Sunday, October 31, 2010

Epiphany whee

Think I've finally figured out how to do the 3/4 view... it doesn't really need to be truly 3/4 so much as just partly on an angle or at least showing some degree of shape and stance other than flat on orthographic.

So I did a dig through google images, got a dragon picture with a relatively close to side view image, then built the vague outline of the skeleton over top of it, making the adjustments as necessary, which included proportional changes, how the wings are built, leg length and construction, alot of stuff like that, then redid it over top as another layer to more accurately display some pieces which need to be more accurate like how the neck, tail and waist connect.

So, now I have a sloppy, basic skeleton to build over top with updated, better looking parts.

As it's getting late, and I'm having problems concentrating, think this'll have to happen tomorrow, but this should allow relatively quick building of the main illustration, which I've been struggling with for weeks now. I figure... 2-3 hours tomorrow probably and it should be done even now that I finally have a template. I should've done this from the start, but I suppose sometimes you just only think of these things when half asleep. Odd, that.

Anyway, here's the sloppy skeleton that everything'll be built upon tomorrow.

Another sheet cleaned up



Nothing much special to say here, just some more stuff detailing key parts which move or need closeup drawings of how they work.

Quickish update for the weekend

Been finalizing some stuff, inking some older drawings for easier scan, redoing an orthographic which came out kinda so-so quality, which honestly the third try on it wasn't really much better than the first, at least proportionally, but seems better on quality otherwise... should be able to just tidy up the proportional issues in photoshop, or at least so I hope.

Tons of drawing work and such over the weekend anyways, but it's come along well so far.

Only have one totally finished on cleanup so far from photoshop, but going to put that here anyway for the time being. Should have several more done by tomorrow.

Shall likely need a little more time tomorrow, and the trip to class is a waste on mondays since it's about 2.5 hours of travel time for a mere 3 hour lab... not to mention once home again it takes a bit of time to settle back in and get things set up again to continue work, so think I'll just spend the time home as it'd be more efficient overall.

Anyways, picture time! This's a quick sample of some design schematics for the robot being made for character design 300. Going with a steampunk stylized dragon concept, so it's a pain to draw... on the plus side, from tests already done in maya, all it really requires for most of the fancy stuff like gears and such is just a few basic models to be done, and the rest duplicated off the primaries. Things like piping and such are highly modular and easy to make/place, and in fact many of these modular pieces have already been modeled, despite technically not supposed to've done so until next quarter I believe.



Honestly, I liked the sketched version of this better than the finalized cleanup version, but that's probably just because I was working on the cleanup so far zoomed in on minor details I could see the tiny imperfections with far greater ease, and zoomed out I notice them since I know where they are now.

Either way, though it has minor flaws I was not able to correct, in terms of providing basic concepts of how particular features work and go together, it's serving the purpose it was designed for just right.

The side orthographic will take up most of the rest of the night; it's pretty much done, but needs some serious cleanup work. Tomorrow should be able to see the final versions of the top and front done, since they'll be based directly off these design concepts and the side orthographic for overlaying, all the dimensions will be properly worked out in final form, which should make those two go faster than todays' work did.

Have to admit, either way, that I'm still much better at the modeling than the sketching of designs... despite having worked on this for weeks and multiple attempts at each, and design sketches ranging back since christmas, still having alot of problems, and I'm starting to think that the 3/4 view is doomed to be hideous no matter how much I attempt to correct it or redo it. Have something like 5 versions of it now or something, and each one comes out bad. Was supposed to do that first and the orthographics after, but this seriously flat out isn't working. Hopefully getting the orthos done will somehow make it possible to get a version which doesn't suck horribly, but I'm not about to hold my breath. Simon may just have to accept the fact that I know exactly how to build it, how the parts fit together, how they operate mechanically, but that the 3/4 view he may get may end up just looking not that good since it's such a complex model.

I could've done something simple, but then again, everyone else is already. I want something that I can be proud of for having taken the hard road and done it well, instead of something on par with a stickfa, which even the top students from years before essentially are. Even Andre, who is constantly mentioned as being one of the top students of the school to graduate, his own robot couldn't even do a proper walk cycle because it was flawed from basic concept, and honestly, I wasn't impressed by the design, which resembled far too much to me as looking like the stickfa with curved legs and an old camcorder for a head, made to look more futuristic.

