Home Technical Talk

How The F*#% Do I Model This? - Reply for help with specific shapes - (Post attempt before asking)

Replies

  • vakdlfjas



    how do i fix this kind of seam? 
    when i do something like this shape ( cylinder rectengle intersection? or something with cylinder )
    sometimes i make them separate objects but sometimes it doesn't work 
    i tried bevel and insert edge loops

  • sacboi
    Offline / Send Message
    sacboi high dynamic range
    Your example looks nothing like the ref, too high density geometry and a hard edged 'curved' topology even without the benefit of a posted wireframe, is fairly obvious when attempting to generate an assumed cloth like material.

    So spending additional time applying forethought will often pre-empt potential issues from occurring, in the first place. Now firstly assessing the basic shape which at a glance is rectangular plus using minimal vertices, if further revisions are necessary.

    A personal straightforward method, without utilising instancing:

    • Create a plane
    • Extrude into a basic rectangular box
    • Subdivide 4x edge loops - create diagonal edges via Looptools addon (or other DDC app equivalent)
    • Select alternate vertices, slight constrained to axis manual adjustments in order to replicate the diamond shape pattern
    • Shade smooth then apply a Subd modifier (lvl 2)


  • Filip5
    Offline / Send Message
    Filip5 polycounter lvl 8
    sacboi said:
    Your example looks nothing like the ref, too high density geometry and a hard edged 'curved' topology even without the benefit of a posted wireframe, is fairly obvious when attempting to generate an assumed cloth like material.


    @sacboi
     Thank you for reply sacboi, but this shape is fairly simple. I you look closely the mattress shape is somewhat different. I woulndt have a problem if those were straight lines, but these are slightly different.


  • sacboi
    Offline / Send Message
    sacboi high dynamic range
    "but this shape is fairly simple"

    Oh c'mon put your thinking cap on! regardless whether an edge/shape is straight or curved the example workflow remains the same.

    ...so how would we create a curved edge using sub-division modleling?!

    EDIT:
    Hint - your attempt is almost halfway there, instead of hard edges...
  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    @vakdlfjas Welcome to Polycount. Consider checking out the forum information and introduction thread.

    Placing support loops along the edges that make up the walls of a cylinder will causing pinching. The topology needs to be routed around the existing cylinder geometry so it doesn't interfere with the smoothing behavior at the base of the cylinder. Resolving smoothing issues around surface details and shape intersections on cylinder walls is a common question and the last couple of pages have some great discussions on this topic. It's definitely worth the time to skim through this thread and look at how other artists have resolved smoothing artifacts on similar shapes.

    Here's a few links to get you started: Wirrexx recently posted a good example of how to use existing geometry as support for shapes that intersect the top of a cylinder. A few posts before that I shared some examples of how to route topology around slices in cylinders and a couple of pages back there's a post about modeling boot treads that could be used to resolve the smoothing issues without having to increase the number of segments in the cylinder.

    The basic idea is to route the support loop topology between the existing edge segments that make up the cylinder wall and direct any excess edge loops perpendicular to the edges that make up the wall of the cylinder. This allows the support loops to sharpen the individual segments without deforming the outside of the cylinder.

    Here's an example of what the process could look like if the number of segments in the cylinder shape is constrained by existing geometry and cannot be changed. Select the cut outs, inset the width of the support loops, dissolve the faces and inset again to add support around the ID and outer wall of the new shapes. Working with such a limited amount of geometry means the mesh will deform slightly and the highlighted segments need to be scaled inwards to compensate for the smoothing stress and bring everything back in line with the existing cylinder wall.

    This approach is quick and works well when there's a limited amount of geometry but it's not highly accurate. Whether or not this is good enough depends on the scale of the object and how closely it will be viewed.


    Increasing the amount of segments in the starting geometry tends to increase the overall shape accuracy but it can make the mesh harder to edit. It's generally considered best practice to evaluate how closely the model will be viewed and use the appropriate amount of geometry required to accurately hold the shapes at the desired view distance. Here's an example of the same process but with twice as many segments in the starting geometry.


