Basics of modelling

[syshex]

Hello everyone, I’ve been learning blender for a few weeks, and managed to get a “working” understanding of how some simple stuff is made. I followed a few tutorials, messed around a bit by myself with some projects, etc.

Now, bear in mind that I’m really really not a artistic person, I can barely hold a pencil and draw a straight line, let alone draw something creative etc. Not a very visual person, but I am trying. I have a few existential questions that maybe someone has the time to help me out, or point me to the right resources. I’ll be glad to do the reading if references are provided.

1- When drawing a model (lets say a House) , the tutorials I find online do one of two things : Start with a base mesh, subdivide, extrude, mirror, etc all in the same mesh, and end up with a one continuous mesh, OR add meshes as required (like 4 walls? no problem, 4 cubes). Is there a rule of thumb as to what method to decide and use for any particular model?

2- Size references. When I seed tutorials, people grab a base mesh and go like : “I’ll make this bigger” . I think of blender size units as meters (I think I read somewhere that the scale was like that) . I saw a teddy bear tutorial that produced a 3 meters tall bear. So the question is, should one not follow this scale? Wont working on a different scale potentially, later on , result in problems, specially using simulations like fluids, smoke and what not? Maybe I’m getting ahead of myself here.

3- Reference images. For most of my pet projects I use reference images. I place them in Front, Left , Top . Go into ortho, and when I’m in that view I see them. I find myself needing to see them when I’m lookin at the model from an angle, instead of only viewing them from those specific views. This is to allow me seeing the references and add details that I cannot see from the reference when I have a model in front of it. The way I went about it to allow me to see the references from angled views is to create plane meshes behind the model and putting textures on them with the reference images. This has got to be a really silly (and not easy) way of going about it. Is there a correct way to achieve this ?

This is it. Probably really dumb questions, but they stress me a bit because I feel that I might be using this tool in the wrong way.

[lensbreak]

@syshex"

1- When drawing a model (lets say a House) , the tutorials I find online do one of two things : Start with a base mesh, subdivide, extrude, mirror, etc all in the same mesh, and end up with a one continuous mesh, OR add meshes as required (like 4 walls? no problem, 4 cubes). Is there a rule of thumb as to what method to decide and use for any particular model?

You can seperate/join meshes/faces/walls/etc. anytime so it’s depending on what kind of scene youre working I think.

@syshex"

2- Size references. When I seed tutorials, people grab a base mesh and go like : “I’ll make this bigger” . I think of blender size units as meters (I think I read somewhere that the scale was like that) . I saw a teddy bear tutorial that produced a 3 meters tall bear. So the question is, should one not follow this scale? Wont working on a different scale potentially, later on , result in problems, specially using simulations like fluids, smoke and what not? Maybe I’m getting ahead of myself here.

It’s always good to work in the same scale so you dont run into problems with of different sizes. However I dont understand why that should not work with simulations, whats the coherence?

@syshex"

3- Reference images. For most of my pet projects I use reference images. I place them in Front, Left , Top . Go into ortho, and when I’m in that view I see them. I find myself needing to see them when I’m lookin at the model from an angle, instead of only viewing them from those specific views. This is to allow me seeing the references and add details that I cannot see from the reference when I have a model in front of it. The way I went about it to allow me to see the references from angled views is to create plane meshes behind the model and putting textures on them with the reference images. This has got to be a really silly (and not easy) way of going about it. Is there a correct way to achieve this ?

There is an easier way indeed. Open the left panel with “N”, go to “Background Images”. There you set your reference images I guess. Add a new image and choose at the axis-menu “All views” or “Camera”. Dont forget that those images are not visible in the final render view nor has any influences to your scene.[/QUOTE]

Sincerely,
Steven

[syshex]

Hi Steven !

Thanks for taking the time !

@lensbreak

It’s always good to work in the same scale so you dont run into problems with of different sizes. However I dont understand why that should not work with simulations, whats the coherence?

