Advice?

[Per-Anders]

well, you should try to break down the objects into their simplest forms, and just really learn the tools so what you see for instance a shape that has a curve on one side you know how to make that shape, or any shap really, it's just getting to know the tools at your disposal, and working on the thorugh process. if you have problemd with breaking objects down mentally then try doing it by drawing them first, and simplifying the object into what shapes you think should be there, and using that. aside form that, practice practice practice

[JamesMK]

Now there's a BIG question :D

Of course, the answer depends a lot on who you ask and what kind of project we're talking about. I would say that the process is not the same all the time, that is, there's no standard procedure that can be applied to anything - at least not for me.
Every modeling task is different.

First of all, you should listen to the helper bunny about learning the tools. But that's only the start of course, allthough an important one. It's not until you know all the tools that you will be able to pick the right one for a particular task. The simple solution is to do what you are doing now - tutorials, or to do what I usually do - work a few silly projects while constantly testing every single button, menuitem and plug-in available and do everything a different way each time, just to get a feel for the difference between function X and function Y.

But that's only the first phase, so to speak: knowing the tools.

The real deal is the second phase - finding a workflow that's good for you. I believe this is a very personal thing....

Brain picking begins::::

I start out with sketches. Pen and paper and lots of sketches. I'd say I draw my object somewhere between three and ten times before sitting down in front of Cinema. Drawing lets me figure out in advance how the object and its details should look and most of the time also how they should be modeled (that's where the tool knowledge comes in handy).

Once I've started working in 3D, it's pretty important to have a solid idea about the final product.
- Is it going to be a still, or do I want to animate it? There's quite a lot of shortcuts and cheats to use if it's not going to be animated.
- If it's a still, how will I compose the final image? The answer tells me which parts need to look good at close range, and which parts will only be seen from far away and thus need not be too detailed.

The final result, still or animated, will have one or several main objects, textures, lights and an environment. It's important to find the correct sequence or priority while working on these various elements. They all contribute to the final result, they are in a way equally important, but they should, in my opinion, be made in a certain order to prevent wasting time on things that turn out flawed and have to be remade over and over.

So, currently I'm following the protocol below:

1. Modeling the main objects and the environment. As far as modeling goes, there's no standard procedure for me. Sometimes I start with the main outline or volume and work my way down into details, at other times I start with the details and add it all up progressively like playing with Lego... It all depends on the subject matter.

2. Make sure that the models look good enough and are detailed enough. Avoid thinking about how textures could make it look better. It has to look really good in plain gray OpenGL - if it doesn't, it's not finished yet.

3. Assigning temporary materials to those models and details while working my way through all the geometry. This means simple materials with a reference texture (a colored grid pattern) mapped on the models. I do not worry about the final look, only about correct map projection, but it's very important to remember that everything is supposed to get a texture or procedural shader later on, so I'd better make the preparations for that right away.

4. Arrange the main objects with the environment objects and throw in a couple of cameras to get a good coverage: longshots, midrange and close-ups. The final composition I have planned obviously dictates what cams I need.

5. Start working on the lighting. I've noticed several times that it's a total waste of time to any elaborate lighting until I've reached this step. The stage should be fully set before I rig the lights. And, I do the lighting before making the final materials as well. Many materials get most of their final look not from themselves but from the environment and lights, that's why I do lighting before materials.

6. Make the final materials, paint textures etcetera and start making testrenders. This is probably the most time consuming step.

Umm... that's about as much as I can make myself write down for now.... :)

[squidinc]

yup, I agree with everything that's been said so far, but as well as sketching, what I've found really useful is to block out the scene using lightly modified primitives, sort of like sketching in 3d, this way you can see a rough idea of what the scene will look like without spending a lot of time modelling it all.

[LucentDreams]

Originally posted by squidinc
yup, I agree with everything that's been said so far, but as well as sketching, what I've found really useful is to block out the scene using lightly modified primitives, sort of like sketching in 3d, this way you can see a rough idea of what the scene will look like without spending a lot of time modelling it all.

Mine aren't as "lightly modified" as I think Squid may be refering to, but definitely agre with him none the less, in traditional mediums a quick sketch done in about 30 seconds to a minut getting the pose and perhaps basic structure is known as a gesture sketch. I've never seen anyreason not to do the same in 3D (granted a gesture model takes a little longer I'd aim for about 15 minutes)

here is an example, I'd show another one but I can't seem to connect to my srever :/

The image I'm working form, used with permission from Michel Gagné (http://www.gagneint.com)
http://cgi.third-era.com/~kaiskai/misc/rabbit5.jpg
The gesture model part way early into the process.
http://cgi.third-era.com/~kaiskai/misc/bunny.jpg

[fretshredder]

thank you all for your truly valuable and helpful advice. I will continue doing what I am doing, and hope that somewhere along the line I can "find my way" so to speak. Also, I hope that after I get a better understanding of the tools I am using I will be able to create something without that sinking feeling of "where to start" everytime I try to model :)

Well, thanks again, and I hope to have something to share soon
:thumbsup:

Keep rockin'
/gz

[Martin Kay]

All of what the others have said, plus you most definitely need to get a very good grasp of what the modelling tools in c4d can do and what they can't do, before you can confidently start out on a project. Whilst the modelling tools in c4d can do most things, there are some things, like complicated rounding that is very complex to achieve easily. Personally I'd advise modelling real life objects, like watches, cameras, hi-fi's and other easily accessible household objects. Its far too easy to bumble along making things up/inventing things and never having to address the 'problems' of simulating real world shapes. Struggling to model even a simple key-ring will teach you more about modelling problems than aimlessly mucking about making robots etc where essentially one 'goes where the software leads you' as it were, and not where you want it to go.

[JamesMK]

Oh, there's one omission here, so I'll add that as well:

REFERENCE IMAGES!

This is vital to just about any project. Even if you're creating something entirely imaginary, there are still details in it that will have, in one way or another, real-world representations. Get some pics of such things and keep them nearby as you work.

[MJV]

Originally posted by JamesMK
Oh, there's one omission here, so I'll add that as well:

REFERENCE IMAGES!

This is vital to just about any project. Even if you're creating something entirely imaginary, there are still details in it that will have, in one way or another, real-world representations. Get some pics of such things and keep them nearby as you work.

Agreed. Google image search is your friend. :) Also, the hardest thing for me is just beginning. So as not to overwhelm yourself, start out modeling very simple things and work your way up. I do the same even with very complex projects. Start with the simple part and let success there inspire you onward.

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.