    Recap:
    Search for existing solutions to similar shapes and topology issues.
    Use the appropriate amount of geometry required to accurately hold the shapes.
    Place intersecting geometry between existing geometry segments or route the support loops around the shapes that need to be supported and away from the underling cylinder's base geometry.

    @sacboi I like the process you posted. It's simple and effective.

    @filip5 Without a wire frame it's somewhat difficult to see exactly what's not working in the original mesh but sacboi's advice is a solid approach. What will this model be used for and how accurate does it need to be?

    Not to be too contrary here but the quilted fabric in the first reference image does have a diamond pattern where the padded backing has been sewn in with perpendicular diagonal stitching. The slight puckering at the peaks of the diamonds is just where the two stitch lines cross over. In the second reference image the stitch lines have a different shape and  (assuming) orientation?

    There's a lot of different ways to get patterns on to shapes so what's the right approach depends on the end use for the model. If this is going to be baked down to a low poly mesh for a background asset in a room then it might just be more efficient to add this shape to the normal texture in Substance and call it good. If this is going to be high poly mesh for product advertising or VFX then a different approach, like modeling everything out, sculpting the details or using Marvelous, might make more sense. Again it all depends on how the mesh will be used and what it needs to do.

    Here's one approach that will hold fairly complex patterns without requiring too much work. Start with the basic dimensions. Subdivide the mesh so each quad segment is appropriately sized to the pattern. Match the edge flow the the general direction of the pattern. (In this example the diagonal edges were created by subdivided the mesh again and using a simple un-subdivide modifier.) Select all the quad faces and run an inset individual operation. Select all the triangle faces and run a normal inset operation. Use weights or groups to add a bevel modifier to the peaks of the diamonds and along the outer edges of the mattress. Adjust the mesh to replicate the pucker between the diamonds by using select by attribute to grab the quads at the peak of the diamonds and shrink them down using individual origins. Apply a subdivision modifier and (if micro texture is desired) add a texture displacement modifier using procedural cloud noise.


    This same process can be adapted to fit whatever the pattern is. Leveraging select by attribute functions will make it easier to quickly highlight and isolate specific loop segments in the mesh. It's also possible to create tillable segments (either on the existing shape or as flat patterns that can be deformed to the correct shape) and use mirroring or instancing to speed up the process.

    Here's an example of how the wavy stitch pattern could be made with this same modeling strategy. Start with the basic shape. Subdivide the mesh to the appropriate density to match the scale of the stitch lines on the fabric. Round over the edges of the shape with a bevel / chamfer operation. Loop select the geometry for the first tileable pattern. Inset the selection and repate the process until there's two matching tile sets and an end cap. Duplicate the tileable segments until the mattress is the correct size and then cap the end. Add a subdivsion modifier and the model is complete.


    With difficult patterns, try working through the topology on a flat plane first. Definitely try to use the minimal amount of geometry and work on the smallest possible pattern segment to keep excess work and complexity to a minimum. Leverage modifiers whenever possible to speed up the work.

    Depending on how the pattern tiles, it should be possible to speed up the process quite a bit by using the minimal amount of geometry required to create the shapes and using the right combination of modifiers. Ideally there would be a way to instance and deform a single tileable segment into the whole shape so the pattern and size can be adjusted at any time but setting that up is well beyond the scope of this post.

    Recap:
    Evaluate how closely players will view the model and decide on an appropriate level of mesh detail.
    Look at alternative workflows like normal texture patterns, sculpting and cloth / sewing simulation tools.
    Block out the mesh and use the minimum amount of geometry required to match the shapes of the pattern.
    Use simple tiling mesh segments and leverage modifiers wherever possible.
  • Filip5
    Offline / Send Message
    Filip5 polycounter lvl 8
    Here I go with the solution I have used - I did lowpoly for required shape, then duplicated that shape and connected edges. Afterwards I used inset and moved faces constrained to axis. I did a plane shape like this, whichô was quite easy. The harder part came in bending the shape into box shape. I had to do some cleanup, but using symmetry modifier it was pretty straightfroward. At the end I turbosmoothed the model by 1 and applied bevel to edge parts. I mirrored the top part to create enclosed shape. I also had to do some retouch afterwards in photoshop. The aim was to export png images, with relatively close look to mattress. For that reason I decided not to go for normal map or texture with height. Took me a couple of hours, could be probably done in half the time.






