Critique => 2D & 3D => Topic started by: Grimsane on May 15, 2012, 05:14:06 pm

Title: (semi)low poly modelling experiments (undeadly)
Post by: Grimsane on May 15, 2012, 05:14:06 pm
448 Quads 512x512 diffuse 4-5 hours modelling unwrapping and texturing, no ref, just further testing my freehand anatomy and painting,
tattoos where an afterthought just experimented with creating a simple yet intricate pattern. colour tests from undead pallid to lively peach
feedback would be much appreciated, and feel free to comment i want to know if it is aesthetically pleasing or not.
kinda reminded me of shadow man with les gads, no?
Title: Re: low & semi-low poly modelling experiments
Post by: Grimsane on May 20, 2012, 05:15:52 pm
figured I'd avoid spamming the forums with threads and just put this here.
Robot experiment, using Vertex painting, and a small texture plane for the eye region.
Title: Re: low spec experiments
Post by: Grimsane on May 23, 2012, 10:59:40 pm
just messing around with an idea, (yeah the textures are rushed, and tile rather horribly)

have done something similar in the past, which fortunetly i had online (that PC i did alot of art and game design experiments on died and so did my backup..)
i only had basic grid based movement, openable chests, and 8 directional camera rotation, was just messing around.
(and yeah when i was doing it i noticed the character style could easily represent minecraft so i put a reference in, i didn't have any dialog system in, but i intended on having the guards complain about finding him outside the outskirts of the city digging up the terrain and wildly punching trees. as an in-joke)

with the new one i am messing with now, i plan to try out sprites instead of low poly square dudes, kinda like breath of fire 3 or something, but I'm already fearful coding camera dependent view angles etc will be a bit complex, that and I'll need some sprites, but i do have some I've already done in mind, anyway just felt like dumping it here, see if anyone is interested
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: Etchebeur on May 25, 2012, 08:23:21 pm
Sorry for the stupid newcomer question but, what is the program you use to model 3D?
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: Grimsane on May 26, 2012, 01:06:49 am
Sorry for the stupid newcomer question but, what is the program you use to model 3D?
Blender, It's very powerful, and doesn't cost a cent :) the interface has gotten pretty user friendly recently too.
It might be worth trawling the online documentation or watching some video tutorials, I'm not an overly tutorial dependent type person but i know alot of people like them and find them immensely helpful and the community for blender is considerably large, and helpful.
it also has its own python based, Node driven Game engine which isn't too bad, especially for prototyping, no need for code for simple things, great for interactive turntables etc.

http://www.blender.org/download/get-blender/ (http://www.blender.org/download/get-blender/)
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: Etchebeur on May 26, 2012, 03:53:51 am
Oh, ty, Im interested enough to try something now, the program is very expansive but i know with time get used to it.   ;D
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: Grimsane on May 26, 2012, 11:47:54 am
good luck  ;D it definitely takes persistence, I've only been modelling 2 years on and off (and pursuing many other creative endeavors simultaneously), I also do high res stuff, but would probably be a fallacy to post that here. :blind:

in theory if you (or anyone else) concentrated and practice almost exclusively in it you'd surpass me in a fraction of the time :D
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: BladeJunker on June 07, 2012, 06:49:53 pm
Well I like the torso as its very elegant in design and the texture colors work well. My only CC is that it seems like a little too much polygon use put towards overall prism smoothing than there needs to be. Best to take this advice with a grain of salt since the resolution range your model is in is tricky one, its that post Quake3/Unreal 2 quality range so its hard to split hairs. My only reference would be Metal Gear Solid 2&3, although your design is very smooth the poly count would rise quickly with more ornate costumes based on this mesh.
The outlining form and or joint prisms seem good as full motion range should deform the mesh well, shoulder looks good but the hip joint might give some trouble.

Cool robot, reminds me of Tobal with its vertex painted surfaces, minimal gouraud shading and textures. The model is a bit too ornate when you consider the low spec texturing, in particular the shoulder indention. Overall I'd consider a little more texture mapping to optimize the mesh more. I always thought the mech based games on PS1 and arcade titles like Virtual On found a good sweet spot in model aesthetics.

