"Speed asset," unknown human skull, 3h

Vrav

Summary: a rather trash speed asset done to measure time versus quality. I have not completed much game art lately and felt the need to make something. This way, in a year, I can do the same and gague improvement or something. Negative qualities I note:

- sculpt was undesirable, very underdetailed at one hour
- the detailing stage I skipped would bring physical asymmetry
- not a fan of photosourcing colourmaps, but did here, via cgtextures
- uv layout far from optimal: empty space could be filled with more bones
- the lowpoly... stupidly dense, haphazard topology, imbalanced texel density
- presentation would be cooler with an actual, simulated physics pile (tweaked)
- needs giant rubies with custom shaders in the eyes or something
- more reference was required to accurately render cranium seams

I may now return to the sculpting stage and produce a less hasty, hopefully superior twin asset. I think it could be interesting to compare the results. For example, I did no painting in Photoshop for this; it was all photo-projected using an application still in beta, mixed with some lumpy AO. Anything I can think of...:

What do you make of the time taken versus the final outcome?
Could this (with reduced tricount) work in a commercial game?
Do you agree with me that the spec map is worthless here?
Should I bother taking more screenshots?
Why am I experiencing deja vu?

Do you ever feel the urge to do a similar "speed" asset, or do you get enough of that in crunch time? Does this shotgunning practice hold much value, even if my personal results are not the prodigal portfolio ideal?

I think three hours is a bit long for this tiny thing. It's just so lumpy/blurry, more like 20 minutes would be acceptable... or 20 more minutes on the sculpt would pump it up a thousand percent. But what's cool is that can be done and the normals/ao can be updated with little issue, and if I optimize the lowpoly, everything can be baked from the various highpoly sculpts again...

OrganizedChaos

Hm. I'm not sure how I'd feel about taking an asset from start to finish as fast as possible. It kind of gives you the result you have now- you're not happy, most other people won't be happy. I guess it's practice, but I like how game-artist.net did it, in stages.
You could do speed modeling
http://www.game-artist.net/forums/game-artist-mini-competition/12983-speed-modeling-competition-2-steampunk-blunderbuss.html
or speed texturing
http://www.game-artist.net/forums/game-artist-mini-competition/13205-speed-texturing-competition-2-001-steampunk-blunderbuss.html

Just a suggestion if practice is what you're going for
(edit- hm, it used to be an hour limit for the model, looks like they changed it up a bit)

Vrav

Well, I am not particularly unhappy with the results. There are simply things it lacks for being rushed. The only thing I am actually dissatisfied with (ie, can't be fixed) is the time compared to the result, and that can only be improve with experience.

I have been thinking about it this morning, and if I can salvage or fix this up with 60-90 minutes more work, I think it could prove an interesting workflow... or study. I don't know, I will return to the sculpt and see what can be done.

Those are good challenges, totally forgot about those forums.