  • Filip5
    Offline / Send Message
    Filip5 polycounter lvl 8
    Hey guys, any clues what might cause this texture disortion ? It is also shifted right after exporting it from max, I ve tried to reset x form, collapse the mesh but it did not work


  • Filip5
    Offline / Send Message
    Filip5 polycounter lvl 8

    Solved by collapsing the mesh and reseting x form via utilities tab in max
  • viqhaas
    Hey guys...

    This is really bugging me its really simple but i cant get my head around it?! The image is really self explantory.

    Trying to make a protein shake lid but im struggling on how to sharper the areas highlighted in orange without distorting another area,


  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    @viqhaas Welcome to Polycount. Consider checking out the forum information and introduction thread.

    Overall it looks like you have a pretty good start with the base mesh and will just need to adjust the support loop routing on attempt 2.

    The vertical deformation in both attempts is caused by the support loops that run off the back of the tabs and down into the hemisphere where they disrupt the even spacing between the segments. There's a few different ways to resolve this so what's going to work best will depend on what the project constraints are.

    If keeping the current mesh density is important then it should be possible to simply terminate the support loops into n-gons at the base of the shape intersection. There will be some minor shape deflection and subtle artifact but it will be limited to the very small transitional area around the shape and it's unlikely that it will be visible when the normal textures are baked down.

    If a greater degree of shape accuracy is important then it will be necessary to increase the number of segments in the hemispherical base mesh so the support loops around the tabs can flow into the existing geometry around the shape intersection.

    The horizontal deformation around the top of the shape intersection is caused by a lack of supporting geometry above the area where the two shapes merge. This can be resolved by simply placing a support loop (directly above the shape intersection) at the bottom of the tabs.

    With subdivision modeling: Preserving the shape of the underlying geometry and either limiting the deformation to the small transitional areas around the shape intersections or consistently averaging out the deformation over a wider area will help reduce the severity of smoothing artifacts. Increasing the amount of geometry tends to increase the overall shape accuracy but also tends to reduce editability.

    This is where balancing shape accuracy and modeling efficiency becomes important. Evaluate how closely the object will be viewed and determine how much geometry is required to accurately hold the shapes. If minor smoothing artifacts aren't visible at normal viewing distances then there's minimal benefit to spending a significant amount of time improving the results. Use the appropriate amount of geometry for the project's goals and the model's use case.

    Here's an example of what this process could look like using a similar mesh density to attempt 2. The existing geometry in the hemisphere acts as support around the base of the intersection and the narrow support loops that sharpen the tabs simply end as n-gons on the surface of the hemisphere. A horizontal support loop is added directly above the shape intersection to control the upward deformation of the shape transition and can be moved up or down to adjust the width of the shape transition.


    Here's another example that shows what this process could look like when geometry density of the hemisphere is increased to match the number of supporting segments in the tabs.



    How much geometry is required depends entirely on how accurate the shapes need to be and how closely the model will be viewed. Triangles and n-gons are fine as long as they aren't causing any major smoothing artifacts at normal viewing distances. It's also helpful to try and maintain a relatively consistent edge width across similar materials since this will unify the shapes and the highlight roll off. Narrower (sharper) support loops read more like metals and machined surfaces while wider (softer) support loops read more like plastics and molded parts.

    The last half dozen pages in this thread have a lot of great examples of common shapes and smoothing artifacts so it's worth taking the time to skim through this thread and see how other artists have solved similar smoothing issues.
  • viqhaas
    wow @FrankPolygon you really are something! 

    I was reading the articles and seen how you helped people, we need more people like you on this earth!!!

