Can I salvage this piece for my Portfolio?

Soul_of_Solace

Hello again,

So, I took an art test recently that I didn't pass (unknown as to why, I mean, obviously it was not enough, but not enough what?). But I spent a lot of time on it so I'd hate to just scrap it in the bin, and I want to get peoples opinions if this is something that I can use in my portfolio as is, or if I polish it up a bit (and of course, ideas on what needs to be polished are most welcome).

Some background on the limitations i was working under. 2500 polygon limit, 1024x1024 texture limit (broken down into a few sheets). Used Zbrush for the Gears and Door, tiling textures for the remainder of the structure.

[vv]48844981[/vv]




So yeah, it'd be nice to save this, but would want to make sure I only save it if it would benefit things.

reverendK

if you realy want to save it i think it'd require a lot of work - maybe more than it's worth.
namely - i would get rid of the terrain and just put it on some sort of pedestal. probably get rid of the sky as it is as well...unless you want to model more of a scene around it.
less tiling textures, or at least a greater variation of them - more edge work on the structure (stones round the arch, corner stones and not just white bricks everywhere) - more detail in the mesh and texture of the gears - more functional details (belts, chains, more support structures on the gear at the top) It could be really cool with a lot of love and thought (and maybe some sound in the animation - clanking and creaking and whatnot) - but may practically have to be redone in a lot of ways.

BARDLER

I think you need to rework your textures a lot. The bricks are very clearly tiling and don't make much sense at all, and the cogs look rather boring with that simple texture on it. To me the castle like thing has a clear fantasy vibe to it so matching the textures to that would have been a good idea.

linkov

I'd say let it go. I don't mean like delete it or something.. Put it into your memory lane folder, just to remember where you've come from and start something new

Soul_of_Solace

Yeah, with the limited texture space I was allotted I was annoyed I ran out of room for a secondary texture that I could have vertex blended in UDK to break up the repetition. All three variants of the gears are sharing the space of a 512 texture, which was obviously not ideal for such big pieces. Making them smaller and having more of them would have run me over-budget in polygons though. Same with trying to smooth up the geometry more.

You're right, it'd take a lot more work to fix it unless I can quickly get some vertex blending in to fix the repetition and create a new texture for the Gears. Worth a shot I guess.

linkov

well, you probably could have saved a lot of tris by not making all these bevels and then use them to make cuts in cogs and texture them with smaller tiling textures.

I'm curious what was the brief? Apart from tech limits? Its ok if you can not tell.

Soul_of_Solace

I'm not sure if I can say or not, but I suppose I can give a general brief and say that the piece was supposed to be an entry structure and door to an area/compelling scene with visual history to it.

The Gears were a decision I struggled with since I wanted to use Zbrush to make them, but in retrospect, that probably was not a sound strategy with only a 512 texture available for three variations. Doing what you said and adding in more geometry to the gears like cracks or breaks in the geo, and utilizing some tiling textures would have been better. Although I don't think the bevels on the supports were adding that much geo, and I was running out of polys very fast.

linkov

I see. And no, I wasn't talking about cracks or breaks. I was talking more about structure. Easily distinguished parts what can be mapped with simle textures. Cogs of that size most likely cannot be carved out from a single piece, be it stone or metal.

Anyway, I guess the biggest problem here is the overall design. That structure doesn't really have any sense of purpose. Well, at least these cogs don't have it. Its like here is that random castle and then some giant cogs around it. Its also doubtful that cog on top can be held by these thin beams.

Soul_of_Solace

linkov wrote: »

I see. And no, I wasn't talking about cracks or breaks. I was talking more about structure. Easily distinguished parts what can be mapped with simle textures. Cogs of that size most likely cannot be carved out from a single piece, be it stone or metal.

Do you have an example of that? Because I'm still not quite understanding what you're getting at since you did first state to make cuts in the cog and I took that to mean a gash through the piece (not a texture or surface cut) and a break would be one of the gears "teeth" removed or disfigured via geo.

EDIT: Got Vertex painting to work, although I of course ran into the problem with very few verts to really do a good paint. I also tightened up the Brick texture so the obvious depressions that made it so obviously tiling are no longer quite so obvious.

linkov

I meant something like this

Mrandk

I honestly don't think you should try to use this as a portfolio piece. Make something new and save this in a folder.

As Linkov stated, the structure doesn't seem to be a functional piece. (I mean yes it is a doorway so that much is functional but there are other aspects that are not.)

The gears appear to be nothing more than giant decorations that move, What is causing them to move? What do they do? Were they intended to open the door? Some of the gears appear to be held up by nothing and merely jammed into the side of the gateway. Some massive wooden beams could have worked as support and they wouldn't have taken up too much texture space/ tris as they are only glorified cubes. The things supporting your central gear on the top are indeed too thin to support such a massive structure. The two pillars on each side of the gear on top appear to be made of what looks to be cobblestone. I have serious doubts of the structural integrity of them.

Aside from the design aspects of this piece there are major technical issues with it as well.

You have a huge difference in texel density between your brick and the gears themselves. Yes that does come from the unique sheet that the gears are sharing but knowing the limitations you were given, making them as large as they are was a mistake. The gear you kept small looks nice and has close to the same resolution as the areas around it.

It feels as though trying to keep the gears in the piece robbed you of texture space and polycount for the sake of having gears.

The piece could use more color variation as a whole, everything has a very green/yellow feel to it. That may have to do with your lighting but I cannot tell for sure. The strong lighting is showing off the hard edges of the lowpoly in a bad way. The terrain surrounding the structure is doing you no favors by being so flat and barren.

In the end the most important part of a portfolio piece is to make it look good, limitations with in reason are fine but when the limitations restrict you so much as these did it doesn't make for good portfolio work.

I suggest if you truly are intent on this gate being a portfolio piece, start from the ground up starting with the design thinking of function and purpose first. Remove the insane restrictions. Focus on making it look bad ass and then worry about how many tri's/ textures/ whatever later.

If you are not intent on this being a portfolio piece make it easier on yourself and work off an already existing concept. There are plenty to choose from at this site: http://www.conceptroot.com/ult/

I hope I was not too harsh and I apologize if I came off as such. Hope this helps!

Soul_of_Solace

I wasn't worried about the design aspects of "this can't exist in real life" since the examples I was given were of object entrances that also couldn't really exist and had "physics" issues, since this is a fantasy piece. That said I did originally hope to add supports, but of course, ran out of polycount (I actually did hit the 2500 barrier on point). But yeah, yours and linkov's points were pretty much spot on regarding the tech issues. Too many for a quick fix, so I will just let this one go and think of it as a learning experiment.