need more greebles for added UMPH, although adding a cannon burn effect as u have done is some what out of the question as the same texture is repeated for each section of the expanding cannon....any suggestions for this dude ?
Peet has just told me he is going to modify his model slightly without changing the uv map, hopefully this will leave me with more options for detail for a more interesting texture
when rl gets out of the way im going to present a more defined texture with attention to more greebles that i havent had time to work on atm, but as tis is my first presentable unit texture its very nice to receive a complement to start with
i hope i can up the standard but obviously i need concept support as im such a nub atm, but im planning on this current state to change very soon lol
as u can see i have plans for "sheeted metal" lolz but i havent thought about the shadows of each panel so to speak as ur texture presents, which i do like!
But yeah i need more greebles for added UMPH, although adding a cannon burn effect as u have done is some what out of the question as the same texture is repeated for each section of the expanding cannon....any suggestions for this dude ?
Peet has just told me he is going to modify his model slightly without changing the uv map, hopefully this will leave me with more options for detail for a more interesting texture
when rl gets out of the way im going to present a more defined texture with attention to more greebles that i havent had time to work on atm, but as tis is my first presentable unit texture its very nice to receive a complement to start with
i hope i can up the standard but obviously i need concept support as im such a nub atm, but im planning on this current state to change very soon lol
Sort've, but not really. It comes to a complete halt, puts the legs down, with dust and stuff, then it can fire. While in "firing mode", it cannot move, nor can it turn.does it "deploy" like the sc tank for terran?
But I don't have a dual-purpose mode, where it can fire while moving, but can fire farther and do more damage while sitting still. That wasn't the purpose of the unit, frankly- I wanted something that filled a very narrow game-design goal of "long-range heavy mortar". I've been wanting this kind of unit in Spring for well over a year- FLOZi's patch, allowing me to SET HEADING, finally allowed me to do what I needed, although there is a rather high cost, because the script has to update every frame (unlike SET MAX_SPEED, SET HEADING doesn't override the wantedHeading variable, so the Unit tries, every frame, to turn itself towards the target).
I plan to implement a tank that can do that, though, as it's one of many SC game design concepts that's now possible in Spring, and it'd be fairly easy to script, as well as possibly fun to play (although, I must say, given the way that Spring mortars work, I doubt very much that it'd work at all like the SC gun, but ya never know- maybe I can do a dual-weapon swap with a low-velocity, high-damage mortar, or something).
-
[KnoX]ElementalGizmo
- XTA Developer
- Posts: 266
- Joined: 24 Aug 2006, 01:33
I just use the standard brushes, usually. Sometimes I use one of the preset Photoshop brushes that is basically some random blobs, but rarely- most of the time, it's not as effective as other methods. The Pattern Stamp tool is the main thing I use to create effects quickly- using a pattern, you can change value and hue at the same time, with random-looking effects. While some people make a big deal out've custom alpha brushes, etc., I don't see the point- either I want 100% coverage, and am using the Pencil or Eraser, or I want soft edges, in which case I'm airbrushing, pattern-stamping, or dodging. I basically do long, sweeping passes with airbrushes and pattern stamp, to give a texture area a general feel, then narrow the brush size considerably and zoom in, to focus on specific areas of detail.
One of the biggest things you can do to help yourself do more realistic stuff faster is that you should keep the amount of flow as low as you can stand. I typically use about 3-5% with the airbrush, and from 10-30% with Pattern Stamp, depending on what I'm doing and how much change I want to see. The reason to keep the flow so slow? Simple- it means that you can make small strokes, and you can have areas of "overspray"- i.e., small mistakes of aim- and they're hardly noticable, so you can use a much more natural "scribbling" motion with your hand, instead of aiming for inhuman perfection. If you want inhuman perfection- well, that's what Bezier curves are for, duh! If you want to impart the sense of billions of random events on an object, however... well, then it's best to have lots of variation, and many small mistakes will seem natural to viewers, because Nature is full of random crap that isn't perfect, too.
One of the biggest things you can do to help yourself do more realistic stuff faster is that you should keep the amount of flow as low as you can stand. I typically use about 3-5% with the airbrush, and from 10-30% with Pattern Stamp, depending on what I'm doing and how much change I want to see. The reason to keep the flow so slow? Simple- it means that you can make small strokes, and you can have areas of "overspray"- i.e., small mistakes of aim- and they're hardly noticable, so you can use a much more natural "scribbling" motion with your hand, instead of aiming for inhuman perfection. If you want inhuman perfection- well, that's what Bezier curves are for, duh! If you want to impart the sense of billions of random events on an object, however... well, then it's best to have lots of variation, and many small mistakes will seem natural to viewers, because Nature is full of random crap that isn't perfect, too.
tbh these sorts of cob scripts would be better done in lua as then with the eventual lua<->AI interface planned for early 2069 you'll finally be able to tell AIs that the unit cant move while firing and you can do things much better, maybe even create a deploy and undeploy button in the UI.