    Keep being amazing. Thankyou!
  • martianlion
    Offline / Send Message
    martianlion polycounter lvl 4
    Hey there! What would you guys suggest to do if I'd want to model a curved plane with a border in the shape of a specific curve - say some decorative wardrobe door panel. Myself I know two solutions for this: either to create a smooth, curved surface on which I paint my mesh with Quad Draw/Shrinkwrap, or to make the mesh flat, and then bend it with deformers; the second is, to create a basic curved plane, with enough resolution, to cut out from it the shape I want. The things I don't like about the first one is creating these "template" objects as a place holders, and still, I need to create the geometry pretty much by hand anyway. Also fidgeting with deformers, like bend, can be tricky to achieve exactly the shape I want. Second solution is my main go to, but still - it often produces too much geometry in the end, (also mesh is unnecesary dense in some places, and it makes it hard to work with, to achieve a really smooth lines), and still I often might have to do a few attempts to find the right resolution that allows to contain all the details. I wonder if you guys have any tips on how to tackle such shapes either in general, or if you have any directions for estimating the right resolution for doing the cuts (like, idk "look for having three points on every 90 degree curve" etc.)

    Here in example A I cut out the shape from the mesh that holds my main curve form; B was painted with quad draw on a template.
  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    @Jossy Welcome to Polycount. Consider checking out the forum information and introduction thread.

    The pinching around the corners of the inset shapes is caused by a lack of supporting geometry. Adding bevel weight to the inside corners of the inset shapes and running the support loop around the edges between the existing shapes should solve most of the smoothing artifacts on the existing mesh.

    There's also a minor shape mismatch between the reference image and the base mesh. The tight corners, when combined with the existing edge flow around the shapes, doesn't leave a lot of room for support loops and the topology in this area can become a little too congested. When blocking out the base mesh try to match the shapes as closely as possible while also providing enough room for the support loops to flow around the shapes.

    Below is an example of blocking out the topology for a mesh where the width of the primary features carries over into the secondary features. The support loops are routed around the perimeter of the shapes and into the inside corners.


    Here's what the topology for the rest of the shape could look like. The majority of the support loops in the shape were added using edge weights and a bevel / chamfer modifier.



    Overall the mesh just needs some minor adjustments and it should be fine.

    The last half dozen pages of this thread cover a lot of common shape intersections so it's definitely worth the time to skim through the thread and see how other artists have resolved smoothing issues on shapes similar to the one you're currently working on.
  • Jossy
    FrankPolygon WoW, Thank you so much for taking the time to help me out. This means a lot, I was stuck on this for a couple of days now, trying to figure it out, I'm so happy I found this page. :) I'll check out the links too.
    Thank You again I will be posting my progress. :D 

  • Blaizer
    Offline / Send Message
    Blaizer interpolator
    @viqhaas here's something i did for my modeling blog that it's pretty similar. You can avoid a huge amount of work using mesh projections with similar polycount/number of edges.


  • JBurk
    Hi there! Back again. I did a bunch of searching and scrolling and I couldn't find a solution. basically, I have a rectangle that I want to set a cylinder in, but the cylinder extends beyond the rectangle. Kind of like a cannon. I am stuck is the portion where the rectangle ends and the cylinder continues. The corner where that event occurs I cant seem to figure out a good solution to avoid pinching or sagging. 
    this is the shape.

    This is my best guess.

    But this is what I end up with.

    What is the proper edge flow, using quads, for joining these two shapes?
  • JBurk
    Okay, I kept playing with it and came up with this.
    still some artifacts thought... 
    Am I just being nit picky? or is there a better way to do this?
  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    @JBurk The underlying topology in both your examples, with some minor adjustments, could work. Which starting topology layout makes the most sense will depend on whether the shapes need a softer or sharper transition.

    With subdivision modeling there's almost always going to be some kind of minor smoothing imperfection and it's important to consider whether or not this will be visible to players at normal view distances. If the smoothing artifacts are confined to a small area that isn't going to be regular focal point then there's often minimal benefit to spending a large amount of time removing the artifacts. When working on small details it's important to zoom out every once in a while and evaluate the overall read of the object's shapes and edge width highlight.

    One thing to watch out for is edge width consistency: If the actual parts are made from a single material (like metal) then try to avoid inconsistent support loop widths around the major forms as this tends to muddy the visual read of the object. Parts cut from metal plates or machined from billets and precision castings tend to have a relatively consistent sharpness to their shapes. Rough cast parts tend to have soft shapes like draft, fillets and round overs that are interrupted by sharp machined surfaces. Mass produced plastics tend to have much softer edges and transitions with a lot of variation.

