How much time do you take prepping before modeling?
polycounter lvl 11
I'm slowly starting to realize that I spend a LOT of time prepping before I tackle modeling a prop. Big or small, I spend a lot of time looking at multiple references, average or standard dimensions, the scale within my scene, trying to figure out how to tackle certain parts of a prop in my head, etc. I guess one beauty of game art is that objects don't necessarily have to BE right, so long as the look and FEEL right, yet I still find myself putting way too much effort with the pre stuff.
Is this a good thing, or a bad habit? Should I not spend as much time thinking about how I'm going to model something, and just go for it (obviosly not just randomly going for it but an educated and reasonable attempt) instead of making sure what I think in my head will work 99%? Does anyone else have this issue, if it even is an issue?
Is this a good thing, or a bad habit? Should I not spend as much time thinking about how I'm going to model something, and just go for it (obviosly not just randomly going for it but an educated and reasonable attempt) instead of making sure what I think in my head will work 99%? Does anyone else have this issue, if it even is an issue?
Replies
I usually just dive into the deep end an find references and do "prep work" as I go. Sure, getting stuff right is great and all, but getting stuff DONE is also kinda cool
just go for it man, You are definitely on a level where you should be comfortable trusting your instinct.
I'm slowly learning modeling takes a lot of time, and so does having to unwrap, texture and light a model or scene.
This is more or less how I do things as well. If I'm making a character, I usually spend half or even a full day gathering references and looking up information about what I'm going to make.
For example, I'm doing some freelance that involves a pretty realistic space suit and I needed to know how the valves, seals, and helmets are constructed. I spent a few hours looking at NASA websites and I now have a reference folder with a couple hundred images and two full space suit technical manuals. Even if I'm not going to directly use any specific real-world design for the final product, having those reference materials lets me understand the shape language and smaller technical details of how these things are constructed, and that leads to a more believable final product.
At this point, I have reference folders that range all the way from old timey Naval tattoos to images of skinned rabbits. All those references probably constitute thousands of images and dozens of hours of research that I can then apply to further projects down the road. In my personal opinion, research should be a good quarter of the work you do when making something new.
This takes me anywhere from an hour to a day.
Then I try things as I go along. Sometimes things don't work out well, so I change some more. And as for things work out well, I run into lack of fine details.
So far as the actual modelling, well during that hour or so i'll have been drawing over the concept whilst talking, and then i just jump in and start blocking out.
Blockouts goes in and out of engine frequently for testing as I go and well, and get refined over time.
usually that is true, but there are exceptions.
for example high quality hard surface objects like cars.
if you make a mistake early on it might cost you days if not weeks of work. so better put some extra steps in preparing your base and references.
For simpler props, a few images will do and jump straight in.
For more complex things, or things that have to be accurate and/or move, maybe half a day to a day of crawling through Google so I can see how the asset is put together and how it all works.
Generally I'll try and grab as much ref at the highest resolution I can, all of the exact same thing (generally you can find something and say "right, there's a lot of good ref for this specific make/model and I like it, so the asset is getting based on this") then gather a bunch of closely related refs if there are parts of the original that aren't too clear. You can kind of fudge the two together from that. As Swizzle said, there's a good chance that a lot of it will never be directly used, but it kind of helps builds up a solid idea in your mind of what the result should be.
Finding reference to me isn't so much about finding correct proportions or real world examples as it is just exploring other peoples ideas. Google is like one giant creative brain that lets me take all those outlandish ideas and mash them into a more focused asset. It's art after all, the most fun is the beginning playing with all the different legos.
Everytime I start a project and I make the concept and idea sound (or concept art look) good the project turns out so much better. I also end up completing it faster.