How To Get Faster... Faster?
polycounter lvl 14
I'm not really sure where to put this question, but this forum seemed most appropriate.
I just want to know, how I can improve modeling speed/efficiency to improve the quality of my work and my work speed.
What kinds of things do you pros do that I could benefit from doing? what workflows, practices, etc.?
Any help (especially a point in the right direction) would be great
I just want to know, how I can improve modeling speed/efficiency to improve the quality of my work and my work speed.
What kinds of things do you pros do that I could benefit from doing? what workflows, practices, etc.?
Any help (especially a point in the right direction) would be great
Replies
I've yet to do it myself, but I'm told that recording yourself and watching it later does wonders for picking out bad habits or things you could improve on.
Customize your Left-Handed Hotkeys. What tools or functions do you use most often? Are there little things you do often that could be set to a macro, action or script?
At one point, I had minimized my Photoshop brush set down with the essentials at the top, no need for 7 size presets for a common brush when 3 would do. Common grunge and spatters at the bottom. Any other special brushes I stuffed into a second library for when I need them. This feels quicker for getting to a brush, less to sort or visually remember locations of in the list.
speed comes with experiance. You can't just jump over this.
The studios that care about time usually often go for "good enough". Most studios aren't Blizzard where it's done when it's done. Quality is like a bell curve - at some point the amount of invested time only yields small improvements. Learn when to stop. The exact time where to stop obviously is different from studio to studio or AD to AD.
Ask yourself if, for examples, skin pores need to be done in ZBrush, or if it would be faster to do them as normal map. Think about the quality difference in the final asset and if it is acceptable for the standard you're trying to achieve. Can you mirror and re-use mesh parts / textures? Do you have a "parts" library where you can grab buttons, buckles, base-meshes for shoes, swords, etc?
This is of course also transferable to other modeling tasks.
Plan your work ahead. Have a clear picture in mind which parts of a character/environment have to be done first, to what parts you move on to next. Spend most time on obvious features - don't get lost in details and waste time on features people will not notice(*). Try to avoid backtracking and re-visiting parts that have been completed. A back-and-forth between different work steps and asset parts can cost time. Cut details if you run our of time, but never skip on polish!
(*) not talking about details, i.e. features of the character here, like buttons or zippers. I'm talking about wasting time fixing 1 pixel seam that's on an area of the model which the average player will never even see or notice unless someone tells them about it.
http://poopinmymouth.com/tutorial/analytical_thinking.htm
some practice:
http://artisaverb.info/Selection.html
just apply this logic everywhere all the time and you'll get times faster.
when i was learning to play the guitar, i focused almost entirely on heavy metal and lead guitar techniques. i found i'd reached a plateau... no matter how often or how long i practiced i just wasn't getting any better. so i started learning to play classical guitar, and within a couple of weeks my lead playing was moving on in leaps and bounds again.
this could also be applied to any other form of learning, i'm sure. learn something related, but different... it's still using the same core skills, but in a new way.
Scripts that reduce a 50-click operation to a single button press.
Learn selection tricks. From simple stuff like "grow selection" to more advanced "select every nth edge". Use paint selection instead of clicking each element. Use negative selections as well as positive. Learn (or make) hotkeys that convert selections, eg. convert vertex selection to contained faces.
Years of practice ;P
The biggest trick to getting things done fast is to know where to focus your effort and add detail, and where not to. When your on a deadline, you learn that you just have to let those things go and focus on the details that are most prevalent and important.
Also, specifically when dealing with environments...block your stuff out right away...figure out your scale, sizing, modular piece fitment...as early as you can. Spend time to make the blockout right...because its a lot easier to fix then than to deal with later with near finished assets.
turn off all windows and programs sounds
no wallpaper, and deactivate all those fancy windows features so it looks like windows 3.1 again so you can focus all processing to main working programs.
practice practice practice
learn all the hotkeys
Keep a stopwatch at your desk, and learn to use it!
Every hour on the hour, maybe do some push-ups or something that makes you GET-UP and REALIZE and hour of time just passed. Just after a few of these you realize when you had a lazy hour and step up for the next hour.
fast, rocking music
caffiene
EAT WELL! Biologically, when hungry, the body thinks it's dying and starts refocusing resources away from creative productivity. Grab a snack to activate your brain.