    A couple of posts up there's a discussion about a similar shape with an intersecting cylinder that sticks out from a flat surface and that post has additional links back to a similar discussion on page 178 about blending rounded rectangles.

    When blending curved surfaces into other shapes there's two basic approaches: placing the intersecting geometry on the existing edges that make up the cylinder walls and adding the support loops around the intersecting shapes or placing the intersecting geometry between the existing edges that make up the cylinder walls and using this existing geometry as support. It's also important to use the appropriate amount of geometry in the cylinder to ensure there's both enough room for additional support loops and enough geometry to hold the shapes without introducing cylinder wall deformation.

    Below is an example that shows the basic topology progression for each approach. Top row is placing the intersection on the existing geometry in the cylinder and the bottom row is placing the intersection between the existing geometry on the cylinder.



    Here's a comparison of wide and narrow support loops placed around the shape intersection generated by placing the intersecting shape directly on the existing geometry in the cylinder wall. To maintain all quad geometry this topology routing requires the number of segments in the cylinder match the topology around the corner of the intersecting rectangle.


    Here's a comparison of wide and narrow support loops with a different topology routing that doesn't require exact segment matching between the cylinder and the intersecting rectangle. In this example the intersecting shape is placed directly on the existing cylinder wall geometry.



    Here's a comparison of wide and narrow support loops when the existing cylinder geometry is used to support the shape intersection. Both shapes are placed with a slight offset and merged. This offset provides space for additional support geometry without disrupting the smoothing behavior of the curved surfaces.



    There's a number of ways to handle the topology around this type of shape intersection and which layout makes the most sense will depend entirely on constrains from adjacent geometry, tools used to generate the support loops (manual placement with loop cut tools or automatic placement with modifiers) and overall mesh quality and shape accuracy goals. Most of these topology strategies work with or can be adjusted to work with modifier base support loop placement.

    As far as the all quads requirement: there's a number of great discussions in this thread about how to use triangles and n-gons appropriately. Working under self imposed, artificial restrictions can really have a negative impact on the amount of time required to model a given shape and the overall quality of the final mesh.

    It's important to compare different topology layouts on ease of modeling, smoothing behavior quality, edge highlight consistency, editing flexibility and topology cleanliness, etc. Blindly following popular mantras like "quads only" and "manual loop placement is superior" can really lead to some nasty habits that become massive time sinks without offering any tangible improvement to the final product.
  • JBurk
    @FrankPolygon
    Thanks for the thorough response again!
    Sorry I missed those previous posts.

    Thanks for the pointers on materials, and the warning against those artificial restrictions. I am a master of those.

  • cookedpeanut
    Offline / Send Message
    cookedpeanut polycounter lvl 12
    @FrankPolygon Would it be possible to make an image dump somewhere, with all the images you've posted in this thread?
    Would be nice to see everything in one place for reference!

    One problem I've repeatedly had problems with, is modelling the scroll (head area) of a violin. Everytime I've tried to model the scroll, I've given up in fits of anger.

    Here's what I have so far:



    This is what I'm having serious problems modelling:



    Maybe someone has an idea how to easily approach this shape...

    Edit:

    Alright, I tried to give it my best shot, although I'm not happy with the scroll, especially where the neck intersects with the scroll. The density is just too low and results in the corner looking too soft. This was done with a lot of vert pushing, I'd love to know if there's an easier way to do this, still.


  • komissar
    Hi there!
    Seems that my question will be too nooby, but I'll try ) 

    Is there any elegant way to resolve topology of supporting edges here without adding too dense geometry around this area?
    Or I just need to do some smoothing iterations to get more geometry to work with?
    What is the right approach in this case?
    Thanks in advance and have a nice day ahead!


  • Sims_doc
    Offline / Send Message
    Sims_doc polycounter lvl 7
    Hi! I'd like some advice on how to approach this edge flow. It's the forward sight post on a AK-47 and there's a tiny nook that is milled into the sloped area on one side of the sight post but, I'm unable to come to a good solution that is still a quad and connects the sloped area to the flat area while still keeping the highlighted edges hard.

