[WIP] Modeling a Scythe

No, i did not (made tesselations or smooth by subdivision) =D.. I used the Smoothing Groups :)

(NOTE: In my opinion, even to higher poly Bladed Weapon's model, applying favorable Smoothing Groups is crucial)


A smoothing group is a group of polys in which one has a smooth transition to the other on the same group.. But denotes its transition on the boundary to another smoothing group. It is very useful to do instead of subdividing.. A low poly model is way easier to handle..

Example (bottom end with and without smoothing groups):
http://i39.tinypic.com/ogmcnq.jpg

It is uncommon for me to apply smoothing through subdivision, but, when i do, mostly it is after having modeled the mesh properly (for an example, if you are well familiarized with 'TurboSmooth' (i mean, you kind of know how it will act), you can make an ridiculously low poly (the lowest possible poly count) character model with some reliefs on the overall format of the muscles, and (after making the lowest poly muscle reliefs) apply the TurboSmooth, as amazing as it can sound, it does not make the model heavy or too high poly, but smoothes it quite perfectly (but you have to be VERY familiar with those forms of smoothing techniques, or you can create a lot of problems on your model (and the geometry has to be very organized if you want to apply smoothing subdivisions without having critical problems..)

Example (before and after TurboSmooth, on the creation of a 'Samurai' character, for UDK) very simple forms and did not generate any 'funny' geometry... but you have to be very careful on subdividing your meshes.. always save first!!:
http://i44.tinypic.com/2e584zp.jpg



I make this kind of model 'thinking low poly' (though this one is not sooo low poly..), to use it in games.. If i subdivide it, it will become too high poly to be comfortable for games..

Try to make it via Box Modeling on Blender, i am very sure you will learn a lot.. =D

Quote:

---

Originally Posted by Sparkler

DIdn't you use anything like a subdivision modifier? (it subdivides and smoothes things out, I'm sure there's something like that in 3ds max too)

---

Its called turbo smooth, but it is not recommended for ingame modeling and stuff like that, but only for High polygonal modeling and renders.

Say for instance I was making a model for WarZ, the limit to the amount of trigonal faces for that game is about 5000 per weapon when used in a 3rd person scenario. To run objects in real time, you must limit the amount of data on the screen.

Now cases where a Turbosmooth/ Subdivison modifier would be used.
High Polygonal modeling (ZBrush for instance) in an effort to bake Ambient Occlusions and Normal Maps. In this scenario you would make a low polygonal model as well as a high polygonal model, and your modeling program would create different kinds of textures based on the differences between it. Normal maps control how light is reflected while Ambient occlusions give a very basic shadow structure.

High Polygonal modeling in an effort to render a product. Since rendering has no limit to the time it takes to produce one frame, you can easily produce objects that would normally be too much for a system to handle. Also, since you can do more than one pass, the system can correct any mistakes it makes on prior render passes.

Edit: Here is a set I created for GunZ a few days ago (from scratch I might add... handpainting was pretty fun lol). As you can tell, it has very few polygons, not even hitting 1000. This is quite normal for a game that came before 2006. (If her butt looks a little big that is because I have to account for softbodies, giving it a little space so the cloth can move freely)

Spoiler:

http://puu.sh/39kpS.png

http://puu.sh/39kH3.png

@ ScreemingFox.
I do not have any idea why you would do that type of modeling with the Samurai figure. You could just as easily create your low poly version, then sculpt the subdivided version of it until it fits your needs. Creating a low poly model for the very reason to try and smooth it seems like a waste of time. I always use my low poly version as a base, then work from there, never make 2 different low poly version in an effort to achieve my high polygonal maps. I mean it is cool and all, but it just seems like wasted time and a really bad workflow when it could be streamlined taking the easier method.

Length looks much better now.

Yes.. and i slightly reinforced the shaft =D
http://i39.tinypic.com/14y0zt.jpg
(ingame)



Here is the link, with everything needed..:
http://www.filedropper.com/infernoscythe
(@edit: Needs some adjustments on the blade, some parts got too thin ;d)



@wucas: That body i did from the Zero.. i had no 'body' to base some clothings, armor pieces, etc, so i though of making one out of a box XD.. i also thought that it would take a lot of time, but it was kind of quick really, i tried to do it without any 'blueprints', and based the arm on my own (i am kind of fat so, some muscles will not show, hehe).., but the only part that will be kept to the final character will be the chest/neck region, the rest will be covered with clothes, etc... I made it with an old 3dsmax (i think 7, without that graphite modeling tools so it was fun to also learn that side panel (of Edit Poly) properly) at work, it took me, kind of 2 hours i think, but the part that took the most of the time was the hands.. XD..

Oh it looks nice, but isnt it a bit toooo big, even for a scythe?

@Wucas and Screaming fox
THanks for the detailed text on smoothing


EDIT: I started modeling this again, but now through box modeling, kind of. I started by doing the handle, with a 7 sided cylinder, cuz I feel like if I do it with a square I'm doing to take up more polys to make it roundish (even with the smooth modifier)

Should I make the handle and the blade 2 different object, or should I just merge them

Yes, it is kind of big indeed. (added to arsenal, named it 'Kaphiroth Scythe' muhuhuhehehe XDD)

I think it would be a nice weapon to monsters.. Specially fliers, since it is a bit disproportional..

You can make with 2 separate objects or model everything from the same mesh, i think both methods are valid.. I prefer to model everything from the same mesh but there are a lot of situations when it is better to model different meshes and then attach them..




You could post here the other drawings from your friend =D

She said she'd like me to model them, not other ppl :( sorry ^^

EDIT:
I've been working and studying the topology of hard surface meshes, this is how half of the blade turned out:

[img]http://i.imgur.com/0523KtA.png[/img]

How is it?