Challenge #3 Topic Discussion

My idea is this:

- 10 000 Polygon Limit
- A beast and a character (rider, handler, slavemaster etc :))

- texturing, probably need to do 2x 1024 maps

and instead of 3 weeks, 4/5 weeks.

What do you think?

Cool Pants and I have already done that, we'll just post our old stuff hey [;)]

How about something really low poly, a few 600 poly characters or something, 256 textures etc.

um, inglis~ what sort of game is that ment for? thought the point of game art was a) creative use of minimal resources and b) getting it done by a short deadline (never worked for a company where i have had 4/5 weeks to work on 1 art resource. usually more like 4to 5 days....)

my sugestion;-

- 2000poly movie star
- 1024x512 texture space
- 3 weeks

DSC

sure it shouldnt take 4/5 weeks to get it done. the large timeline is to allow people to jump in if they found out late about the challenge.
just look at this challenge for example, 3 weeks- ample time and over half of the entries arent completed.

it could be 5/6k...just throwing out a number.
that figure just came up in the irc channel.

Movie star! that's quite a cool idea [:)]

I bags the Porn star!!

damn! i was joking around saying Dirk Digler lol

how about to make it a bit more interesting 3D Caricatures of movie stars, just to give a bit more artistic freedom.

Good idea inglis, maybe film or music stars? If anyone does Ozzy Ozborne though I'll throttle them. lol

Since there is a prize at stake here, maybe a Movie Star and something that represents them as a character, or something they are known for?

e.g Dame Edna, sitting on top of huge pink funky glasses..

just an idea...

How about something other than a character, a vehical or structure perhaps?

I havent done either yet and i find the challenges are great to keep me motivated enough to finish.

Ooo, Dan has a point. Vehical and structure work aren't exactly my strong points, as I've hardly done them. And level design and creation is an important aspect in gaming.

I'm all for the movie character, but this'd be really good.

How about ancient warrior?

Structure/Vehicle is a good idea :)
Im up for that.

I like the caricature idea... it would give people a chance to venture out and try something new. As long as everybody makes Bruce Campbells... [:D]

Scott.

Or Miss Geosphere 2003 Pamela Anderson. [;)]

Pantmonger

We should stick to low poly stuff regardless i think, most of us here are looking at the games industry as a starting point of our modelling career and these challenges help flesh out those show reels :)

Anyone have an idea what the poly count for a ingame structure in one of the current game engines is?

Perhaps make the challenge be to build a structure, vehical and unit for a strategy game like C&C Generals.

Wait for a suggestion from Rod Green.. he's the one who organised the prize for the next challenge, so I'm happy for him to suggest what he wants for the challenge.

Voting for the first challenge was criticised for taking too long the votes/comments to appear from the judges.. and obviously, the suggested user votes that's happening now wasn't the good idea it's turned out to be.. so for future challenges, I'm not going to be chasing up judges for votes (judging and commentary takes time for industry people, and I can't be annoying the same people over and over whenever votes are needed).. so judging will be voluntary from anyone in the industry, as long as they are in an artistic field. They can email me their commentary and votes. If I don't get at least 5 votes, I open up the voting to the public wth a voting form.

Since there's an Infogrames Melbourne House prize for the next round, I suggest Rod and a few of his art workmates judge the next one. [;)]

I don't think we should kill the user votes.

(A)I just think we need to make sure the rules are clear on the vote thread and that people adhere to them.

(B)Don't change it just because I pulled people up on the rules, its not my place to do that its Souri's or in this case Rod's also. Won't happen again, just after the first challenge it got my blood boiling.

how about a collaboration? i.e have teams of 2 modellers working on a set concept. or maybe have a set concept for a game and everyone makes an asset for that game concept.

Entries being judged on how well they suit the game concept and how well they have been executed (mesh/uvmap/texture)

The closer we make these challenges to the sort of things you would be asked to do in a modeller position the better it is for all of us hopeing to break into this industry.

I don't think there would be enough entrants for collaborations just yet. Teams of two would cut the number of models seen in the challenge by half, so we could be seeing only 3 or so entries. If we had hundreds of people participating, teams would be ideal. [:)]

Why not a real High detail model like 100 000 polys, I'm suggesting this because Normal mapping is coming in now and having highpoly skills are going to be come rather essential in the future. How many game artists have that skill now? Its certainly worth learning.

Personally I feel we should stick to the low polys...and its a fact that, the better you become the fewer polys you need...so go hard and go low! Also I would like to see this new challenge be something other than a character...we all love to modell characters (my showreel is 80% full of them) but when it comes to finding a job, simply sructures and vehicles are a strong point.A good friend of mine just landed a job, and his showreel consisted mainly of vehicles,(jeep,plane)and real simple stuctures like a security fence,rock face,tin shack...nice and neat,great texturing.
I need more of these on my show reel!...how about you guys????

Thats what i was thinking Timbrex, i definately need more experience with structures/vehicals (havent actually done any) and low poly work is more technically challenging than high poly in this setting.

I said i was up for a structure challenge before- and i still think its a good idea.
i would like to work on a fortress like structure such as those in Tribes2. maybe a little higher poly count than those as the outside of most of them were quite flat. (from memory)

Same here, all I ever do is characters so something not character related would be neat.

Poly and texture limit at the moment seems realistic and they're nice numbers to play around in, but good point about normal mapping. Although under the current guidelines I guess you could create the model using normal maps, as long as the final map was no larger than 1024 and the model no greater than 3000 polys.

The caricature idea is a good one as well. Definantly good for something different on the demo reel.

Stef K

If going the vehicle route could have something based around a techno soldier model from challenge 2, design a vehicle to accompy them.

Of course all models from challenge 2 are copyright by respective authors, but generally its still possible to design

I like the caricature idea alot but it may be a good idea to do something a little different than character modelling just to show some range.
Not to keen on vehicles, a level would be good.

How about a fortress like Inglis said; present, medieval, alien or sci fi tech.

Fluffy/Badvibes: I'm pretty sure you can paint your own mormal maps like you do a bump map. ID uses external models in Doom 3 but I don't think its the only way of doing it. Correct me if I'm wrong. [:P]

Saying that I actually wouldn't mind doing a high poly model, its been a while, Inglis would win though but hey it'd be fun still. lol

I'm side tracking things a bit but Crytek have a trial for there normal mapping 'Polybump' plugin for download at...

[url]http://www.crytek.com/polybump/index.php?lang=eng&sx=polybump[/url]

I'm pretty sure this is the same tecnique they use in Doom III. I haven't used the plugin but the included viewer and examples are cool.

Looking at the source for the normal maps, you could paint one by hand but it'd be a real brain bender[:)]

Crytek's got some great stuff, have you seen the demo for their new title, mmmm tasty. Totally realtime editors with realtime deformation of everything.

I may be totally missing the point of how this normal mapping is working - but wouldn't you get a similar result if you just bake a texture map of a high res model with nice textures and lighting down into a single color map (and maybe a bump map if the engine your modelling for uses it)and apply it to a low poly version of your model?