    I believe what, I have right now should work functionally but i'd like to know if there's a better way around this?
  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    @martianlion Modeling the complex shapes of cabinet inlays flat then deforming them into shape can be a good option. A lot depends on what the final model will be used for and how accurate the details need to be. A couple of other options would be to use floating geometry or to create a clean high poly of the basic shape and do a sculpting pass on it in ZBrush. Maybe someone else has some other ideas about how to add repetitive surface details quickly and efficiently?


    @JBurk You're welcome. No worries, you're doing well, it's all good.


    @cookedpeanut Categorizing them by shape and making them visually searchable is something I've been looking at. Another thing I've been considering (if there's interest) is a high resolution collection of just the final cage meshes with their unsmoothed / smoothed wire-frames and subdivision previews.

    What you came up with for the scroll looks pretty good. Depending on where the soft areas are you might be able to add a couple of support loops to sharpen it up and end the loops in a triangle or n-gon without effecting the surrounding shapes. Depends on what the overall goals for the project are and how closely the model will be viewed.


    @komissar Welcome to Polycount. Consider checking out the forum information and introduction thread.

    A couple of pages back in this thread are a few discussions about similar topics that may have the answers you're looking for. There's a lot of knowledge in this thread, posted by many different artists, so it's definitely worth the time to skim through the thread and look at how other artists have solved similar issues.

    Here's a few links to get you started: On page 177 there's a discussion where Wirrexx demonstrates how to use floating geometry to simplify complex surface geometry. On page 178 Sacboi provides another example of how floating geometry can be used on a variety of shapes and surfaces. A couple of posts down from that Welcj asked a similar question about handling excess support loops on flat surfaces. One of the answers to this question covers using triangles and n-gons to reduce mesh complexity on flat surfaces.


    @Sims_doc If there isn't a specific technical reason the mesh needs to be all quads then there's often marginal benefit to investing a significant amount of time into manually generating quad grid geometry on flat surfaces. If there were any smoothing artifacts then it would be worth considering whether or not they're visible to players under normal viewing conditions.

    The mesh in the upper right corner of your first image looks fine but some of the excess loops on the flat areas could be removed to simplify the cage mesh. The rest of the topology layouts look like they should work but make sure the side of the front post is supported with edge loops like it is in the mesh in the upper right corner.

    Avoiding unnecessary mesh complexity will increase the editiability of the cage mesh and flat surfaces can be a good place to end edge loop segments. In most cases: if the mesh is properly supported, has good edge flow and subdivides without causing any major smoothing artifacts then it should be passable.
  • komissar

    @komissar Welcome to Polycount. Consider checking out the forum information and introduction thread.

    A couple of pages back in this thread are a few discussions about similar topics that may have the answers you're looking for. There's a lot of knowledge in this thread, posted by many different artists, so it's definitely worth the time to skim through the thread and look at how other artists have solved similar issues.

    Here's a few links to get you started: On page 177 there's a discussion where Wirrexx demonstrates how to use floating geometry to simplify complex surface geometry. On page 178 Sacboi provides another example of how floating geometry can be used on a variety of shapes and surfaces. A couple of posts down from that Welcj asked a similar question about handling excess support loops on flat surfaces. One of the answers to this question covers using triangles and n-gons to reduce mesh complexity on flat surfaces.

    Thanx a lot, Frank! Will check it out!
  • Grim06666

    I want to create a polygroup of the alpha using the slice curve, but theres no drag rectangle option. Is there another way to do this or maybe I'm missing something.
  • komissar
    wirrexx said:
    To add too  @FrankPolygon post, a lot of you guys forget that you can use floaters, saves time and is less hassley.


    * Create your shape
    * Loops around the area you want to panel
    * Copy the face you want and "intrude", create support edges (depends on how hard you want them
    * turbosmooth x 2 and you have the results =)

    I made a softer version but you can easily go harder on the edges

    Great Idea!
    But if floater will have some thickness and contain any panel breaks (inner extrusions), will they overlapped with the main shape and disappear in it?
    Or I just don't get it well?
     
  • Blaizer
    Offline / Send Message
    Blaizer interpolator
    @Sims_doc Hope this helps, it's quads only.


    And this is a solution for another problem posted, but with different bevel radius.

