*LATEST* (more later in thread)
I've modelled this real time lightsaber based off of Obi wan's third lightsaber. I've been playing a lot of Swtor with friends and that is where the motivation came from ^^
My texturing isn't finale, i'm looking for critique and such. Also i have no idea how to handle the actual blade part.
(The waviness in the normals needs to go i think)
EDIT: Ah! I've realised i've had my screens saturation etc.. messed with (flux for eyes) so these renders are a bit colder than i thought..
Replies
It seems close enough to the "actual" Obiwan third light saber.
Any new thought about how you're gonna make the light sword?
Regarding the texturing, maybe you can do a very light dirt pass?
I also agree that it needs a dirt pass and some wear and tear. Feels too new right now.
I don't quite see what you mean about the normal map scalloping
For FPS weapon of this size you could go higher and your bake will appear better (the waviness that you speak of will be reduced).
Especially on the top cylinder at the end of the weapon since that is what you would mainly see from an FPS view.
Even things like the belt clip can be a lot higher (if you've noticed most FPS grenades these days have circular pins that are very smooth, not chiseled).
A rebuild shouldn't take much time to do,(a simple cylindrical extrusion) and you have the materials sorted out already.
Hope it makes sense.
http://www.polycount.com/forum/showthread.php?t=81154
It would be much better that you learn why it's being caused in the first place.
The lightsaber looks nice though. You may want to make the top flat part a little thicker.
Wat? The OP in that thread addresses cylinder normal maps in great detail. Did you not bother to even read it?
It's a bad habit and a hack because of actually trying to solve and understand the bake errors you're just covering it up, or worse, you're passing it off to other artists to deal with problems that you could care less to understand and get right.
Most (not all) baking errors occur because of human error.
It's also not a universal solution, and becomes much harder to do the more complex the asset gets, so it's good to understand how to tackle the problem in different ways. Additionally, a normal map is a complex data set of math which is not only represents details but shading compensation. The more you paint over bits, the higher likelihood you will introduce smoothing errors or other undesirable artifacts.
Funny thing about this whole thing is that one of the classes I teach actually has a lecture where I drill into the students that painting over a normal map is a bad thing. That lecture revolves around symmetrical UV seams. I find though with mechanical models like this, where the UVs are straight, manipulating them isn't so bad.
You're right about iterative approvals, though. Pain in the arse if you need to go over the model time and again. but sometimes pipelines don't allow for time enough to make a high poly.
I once had a contract that required 6 characters to be sculpted, painted, render meshes created, textures bakes, rigged and skinned and finally posed and rendered all in the span of one week.
The deadline was firm as it was for a convention and they couldn't alter that. All that and I signed the deal without having a computer on which to work. When that happens you take any and all shortcuts you can. Sometimes the results aren't pretty, and I would never recommend it for a proper pipeline. But that work needed to get done and in a hurry. A few hand painted normal maps there save me hours and hours of time.
Hamish, clean up those normal maps and dirty that sucker up.
If i try and use transparency all on one obj however, the hilt becomes see through. I have modified the albedo's alpha so i'm confused on this one x)
And also the glow on the lightsaber is thanks to bloom.
PS: I suck at taking good renders
C&C is more than welcome, i failed to point out i'm a student btw ^^