I really like your cube world and although the tiling is obvious in the textures I do love the bolder colors despite the rudimentary design, nothing wrong with conventions since we take comfort in them.
I like the square dudes as they fit the setting well but if you're going with sprites like BOF3&4 you should add more soft slopes to your terrain as in ramps or reticulating splines plus more ornate mesh tiles to close the aesthetic gap between the smooth sprites and the angular polygonal world mesh.
Lastly I'd go with a softening or blurring of texture along the edges of your polygons to reduce the look of angularity even with meshes that are cubic by nature.

Well I see you dabble in a bit of everything much like myself, I like your 3D stuff and look forward to where you go with it. :)
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: Grimsane on June 12, 2012, 03:32:26 am
only just noticed this D: browsers been exclusively in pixelart for awhile..

first up thanks for taking the time to give feedback it's appreciated :) and agree with all your points, it was rather speedily done so being described as elegantly designed is a great indication I'm going in the right direction and yeah for instance he has 10 poly loop on the arms, should reduce to atleast half that for optimization sake unless I go in the other direction and go high poly, which I won't... because I much prefer PS2 kinda spec models probably set a 5k cap or there abouts including minor costume and props.

still early days though Haven't been modeling too long, should probably dump some of the better stuff I do, and also that's the reason I stated semi low spec in the topic, really low spec art like DS style 3D is something I'm yet to attempt successfully, although I have dabbled in extremely low poly. I'm itching to animate my 3D creations though.

and yeah the robot isn't very optimized and I decided i was going to use loops to do some white paint lines, vertex painting is rather interesting method (I like how ZBrush use quadrilateral Vertex painting and make it out to be something original by calling it PolyPaint™) I might eventually unwrap the UVs and bake it down onto a low res texture and optimize it. I keep jumping from model to model, trying to learn as much as I can from each, I'm essentially teaching myself. recently started getting game models and analyzing some of what they do. high density models are quite intimidating this early on. but I've seen a few videos recently too and I'm surprised how primitive they start, most modern "next gen"/high detail models are simple base meshes, sculpted to an immense level of detail, then simplified and retopo'd but of course work-flows vary from artist to artist.

and indeed dabble in anything and everything that interests me. Yay for dabbling ;D

and yeah I've been contemplating doing some spline style sloped terrain, I'm considering breaking it down though to something efficiently loadable like a quad distributed heightmap or something, but model data isn't that large for low spec, and I'm not coding for any form of limited hardware so maybe I shouldn't, but I will experiment, might prove to be faster for asset creation. (I'm doing a degree in game art and design next year and hope to make some cool demos and hopefully network with some awesome programmers, although if any programmers are reading this and interested in small scale artistically driven game demos PM me I always love to do personal projects whenever I can and my programming is extremely limited so I rarely break past the Art phase)

OffTopic-ish: speaking of Metal Gear Solids, there is something Japanese models. personally they are like the pinnacle of where I'd love to get to skillwise, they have a sense of elegance and precision and something not quite the same in most western games, also I've seen quite alot of western games that have high quality and high detail textures, but poorly formed meshes. in motion sometimes it's not so but nigh unnoticeable, and it's mostly only evident in the earlier era of 3D games. but just something about Japanese (and even Korean games are quite often as nice) proportions and aesthetic ideals.. maybe it's as simple as longer legs :lol: but there definitely is a defining factor of characters in particular that have such a wide appeal, for me atleast.

wow that got long ::)
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: BladeJunker on June 12, 2012, 09:00:21 pm
I'm glad we were on the same page as far as PS2 level quality since I wasn't sure if you were going higher or lower. :)

Yeah I quite enjoy the art of low poly counts like those of the DS or early 3D meshes, that balance between visual appeal and the rendering thoroughput has always been a fun challenge. I was such a clunky worker my friend always did the rigging which I really should have tried to learn, I guess I still can as Blender always resides on my HD.

Strangely enough I became interested in high poly way before it was even practical for systems which lead me to an LOD obsession sometime around when Team Fortess2 was first announced and Shiny had made Messiah. Even though I learned a lot about optimization of meshes I really missed out on actually making games because of it, my friend and I were making the next Deus Ex but I kept looking to the horizon and not on where I was or what I was doing. He has a good job and just got married so he doesn't have time for my flights of fancy these days. :-[
Yeah Zbrush what a powerful tool, man its easier than ever to model 3D now, I hope to some day get my hands on that. That certainly is the key to high detail work is building upwards from simplicity, believe me I've tried many times to jump the ship and model directly in raw polygonal high detail and it never goes well.
PolyPaint™, what a crock. :yell:

Heightmaps are good, I like them but my art skills are weak for complex organic textures so I kept doing meshes by hand. In open world games they are a godsend as modeling mountain after mountain is wholly impractical. I think you'll collaborate with someone eventually as going for a game art and design degree is the ideal place to meet like minded people for game projects both professional & personal.