  • G0056
    Offline / Send Message
    G0056 polycounter lvl 2
    I'm trying to model the receiver of a double barrel shotgun, and I'm struggling with this tricky zone, any tips to model this shape? Btw, I'm using blender (and I'll probably use zbrush and zremesher to create the high poly if necesary)

    Examples:



    My model right now:



    I tried doing the rest of the receiver first, and then using a round cube to create the shape, but it didn't work well. How could be the best way to approach that shape?
  • sacboi
    Offline / Send Message
    sacboi high dynamic range
    Check out previous advice offered by @FrankPolygon & @wirrexx


    Although, personally I'd boolean the main shapes first, when blocking out then optimised mid poly chamfer workflow for detailing.  
  • sacboi
    Offline / Send Message
    sacboi high dynamic range
    komissar said:
    Hi there!
    Seems that my question will be too nooby, but I'll try ) 

    Is there any elegant way to resolve topology of supporting edges here without adding too dense geometry around this area?
    Or I just need to do some smoothing iterations to get more geometry to work with?
    What is the right approach in this case?
    Thanks in advance and have a nice day ahead!


    Learn too work with existing geometry as supports when for instance introducing alternate axis aligned transitioning topology, there are many examples already posted over the last 5 pages alone. 
  • wirrexx
    Offline / Send Message
    wirrexx ngon master
    @sacboi <-- what this man is saying, many examples over the pages. Please use these. @FrankPolygon has been continously working hard showing what one can do.
  • rogi92
    Offline / Send Message
    rogi92 polycounter lvl 3
    Hey have nice day : )
    Please, how i need fix the topology to get rid of this pinching?
    Thank you . )


  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    @rogi92 Welcome to Polycount.Consider checking out the forum information and introduction thread.

    Pinching around shape intersections on cylinders is a common problem that's discussed a number of times in this thread. On the previous page there's an answer to a similar question and on page 177 there's another discussion about adding square shapes to cylinders.

    Take a look at the answers there and try applying them to your shape. If you run into any other problems post here with an explanation of what the problem is, what solutions you have tried and what results you'd like to achieve.

    This thread is a great resource to see how other artists have solved similar smoothing problems. It's definitely worth the time to skim through and look for examples of similar shapes.

    Floating geometry is placed above the surface of the mesh so any indented areas won't poke through the underlying geometry. There's different strategies for blending the shapes together and sometimes it's appropriate to leave a hole in the underlying mesh and pass the floating geometry through it. In cases like this the floating geometry will have a flange around it that covers the hole so there won't be any baking artifacts. The Polycount wiki has a small section on floating geometry.

    To add to what Sacboi has already recommended: here's a couple of additional examples that are relevant to this shape adding intersecting cylinders to spheres and blending rounded shapes into rectangles.

    A good place to start would be to block out the basic shape of the ball fences (at the end of the breech) with a couple of spheres and adjust the number of segments until it all lines up with the rest of the geometry in the adjacent parts. Merge the shapes down, fill in the top portion and carve out a hole for the top lever and blend in the firing pin ports.

    Take a look at the recommended examples then try blocking out the shapes and connecting the geometry as best you can and post the results. This way we can provide more specific feedback on any issue you encounter with this particular shape.
  • komissar

    @komissar Welcome to Polycount. Consider checking out the forum information and introduction thread.

    A couple of pages back in this thread are a few discussions about similar topics that may have the answers you're looking for. There's a lot of knowledge in this thread, posted by many different artists, so it's definitely worth the time to skim through the thread and look at how other artists have solved similar issues.

    Here's a few links to get you started: On page 177 there's a discussion where Wirrexx demonstrates how to use floating geometry to simplify complex surface geometry. On page 178 Sacboi provides another example of how floating geometry can be used on a variety of shapes and surfaces. A couple of posts down from that Welcj asked a similar question about handling excess support loops on flat surfaces. One of the answers to this question covers using triangles and n-gons to reduce mesh complexity on flat surfaces.