Like, if I was to model a Cup , but make it 5 meters tall and 1 meter in diameter, when filling it with water, wouldn’t the water physics make it behave like I was filling a room , instead of a small glass ? ( I mean splatter wise and droplets wise)

Once again, thanks so much

[Quantum_Anomaly]

Don’t be so hard on yourself for not being artistic. If you remain inquisitive and keep asking “what if?” or “I wonder how…” you will be on the right path. Keep guessing about how various things are made in Blender, then try something out, if that doesn’t work try something else or research what others have done. Keep at it and you will make progress.

1. As lensbreak said, what starts as one mesh can be separated (P key) into multiple meshes. Or you could start with several and join (Ctrl J) them. After joining you may need to Remove Doubles (see the W key menu).
   Once you know how to do that you can liberate your way of thinking about it. Also knowing how to change the object’s origin point is important (there is a button in the Tool Shelf but I have assigned the \ key to do this since I use it often).

2. Sometimes you can’t model things to scale because the 3D view will rotate awkwardly and become difficult with objects that are extremely large or extremely small. But if you have to scale up or down do it in powers of 10 so that you can easily know what the measurement was supposed to be by moving a decimal point. Also know that the measurements you take of any object will be inaccurate if the object’s scale has not been applied - that is if the scale is not X:1 Y:1 Z:1, so make a habit to Apply Scale (Ctrl A) except in cases where you have a reason not to.

3. JA12 had a great suggestion of adding empties set to display images. You can adjust their transparency too. As far as the object blocking the reference images I have two suggestions. A (the quick way): Press the Z key for Wireframe View when you need to see through the model, or B (the more clever/elegant way): After giving some transparency to the image-empty set it to display as X-Ray so it will always be overlayed on top of everything else.

Other useful modeling things to know…
Ctrl N : Recalculate Normals
Shift Ctrl Alt M : Select Non-Manifold (long shortcut, but worth memorizing) Helps for finding inside faces or other problematic geometry
Ctrl L : Select Linked
H : Hide selected object or selected part of mesh
Alt H : Unhide hidden
Ctrl I : Invert selection

I could go on and on listing modeling operations that are useful but I don’t want to overwhelm you.

Congrats on entering the world of Blender and good luck!

[syshex]

Hi JA12,

Thanks so much for taking the time to answer here. Also, the reference links provided proved to be a great read. Just realized a couple of problems I did not know existed, and ways to go about it. The bit about concave polygons was something I had to idea, and had been doing it for some of my tests. Once again, great references and tip.

Also the scale explanation really did bring it home for me. Did not know the units could be changed. I realize that I could have searched for it before asking, so I apologize for that. It is just that this is all too new and it never occurred to me that setting would exist.

Thanks so much for the references and thorough explanations.

[syshex]

Hi Quantum Anomaly

Thanks so much for this quick cheat sheet. Did not know about the P and J shortcuts to join and separate.
I have a few ideas I want to give a try at making and can’t wait to try out the X-Ray trick you mentioned.

My main problem with using wireframe while looking at references is that I don’t really know where in the mesh I’m subdiving and extruding (if in the middle or at the edge of the mesh ) … I seem to need to see some faces to know what is happening. I’m sure it will improve with experience. The point pointed out by JA12 of using dual monitor and having another perspective view on my second screen could actually help with this bit. Will try it out later on.

The problem with lack of visual creativity is that I cant draw anything without references, and even with reference images it gets tricky because I cant imagine depth. Like drawing a persons head : front and side perspective, I have a lot of problems with eyes and cheeks. This is just an example , tried to make a head … and quickly realized I jumped to the deep side of the pool. I’m now working on easier things, like Cups , bottles , rocks and pieces like that

[syshex]

Hi JA12, once again thanks so much. This does help me a lot. I’m relaxing a bit more now with the scales and that.

I’ll try to put everything said in these reply to use during the weekend and next week and will update with newer doubts and thoughts that might occur.

[fredgomes]

Thank you for this informative thread.