Home General Discussion

"Don't sweat the little things... and it's all Little Things" - Workflow question...

polycounter lvl 18
Offline / Send Message
JasonLavoie polycounter lvl 18
Hey all, I'm sitting here just about to chug along on this project of mine, and the one thing that has become painstakingly obvious to me and my workflow is... I tend to always focus on the smallest of details no matter how small or "unimportant" the piece is.

So this is a bad habit that has followed me for a long time now, and I was wondering if any of you gracious chaps and chap-ets have any tips or suggestions that can help me view the "bigger" picture as a whole, instead of focusing only on those small pieces.

Condensed: I focus to much on the small details and would like some tips to help me pull away to focus more on the piece as a whole.

Replies

  • sprunghunt
    Offline / Send Message
    sprunghunt polycounter
    Work in stages. If you rough out your work in one go then you'll be able to see the big picture first. The more stages you do it in the less work you'll have to do because you'll spot problems before they become a lot of work to fix.
  • DarthNater
    Offline / Send Message
    DarthNater polycounter lvl 10
    You know what helped me a lot, look at you're work, at the scale the player will see it.

    Case in point: I used to sculpt the shit out of bricks. Make the damn things looks perfect, sculpt every crack, every rough spot... you name it, I sculpted it. Well, zoom out to about the size you think it will be shown on screen (or in your texture) and see how much of that information is lost.

    What I'm getting at is, only work as far as you're piece needs worked. I used to be the same way so I know what you're going through. Another thing that helped me: Give yourself a target 'time limit' for each area of you're piece. I only get about 45 minutes a day (and usually only 3 to 4 days a week) to work on my portfolio. When you work in tiny windows of time like that, you learn where you need to focus the most time on.
  • sprunghunt
    Offline / Send Message
    sprunghunt polycounter
    double post...
  • adam
    Offline / Send Message
    adam polycounter lvl 19
    Mid last year when I was putting in an effort to do more high poly stuff for my portfolio Paul (Eraserhead) would give me this piece of advice whenever I'd get stressed or hung up on somehting, that I now tell myself with *everything* I do: Just move on.

    If it looks OK, and not HORRIBLE, and its a problem you know can solve or a hurdle you know you can just jump over at another time, just move on. I've saved so much time and stress by that. It's simple, but its so right.

    I'm now using this with my current project: 'Building an Environment', and so far, I've been nothing but happy and stress free working on it.
  • moose
    Offline / Send Message
    moose polycount sponsor
    ^

    think about your scale. Also, you can't have detail without greater form - so try getting the shape of something down first, and let your brain work on the details while you soidify a shape.

    Like Sprug said, work in stages. For example, if you were to make a baseball glove, you shouldn't start by making all the stitchings and leather seams first. You'd want to make the overall "cup" shape of the glove in a blocky form. You then refine that down into a more "glove" like shape, then refine the overall shape into having fingers, and netting. Then hit small details like leather seams, "areas" of leather. And finally go in, and put stitchings all over the seams. Could also think of it like how a tshirt or clothing is made - you wouldnt stitch/sew a shirt together without preparing all of the bits and pieces first! It sounds obvious - but forcing yourself to work in the hierarchal-shape method will keep you on track and not let you get absorbed in details.

    It is easy to get lost in detailing once you get excited about something - but you can lose a lot of time making every detail perfect, when you realize later the shape is off, the form doesnt flow, or the design doesnt work.

    Details are fun, but always keep the larger picture in mind.
  • moose
    Offline / Send Message
    moose polycount sponsor
    WHO BROKE POLYCOUNT
  • bounchfx
    adam wrote: »
    Mid last year when I was putting in an effort to do more high poly stuff for my portfolio Paul (Eraserhead) would give me this piece of advice whenever I'd get stressed or hung up on somehting, that I now tell myself with *everything* I do: Just move on.

    If it looks OK, and not HORRIBLE, and its a problem you know can solve or a hurdle you know you can just jump over at another time, just move on. I've saved so much time and stress by that. It's simple, but its so right.

    I'm now using this with my current project: 'Building an Environment', and so far, I've been nothing but happy and stress free working on it.


    Can you elaborate one what situations this might apply too? specific example perhaps?

    you have me intrigued
  • DarthNater
    Offline / Send Message
    DarthNater polycounter lvl 10
    moose wrote: »
    WHO BROKE POLYCOUNT

    Someone from Game Artisans... Bastards!

    Anyway...

    Another thing, don't be afraid to add details in while texturing. I think a lot of people feel they need to model EVERYTHING. I think a lot of things just look cleaner if they are added in via textures and not sculpted at all...
  • Firebert
    Offline / Send Message
    Firebert polycounter lvl 15
    There was a time back when I played music a lot, and I mean A LOT. Was in a band (who the shit wasn't?!?!) and we had a studio that turned into a nightmare wreck of cables and crap all over the place. To top it off, we were all uber geeks that collected computer parts that would get laid out all over the place. Just an overall disaster zone.

    We decided to clean the joint up and really make it organized and presentable. I had a mostly full time job while the other guys worked only a handful of hours a week. I would come in and help one day, and when I would get back the next, the place would still look the same. Asking the guys what they got done, they were like "well we were able to organize this chest of hard drive screws!" *facepalm*

    That's when I decided to make a motto that I live by to this day. It helps with almost everything you set out to do because it really pushes your motivation as you see big progress and accomplishment.

    "Mass before mess."

    Anytime I start to see that I'm honing in on where to cut that poly JUST RIGHT, I remind myself, "mass before mess".
  • aesir
    Offline / Send Message
    aesir polycounter lvl 18
    I'm envious of you detail oriented guys. I'm a big picture person and often times get lazy with details... It's something I've been working on for a while...
  • JasonLavoie
    Offline / Send Message
    JasonLavoie polycounter lvl 18
    Very good advice all around, thanks guys... I think its just matter of me pulling myself away from those small pieces and keep on moving forward on the project as a whole.