I suppose that's to be expected though, the majority of the class, and as far as I can tell, modelers in general, do not have much innate imagination of their own, but simply convert anothers' dreams into something closer to reality. Not an artist, despite it being artwork. A strange concept really.

I'm willing to bite off perhaps a tiny bit more than I can chew at the moment, if it means something I can proudly display on the final reel and learn a great deal from. I'll make mistakes, as I've learned in several past projects, but those mistakes teach one how to do it better the next time by understanding where the flaws are now. I'd rather come across these problems while in college than when getting paid, as many as possible anyway. If it looks like something other than a stickfa with some armour glued on the sides, then all the better.

This of course seems to clash with Simon's school of thought being to do something simple well. While this can be effective at times, it also doesn't really show much else either, and severely cripples the number of learning opportunities. I have admittedly done so a few times now, when on deadlines or with a large course load to bare, but on special projects, the ones that mean the most, such as this robot design, it needs to have that extra touch of care put into it.

Practically any other design I've seen done before that they've put on the wall posters in the school, I could do without even trying. They're simple, and though they look nice, they are ultimately so simple that I don't think they're honestly going to impress anyone when it matters.

I just hope I'm not signing my own epitaph down the line somewheres right now is all. Maybe I am, maybe I'm not, I suppose there's no way to tell, only thing I can do is my best, and strive to go beyond even that. Hopefully this will eventually showcase such in the manner I imagine it shall when completed.

Anyway, shall toss the other stuff up tomorrow, still have alot of work to do. Some parts may need to be finalized in photoshop a bit later if I keep having as much trouble as I expect to have with the 3/4 view.

Time to grab a quick meal and get back to work I suppose.

Tuesday, October 19, 2010

Walkies!

Two new videos today; updated walk cycles!

It's mostly one walk cycle with two versions; in version 1, the walk is normal, heading towards the camera. In the second version, it's a silhouetted form, to give better idea of the shape and animation without being distracted by the model itself which can draw attention away from the specific motions. After review, I've noticed several very slight things which kind of bother me, but kind of look natural too so I'm on the edge as to whether to bother changing them or not.



Thursday, October 14, 2010

The eyes! THE EYES! =O




A quick 24 second video (well 12 twice, two camera angles) of the upgraded eye model. This is manually animated as no rig controls have been added as of yet.

This shows the full range of motion possible for the model.

Also, it's hard to tell with this low res video since it's been compressed twice, once by aftereffects, and once by youtube, but there's small tiny gears inside of the eye which are rotating to turn the gyroscope the lens is built upon.

Overall quite pleased with the design!

The second eye will be slightly different, thinking of a telescope like assembly, which should be much easier to model and look appropriate as well.

Oh and there's some flickering of the shading on it; this's a compression error, just ignore it for now.

Thursday, October 7, 2010

Beginnings of... A MONSTER :O

So we need to design a robot for class... no big deal. I've been working on the basic concept off and on since Christmas now, so know what I want to do with a pretty good amount of certainty at this point.

So far, have the eye design pretty much modeled, just need to correct one error involving the rotational axis of several sections being incorrect, to which I'm waiting on a response from Matt with an explanation on how to do that since it's not working the way I expected.

Anyways! Here's a quick video of the iris design in action, followed by a semi transparent render to show the completed housing assembly.






The way this's set up, the iris blades can close in a circular pattern, similar to a camera iris, or the iris of the stargate in SG-1. Same concept and looks nice, while leaving a gap to allow the inner 'pupil' to see.

The iris is mounted on a gyroscope which allows it to rotate vertically and horizontally, with small gears to turn the gyros.

Furthermore, the 'pupil', is a mechanical construct of 4 cameras; infrared in the center, for primary vision capabilities, with smaller red/green/blue cameras around the central section which provide secondary inputs so that things which are cold or heat shielded don't go invisible.

The pupil can also rotate slightly in relation to the iris assembly, to give the impression of an eye looking around.

Behind the pupil and iris assembly, is a reflective surface which allows light to be reflected back, vaguely similar to a cats' eye, for adequate night vision.

It's pretty complex overall, but the rigging honestly shouldn't be that bad, and it'll be wonderful for a nice closeup shot.