After that go practice art in your new cocaine/speed driven state.
Take a 15 break every 2 hours (or when you are "stuck") to exercise and snack.
Adij: I try to never sacrifice quality for speed, but when it comes to getting stuff done for a class project, its hurting me that I can't have both when I really need to.
Kwramm and gsokol: Hadn't thought of progress like a bell curve before, I'd only thought of it linearly--I'd thought that the more work I put in the better it would be, even with minor details. Also, hadn't thought of putting together an asset library. I have one for textures, but I see an asset one would be really useful as well.
d1ver: great stuff, I'm definitely gonna start applying those concepts today (for both work and life)
almighty_gir: So if I apply that to Game Art, does that mean work on a character every now and then if I usually do environments and props?
cholden: I actually started doing that just last night, doing pushups after working for a long time. (more as a get back in shape thing than a refocus thing tho). And typically when I work, the only other program I have open is iTunes, playing either game soundtracks or something Hans Zimmer related. ....Not really sure what you mean by the stopwatch however...
seforin: I just laughed so hard I think I scared my neighbors.
I know that practice makes (almost) perfect, but my issue has been that my the stuff I try to practice on tends to take a lot longer than I'd like, and I realize now (thanks to you guys) a lot of it has to do with repeating mistakes, not planning out well enough, and/or inefficient workflows.
so imagine you've modeled, sculpted, baked, and textured everything, hand it over to the lighting guys, who then tell you "the spec is too high", or "the diffuse just doesn't read well in the light"... all the changes you then have to make, right?
but if you have some lighting knowledge, you can anticipate that and incorperate it into your workflow for building those assets... which then saves you time in future, and builds your working knowledge of materials and their reaction to lights.
anyway, i'm semi-rambling. all i know is that it works for a lot of things i've learned, so i can only assume it would work on this too?
Makes sense, I'll definitely have to give it a try. Thanks again
... is it free?
Not only did I get in shape, but now I work at 125%.
Yep.
http://www.manictime.com/download/
So my only advice to you, assuming you've never been employed yet, is to just focus on quality. Make the best quality environment/character/prop/whatever that you can. As some have mentioned, there is often a trade-off between speed and level of quality
Three keys to getting faster, faster - from my own experience:
Optimization (workflow)
Variation (differing content/style/reference)
Experimentation (new ways with new techniques, often times this happens as a result of variation)
Optimization is your workflow, shortcut keys, hotkey scripts, and having done things enough times to know the most efficient way of going about them (what will work and what won't).
Variation is trying new things. Characters, animals, environments, props. Drawing, painting, sculpting, reading - taking in knowledge from having done anything possibly related to your craft, trying to become "well versed".
Experimentation is fucking around. There are always multiple ways to achieve an outcome, some work well only in certain situations, there'll be some balance of pros and cons depending. As mentioned I'd say this sort of ties into variation, maybe trying to do the same essential idea in a few different styles or with different restrictions.
I'm not a 3D artist, but it's really all the same I think. Where I might experiment with doing the same image starting from black and white, color, or only with one brush - you might do it with a different poly budget or texture restrictions. Or you might even do some painting, and I might try to learn some 3D stuff. I find that I've improved dramatically just looking at the stuff on polycount - even though most of it isn't my particular "craft", it's all related.
That's just what I've felt really makes a difference in my own work, anyway. It just seems like at the end of the day, the people who whip out awesome stuff really quickly can say "Well I did that before", or "I fucked something similar up one time a few years ago and had to fix it...". Seek out failures and run away with every ounce of knowledge you can!
although sometimes speed is a quality all by itself
ayyup! it's incredible how much you'll pick up, especially in the first months
Thanks! Again, great advice and greatly appreciated. After taking in everyone's comments, I think the best things for me to do from here are as follows:
1) Plan out work more thoroughly, and anticipate things that may be new or difficult
2) Make better use of optimization techniques/hotkeys/etc.
3) Try new things and gather knowledge through more than one method
4) Be mindful of time, but take breaks from work to eat/do pushups/ refocus every so often
5) Learn when to stop.
6) Go for quality practice now, and speed will come in time.
Thanks again everyone!
"...and then I wake up" :poly142: