Technical Difficulty

[Josh1587140445]

I am having technical problems with a model I am currently working on. I mean it is nothing that can not be easily solved. But none the less, it is a matter of detail.

 

I am a very technical person who loves art. But I do not know where to draw the line when it comes to building a model. I try and model every detail. Which as we all know, the more stuff in the model, the longer the rendering time. Like right now I am working out how the gable connects to an archtectural element and trying to model it as how I would draw it with a drip edge and everything.

 

Where do I draw the line on what to model, and what not to model? Or how would you guys handle this detail? A simple plane to fake it, or do you actually cap it and give the piece some depth? But overall, how much detail is too much. One big problem though. The BOSS is very, can we say, indesisive. So the model gets built, and then views are picked from there.

 

So please help me SIMPLIFY my life.

[STRAT]

how much experience have you got?

 

i ask because as the jobs go by you'll soon learn about the level of detail you can add offset against the time constraints you are given.

 

when i started i always used to model every minor detail i could find to model. didn't particulaly make a massive polyed mesh but certainly took ages to complete (and render).

These days i model only what i think will be seen.

Look at building photos in magazines - exactly how much minor detailing do you actually see? not as much as you think.

I hardly ever model depth to window frames anymore for example. i speed through making a basic model and make it 'realistic' by using materials and effective lighting.

 

Besides, whats the point in chamfering a curved handrail with hundreds of faces to get it looking smooth for example when you'd prolly never even notice it in the final image?

[Josh1587140445]

Not a lot of experience. Plus I have taught myself most of what I know. I did alot of modeling in school but it was over a period of time. And I started working full time this past May. And in that time frame I have done 6 models, not including the one I am working on now, and the one that is slated for next month. But what is nice is that both myself and the office are in a learning curve. I am the first to offer this, so they kind of cater to my needs at this point and time.

 

But I am worried about the day I get a project and they need it quick.

 

So what do you mean by BASIC MODEL? How do you go from start to finish? I do the technical building of the model in AutoCAD, and then add doors, windows and everything else in VIZ.

[STRAT]

i model 99% of the building in autocad. i find it quicker and more precise than viz/max, but thats just my personal preferance.

 

by basic model i mean i model the shape and basic structure all in 3d faces first, which gives me a shell to work on. then i methodically model in basic windows, doors, stairs, etc into the shell.

 

then depending on time constraints i'll go back to elements in the drawing and 'beef' them up or add even greater detail to them.

[siliconbauhaus]

this comes from years of experience....

 

do yourself a favour. dont bend over backwards and pull something out of the bag at the last moment on a rediculously tight deadline...they will expect it from then on. if you dont think you can do something withing the time limits say so from the start.

 

you are offering your employer something that they dont have so make sure you dont get exploited.

 

if all his makes me sound bitter it's because I am. I've had companies learn what I can do and go out and try to bring in work based on my skills alone, yet make no attempt to compensate me for it.

 

as for modelling, I'm with strat. I do almost all of mine in acad but I am starting to do more in max. the biggest problem with modelling in max is understanding max's method's with things like UCS's and such. If I was starting from scratch I would have preferred to not start modelling in acad and gone stright to max.

 

as for what to and what not to model, you should always approach your project with some sort of storyboard established and get your boss to approve it and stick to it. this way you know what views your going to be seeing and therefore can put more time into those areas and less time in places you hardly see.

 

my 2 cents worth

[nisus]

Hi josh,

 

Two tips:

1) sketch out your rendering on the forehand. This will aid you in knowing what will be visible and what not and will aid you how much detail you need to draw, i.e. how big each detail will be on the final image. To understand this better, scale up your thumbnail sketch to the final output resolution and count the pixels you will need for a door / window etc. If you only have about 3pixels for a door, it's useless to draw a doorknob!

2) look at pictures and do the same procedure as before (scan and scale at full resolution). Look how many pixels that dog or footbal is etc.

 

rgds

 

nisus

[kid]

the more I do this 3D stuff the less detail I actually model and the more I do with clever texturing, bump mapping, displacement mapping. I'm sure I'm not alone in this progression. But knowing when to stop will and can only come from experience. There's no formular for it (except nisus's ).

 

[ January 11, 2003, 09:25 AM: Message edited by: kid ]

[ingo]

More polygones doesn't mean more rendering time. If you work like Josh and build high polygon count modells, texturing needs less time because you can work with simpler textures, which means less rendering time. Everyone has his own kind of workflow, and of course animation is a different beast.

And if you sketch your views like nisus you have something you can rely on and your client/boss knows what you are working on.

 

Just my 2 cents