Yes the Japanese have a keen interest in beauty regardless of medium as they make the slickest comics, games, and animation. Sometimes that pursuit could get them into trouble with early 3D as games from there could be prone to more popup and clipping than a western developer would do for the sake of smoother graphics. That's what impresses so much about Konami and MGS as they found a great sweet spot between mesh resolution, level scale, and degree of world simulation.
Oh I know, the whole world is getting into game development including China & India too which are making good products too, actually a lot of game companies and animation studios hire from all around the world so you get some pretty awesome talents brought together. ;)

Lol long legs yep that is an art ideal globally I think as the statuesque form pops up constantly regardless of nation. I've created to those standards often but lately I've been trying to get range, gradient, or contrast in body types even within tiny sprites as I'm tired of the homogenization you get with everybody being built to perfection. :lol:
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: Grimsane on June 25, 2012, 08:20:14 am
"making the next Deus Ex"
sounds like an ambitious project, up my lane of interest too, I'm quite fond of emergent game systems, cause and effect and utter player freedom toward actions and exploration ;D create worlds for people to become immersed in and explore etc etc.

currently working on an assignment for my animation course and have thus far speed modelled this character concept
rather low poly and hasn't been optimized yet, still within PS2 character polycounts.

and yeah game development has been taken up all over the globe, sadly some of the jobs in india and china directly correspond to the closing of alot of companies all over the place, Australia took quite alot of hits in the last 5 or so years, especially the state I am in there are barely any active studios at all, besides casual games, and they aren't the same :(

2500 poly or so , still WiPy, on texturing will rig him up soon, spent about 5-6 hours on him so far relatively fast modelling spent couple hours modelling the head and another couple on the body, lil bit on unwrapping and dunno few hours on the texturing (keep getting distracted so it's hard to gauge, but it's faster than alot of stuff I've attempted so far in some regards)

was told it was too colourful (borderline racist if you ask me  :lol:) so had to dull it down, might need to push white make up, it's for a mime thingy, so got a casual looking mime kid going, and oh yeah might end up in sepia, but not sure

seems colour is more likely, and update:
some rather rough story board slides, with an edited version for the boards, with nod to a certain someone on the forums pseudo-trademark I'm sure it's obvious :lol:

oh yeah he's rigged now if it wasn't obvious, back to texturing tomorrow
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: PixelPiledriver on July 02, 2012, 01:14:36 pm

Your boards are at a very early stage and could change significantly, but thought I would mess with the idea a little:

Some Advice:
1. Interaction: have the characters touch each other, affect the others mood, or force the other to do something.
2. Moments: Make a panel for anything significant. Motion, rest, action, start (I forgot to do a title), end, etc. This helps create Timing.
3. Facial expressions: At this stage just paint them directly onto the image. Even if they are very simple they will help you later.
4. Arrows: On a still image sometimes a static pose can look like motion, or motion can look like a static pose. Use arrows to explicitly mark motion. This also helps create Timing.

Playing on the fact that he's a mime could also make it more interesting.
Cool Stuff!
Title: Re: (semi)low poly modelling experiments (undeadly)
Post by: Grimsane on July 02, 2012, 02:16:17 pm
wow nice work thanks  :D
that was like a fast partial slide test while testing out my freshly rigged character,
and yeah I actually had some interaction planned he does the same but with his hand, your adaptation is well within the brief too, and initially I avoided doing the step in it solely because another class mate was doing it, but I think he has since abandon it, so I may well pursue it and I think it'll be easier to fit within the 30 second time frame than what I was going to do, I think I'm liking the dynamics of this option more too, especially now you've helped me visualize it more and I like the idea of being knocked of his feet ;D going to try making some blend shapes for facial animation too

thanks much for the input :) great use of action lines!, and great points I should get in the habit of using them no matter how early, especially when showing them to others ( have learnt that too)

titles are independent of the 30 second time frame so I can go crazy  :D but I can't ruin the punch line ::) got a few ideas worth pursuing, I'll be sure to keep this thread updated with my progress, thanks again ^ ^