    Adam - To be honest, seeing your work (and defiantly your current project) has inspired me to take on how you think, or at least try anyways. You constantly pump out those pieces for your environment, and they look great, but you keep on moving forward and not getting hitched on those small details. And as you said, you can always go back and fix what needs to be fixed at a later time.

    Firebert - Maannn, a messy studio makes me cringe dude. At times its nice to have all those wires running around cause it kinda gets me in the mood for recording, but if it gets too out of control i start to focus on that instead of playing.

    Either way man, thanks for the advice :)

    Im gonna try and use what everyone has said here and see if I can pump out some good work at a faster pace.
  • Vailias
    Offline / Send Message
    Vailias polycounter lvl 18
    Another suggestion, would be to take your end texture resolution into account.
    If you are sculpting down to more polies than your normal map has pixels you may be over doing it. :)
    512 texture = 262,144 available pixels
    1024 texture = 1.05 million available pixels
    2048 texture = 4.2 million available pixels

    So with a perfectly used UV space, ala a wall or floor texture, you'll have that many pixels worth of normals to play with, or essentially that many polygons to bounce light off of. You could go to double or triple that polycount in your high res, just to ensure smooth sampling, but those numbers can at least give you a starting benchmark for "will this even show up?"
  • Mark Dygert
    Step back, squint if it sticks out fix it, if it doesn't then stop worrying about it.
  • mLichy
    Everyones said most of what there is to say I think. But I will say that being under a tight time constraint also helps u learn to manage time better and not sit and dink around with little details or whatever.

    I tend to mass out stuff first, but I hop around the model when I get bored of a section or feel its far enough along to get the idea. Sometimes I will find myself starting to detail too much, then I stop and move along.

    Also, like others said, zoom back a ways, and if it doesn't change the silhouette or seem quite obvious, forget it, unless you can get the rez, or ppl will FOR SURE see it up close.
  • vcortis
    Offline / Send Message
    vcortis polycounter lvl 9
    One more thing for me to add. I don't have any professional experience yet, but a general rule I follow is this...

    How close is it to the viewer and how important is the object.

    In my current piece I'm going balls to the wall on detail, but smartly trying to apply the above rule.

    I know that if this environment was in a game that the viewer would be on ground level. Therefore everything that is on ground level is very detailed. The higher up and farther you go from the viewer the less detailed things become. I start cutting polys, textures, etc. down because the viewer will never be close enough to tell, or the items will be masked in shadows etc.

    For instance the barrels on the floor in my piece have some definition other than just a cylinder with a texture slapped on it. But the Barrels that are up in the loft section are just cylinders. Yeah you can tell if you look really close, but if you were running around with enemies attacking you and engaged in the actual gameplay you'd have no clue.

    A perfect example of this is a FPS. Weapons which are right up in the players face 90% of the time are high poly counts and high res textures. That piece of wood on the ground not so much.

    I often find myself doing what you're doing and consequently things take longer than they should sometimes, but I keep reminding myself of the above and it helps.
  • Kwramm
    Offline / Send Message
    Kwramm interpolator
    > You know what helped me a lot, look at you're work, at the scale the player will see it.

    very good advice, especially if you're doing production art.
    For personal pieces it can be tricky to determine what you want to show, because you're not limited by an engine to a resolution or camera perspective.

    Also keep in mind "what will the audience/player focus on?". e.g. on Characters it's usually the area around the face, rather than let's say the feet. Keep in mind that most people who are not artists won't look in every little niche and cranny to look for not so polished stuff. So use your time on things that matter.
  • mLichy
    Oh yeah, I also forgot to talk about what type of game your art is for. This has a very big impact sometimes on the amount of detail u put in.

    If it's going to be a first person game, you need to be alot more aware of seams and details in some cases, as well as texture res.

    But if you are making something for a RTS or 3rd person game, it's not quite as important or obvious/noticeable all the time.
  • crazyfingers
    Offline / Send Message
    crazyfingers polycounter lvl 10
    I've struggeled with this a lot myself, I'm getting better, but it's been a long process, often times I think i've gotten better simply because i've burned myself so many times pushing forward with no game plan, learnin' the hard way. Recently i started a very simple scene, with simple objects and terrain so it would be easy to focus on the "big picture" and it's helped me big time. All my textures tile and the UV's are all super simple unwraps that can accept these tiling textures so i can swap in new textures and tweaks very easily. If you really want to change your ways, make it easy on yourself, do a very simple scene and focus on the workflow you want to change, simple art can be great art too, especially if you focus on the "big picture".

    Make it a point not to put in finished assets right away, and use a general 2d concept or reference before you ever put your first asset in 3d. Get everything in your scene after a basic blockout and very quick texture pass and just start hacking away at the weakest part of the scene or what looks the most "off" from the concept, by attacking the weakest areas you'l ping pong across the scene and build it up as a whole rather than finishing this, finishing that and losing the overall vision, and don't tweak your lighting too much, adjust your textures to the lighting. A well executed scene with an eye for the big picture will look good in all kinds of lighting. What a lot of this boils down to in the end is familiarity and speed with the art pipeline, if you're going to focus on the big picture you need to be able to adjust anything and everything quickly. There are millions of little things you can do here and there to speed things up and allow you to quickly dart around in the process of creating the scene. Hope that makes sense.

    Edit: One last thing, quality over quantity. Cant' stress this enough, a few very modular assets can really quickly build up a scene, the less objects you have the more those objects affect the big picture, and the more you reuse textures the more cohesive it is. You can get a full scene done and then add accenting props after you have the scene fully realized.
Sign In or Register to comment.