As for the models itself, a general note in design.
In military chassis design its best to present the smallest area possible to the enemy. Smooth round surfaces giving a smaller surface area than big wings and jutting out corners. To this end if you look at the existing designs in games and real life things like arms and stuff retract from the main chassis when it deploys. That way the assembly and machinary is less vulnerable to enemy fire.
However your deployment arms are all on show. And the method used of pulling them up into the air and supporting them in their vertical position is highly inefficient and from a military design perspective, most undesirable.
Thus it would be more appropriate if your deployment arms were melded into the main body of the unit, and when deploying moved out a little bit then went downwards. This would be the best design as its far more efficient and provides a far smaller surface area for weapons fire, as well as providing shielding to the main mechanism rather than exposing it directly to the outside.
As for the models itself, a general note in design.
In military chassis design its best to present the smallest area possible to the enemy. Smooth round surfaces giving a smaller surface area than big wings and jutting out corners. To this end if you look at the existing designs in games and real life things like arms and stuff retract from the main chassis when it deploys. That way the assembly and machinary is less vulnerable to enemy fire.
However your deployment arms are all on show. And the method used of pulling them up into the air and supporting them in their vertical position is highly inefficient and from a military design perspective, most undesirable.
Thus it would be more appropriate if your deployment arms were melded into the main body of the unit, and when deploying moved out a little bit then went downwards. This would be the best design as its far more efficient and provides a far smaller surface area for weapons fire, as well as providing shielding to the main mechanism rather than exposing it directly to the outside.
-
[KnoX]ElementalGizmo
- XTA Developer
- Posts: 266
- Joined: 24 Aug 2006, 01:33
okay been pratting about, got the top section now in panels and perfectly tiled so u cant see the stupid lines from the 4 triangle that make the square up as in the las pic.
Used some very gentle beveling on this for each panel so its not so bland now, any better?....further suggestions?
@Argh
tnx for the tips dude, but i think AF might be on something with the hole deployment arms. Heres some pics i just found of the tank in both modes..
having the deployment arms internal just looks better in these pics, it also makes more sense to shield them within amour than exposed and highly vulnerable.
Is it going to use a 2phase hp cob setting similar to the Zeus that Noruas has done to the XTA Zeus?
@AF:
First, about the script... if you can write a LUA telling the Unit that it may not MOVE or TURN (turn being the biggie- making something not move in Spring is trivial now) that will hook through LUACOB (because, duh, it must get started/stopped based on animation timing) ... I'll be more than happy to use it!
As for the rest of the critique... thanks for your input- I think it's incorrect, and greatly underestimates the real-world issues involved, but I appreciate the thought. Here is additional food for thought.
@EG:
Those are two completely different models, lol! And, in the both cases, I see no armor covering the "foot"- they just used the "foot" to cover the rest of the mechanical details. But yes, it would probably look cooler to do that, I agree. However, I am more interested in getting more models done, than in reworking one that already works, for a small increase in coolness. I will certainly look into that script, btw, but I'm pretty sure I can ON/OFF a tank-mode script with what I already have, frankly.
First, about the script... if you can write a LUA telling the Unit that it may not MOVE or TURN (turn being the biggie- making something not move in Spring is trivial now) that will hook through LUACOB (because, duh, it must get started/stopped based on animation timing) ... I'll be more than happy to use it!
As for the rest of the critique... thanks for your input- I think it's incorrect, and greatly underestimates the real-world issues involved, but I appreciate the thought. Here is additional food for thought.
@EG:
Those are two completely different models, lol! And, in the both cases, I see no armor covering the "foot"- they just used the "foot" to cover the rest of the mechanical details. But yes, it would probably look cooler to do that, I agree. However, I am more interested in getting more models done, than in reworking one that already works, for a small increase in coolness. I will certainly look into that script, btw, but I'm pretty sure I can ON/OFF a tank-mode script with what I already have, frankly.
-
[KnoX]ElementalGizmo
- XTA Developer
- Posts: 266
- Joined: 24 Aug 2006, 01:33
but PLEASE refer to me as Giz or Gizmo. "EG".....not cool man 