Best Of
Re: Stuck: My Topology Journey Has Hit a Wall
You can model the gun using sub division (sub-D) modelling or non sub division modelling (non sub_D hard surface modelling). Look up both these definitions on the web. With both forms of modelling you have to have a strategy that works toward an end result. The image you are working over has some nice details but you ignore them in your block out. Particularly the end of the barrel and the back of the top assembly.
There are loads of tutorials available for just this. You can learn everything you need from the interwebs. When following tutorials you need to be flexible and search for principles that work for you. Don't just look for guns, look at any mechanical modelling you can use as a solution, often they might offer ideas and methods for one part and you may need a few until the light comes on and you get the feel of modelling.
Look for the principles of sub-D and non sub-D modelling. An example :
https://www.youtube.com/watch?v=c-u3APkEVH0
There are loads of tutorials available for just this. You can learn everything you need from the interwebs. When following tutorials you need to be flexible and search for principles that work for you. Don't just look for guns, look at any mechanical modelling you can use as a solution, often they might offer ideas and methods for one part and you may need a few until the light comes on and you get the feel of modelling.
Look for the principles of sub-D and non sub-D modelling. An example :
https://www.youtube.com/watch?v=c-u3APkEVH0
kanga
1 ·
Re: 2D chain texture
Make a trim sheet.
http://wiki.polycount.com/wiki/MultiTexture#Trim_Sheets
http://wiki.polycount.com/wiki/MultiTexture#Trim_Sheets
Eric Chadwick
1 ·
Re: Early-Late 2000s CGI/Game Models
Oh my goodness I did these tutorials back in the day. Am I old now?
artquest
4 ·
Re: BRAWL2 [Vehicle] Rust Rhino, Hang Li
You work fast i like the concept and colors, haven't gotten to my "vehicle/environment" stand for my entry hope i also do not have to go oj with it, haven't at all played with colors for it and still debating on even including it for the finals. Looking forward never-the-less to more of this.
iam717
1 ·
Re: How The F*#% Do I Model This? - Reply for help with specific shapes - (Post attempt before asking)
I made this post aggravated at something else, apologies if I seemed angry 
Does this look right?
Does this look right?
5rettski
1 ·
Re: How The F*#% Do I Model This? - Reply for help with specific shapes - (Post attempt before asking)
5rettski said:I don't even know where to start on his nose. I'm just tired already, I feel like I'm running in circles. I don't know where to extrude from to make the shape etc. I need help.
You just started, go on, finish a nose... see how it looks smoothed, shades, behaves from various angles. learn from your mistakes, do another one.
learn and repeat. getting gridlocked at like 8 polygons in isnt gonna solve your issue. dont expect the solutions to be instant. make mistakes, learn, build experience.
Neox
4 ·
Re: How The F*#% Do I Model This? - Reply for help with specific shapes - (Post attempt before asking)
@5rettski There's a few different ways to approach organic subdivision modeling but it's generally a good idea to start by establishing the proportions with a block out of the facial planes to create a base mesh. Part of this block out process usually involves creating basic loop paths around key facial features (eyes, mouth, nose, ears, etc.) using either C or X topology patterns.
Some good examples of the different topology layouts can be found in the polycount wiki:
http://wiki.polycount.com/wiki/FaceTopology
http://wiki.polycount.com/wiki/Subdivision_Surface_Modeling#Organic_Surfaces
http://wiki.polycount.com/wiki/FaceTopology
http://wiki.polycount.com/wiki/Subdivision_Surface_Modeling#Organic_Surfaces
There's also videos from the period (around 2004-2007) that cover the hows and whys of different facial topology layouts:
https://youtu.be/im9XZf2C1s0(C topology was often used for
more stylized characters and X topology was generally associated with
more complex character models that had a lot of skin folds or other fine
details but that's just an observation and is more descriptive than
prescriptive. Often the decision to use one topology layout over the other was made before even starting the block out but it's still possible to re-route loops and change the topology while modeling.)
The basic shapes and topology flow around key facial features can be created by extruding edge strips and connecting the space between them.
Some artists prefer to start by outlining orifices like the eyes, nostrils, mouth, etc. then move on to defining hard landmarks like the brow ridge, nose bridge, chin, cheeks, etc. This is supposed to make it easier to connect the adjacent edge strips since the segment count can be adjusted to match before any of the actual modeling work is done.
Other artists might develop the facial features in isolation, creating an entire nose, mouth, eye socket, ear, etc. then connect everything together and adjust the topology as needed to create facial lines and folds. It's also possible to strip model the muscle groups and face planes then connect them to resolve the topology flow but that seems like a lot of unnecessary work.
Modeling the nose in isolation could be done by starting with the bridge and modeling outwards, as shown below, but it's probably easier to start by defining the Alar cartilage and nostril then expanding out to the tip and up into the bridge. (There's an example of that approach in the wiki.)
The modeling process itself is fairly straightforward: load up a front and side view image then insert and extrude edge strips to define the major planes and loops that make up the shapes. Focus on creating good topology first then work on pushing and pulling everything into shape. Having good topology that flows nicely helps avoid the pitfall of spending a bunch of time making something that looks great but is difficult to connect to adjacent topology.
Organic subdivision modeling is all about establishing the loop paths then pushing and pulling everything into shape. Try to really nail down the larger forms in the first pass then add the loops to support smaller details later, once everything looks proportional and accurate. Otherwise it's just too easy to get lost in the details and it's often difficult to attach overly dense meshes to each other without chasing poles and triangles around the entire face.
The example below is by no means accurate, it's more of an approximation based on the older version of the character and it's been close to two decades since I've modeled a character using pure subdivision, so definitely study the topology references in the wiki.
If you're using Blender there's a somewhat obscure approach to refining
the shapes in subdivision models that was used occasionally in the late 2000's.
This workflow involved strip modeling a character's base mesh to
establish the loop flow then instead of manually pushing and pulling the
verts and edges in edit mode, just switch to sculpt mode and use those
tools to push the geometry around with the subdivision preview enabled.
To be clear, the purpose of this workflow isn't high poly sculpting, it's to enable a cleaner visual preview of the shadows and highlights on the surface and make it easier to push and pull the geometry of the base mesh into shape so the final subdivision model has a more refined feel. It's also way easier than tweaking individual verts with the move tool and tabbing in and out of edit mode to hide the wire frame.
Another thing that really helps with character modeling is to have other references up and regularly orbit around the model while switching the camera from ortho to perspective mode to get a better sense of the volumes. Looking up art anatomy references and learning the basic muscle groups can also be really helpful.
There's a few artists that still post here who specialize in characters or otherwise have more expert knowledge of anatomy so hopefully they can chime in and share their experience or memories of subdivision modeling from this era.
FrankPolygon
6 ·
Re: Early-Late 2000s CGI/Game Models
This was partly touched on in the other thread, but basically, like this:
http://www.theminters.com/misc/articles/derived-surfaces/
Bay Raitt, who contributed the images in this article and led the Spiraloid forums (not unlike Polycount, but focussed on SubD modeling), later worked on Gollum.
@Neox also produced some very clean SubD models early on.
http://www.theminters.com/misc/articles/derived-surfaces/
Bay Raitt, who contributed the images in this article and led the Spiraloid forums (not unlike Polycount, but focussed on SubD modeling), later worked on Gollum.
@Neox also produced some very clean SubD models early on.
Noren
1 ·