    Thanx a lot, Frank! Will check it out!
  • Blaizer
    Offline / Send Message
    Blaizer interpolator
    @rogi92 this have been explained several times, but i understand that searching 181 pages is a nuisance XD

    With curved surfaces it's a must to add more geometry and if possible, to have a "quad grid" topology to avoid artifacts and tension issues. Here's an example:



    Also, avoid edge propagation and don't add edge loops that will destroy the curvature.


  • Toby3D
    Offline / Send Message
    Toby3D polycounter lvl 6


    I'm not really experienced with booleans yet, but trying to play around a bit with them. The middle piece is what I used on a simple thin cube to carve out a hole, but I'm not sure how to clean it up as efficient as possible, is A, B or C fine, or is there a better way? Any good rules when working with booleans? 
  • wirrexx
    Offline / Send Message
    wirrexx ngon master
    Toby3D said:


    I'm not really experienced with booleans yet, but trying to play around a bit with them. The middle piece is what I used on a simple thin cube to carve out a hole, but I'm not sure how to clean it up as efficient as possible, is A, B or C fine, or is there a better way? Any good rules when working with booleans? 
    If this is LP, all of them work. However UE4, Unity does not like long small triangles, in my experience. A modified version of C with less "long" triangles would be my choice.

  • rogi92
    Offline / Send Message
    rogi92 polycounter lvl 3
    Have nice day everyone,

    please i need you help guys,

    Im stuck here, i dont know how to start can you give me idea?

    Is good to start with cylinder and edit poly by poly to get the shape or is better to make the shape first and at last make the cylinder shape?

    Thank you for tips guys.

  • wirrexx
    Offline / Send Message
    wirrexx ngon master
    rogi92 said:
    Have nice day everyone,

    please i need you help guys,

    Im stuck here, i dont know how to start can you give me idea?

    Is good to start with cylinder and edit poly by poly to get the shape or is better to make the shape first and at last make the cylinder shape?

    Thank you for tips guys.


    Tip of the day: firstly show us what you got. If you have not started, there's another possibility! =)
    Draw how you would imagine the wireframe to be over that piece! =)
  • PixelMagic
    Hey folks. I am trying to finish this model. I know I can just boolean in the windows and call it done, but in the name of learning proper toplogy, I am trying to keep the model Sub-D all the way to final. Thing is, the windows pretty much go against the flow of the rest of the surrounding surface. Any recommendations on how to connect everything up? I have included FBX and OBJ of the model as well if that helps.

    https://www.dropbox.com/s/znziyuu2f76ms4k/queenship_windows.FBX?dl=0

    https://www.dropbox.com/s/hbhbxv3nuaa2953/queens_ship_windows.obj?dl=0


  • sacboi
    Offline / Send Message
    sacboi high dynamic range
    @PixelMagic  I'd recommend looking this, it's honestly the best solution for your issue.
    https://polycount.com/discussion/comment/2724988#Comment_2724988
  • rogi92
    Offline / Send Message
    rogi92 polycounter lvl 3
    wirrexx said:
    rogi92 said:
    Have nice day everyone,

    please i need you help guys,

    Im stuck here, i dont know how to start can you give me idea?

    Is good to start with cylinder and edit poly by poly to get the shape or is better to make the shape first and at last make the cylinder shape?

    Thank you for tips guys.


    Tip of the day: firstly show us what you got. If you have not started, there's another possibility! =)
    Draw how you would imagine the wireframe to be over that piece! =)
    Hello, 
    sorry for delay i was off.Actually i just started and i have this:https://imgur.com/QtGur7f
    But im dont know how to make perfect round shape as you can see on the picture.
    How i can do that?
    Thank you.
  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    @rogi92 Manually extruding and filling polygon strips can be a viable modeling strategy but some of the shapes in that mask could make it a tedious and time consuming process. Since you already have a solid start to outlining the shapes: consider streamlining this modeling strategy by using Blender's curve system to layout the major lines of the shape (later converting these curves to mesh) and using Blender's Grid Fill function to automatically generate clean geometry between the lines of the shape.

    Edge extrusion, polygon strip modeling is a fairly straightforward process but the shapes it generates aren't always perfect and it can take a lot of planning, manual reworking and luck to get everything to line up correctly. This process can be quick and tends to work well on simple objects but the drawbacks can become problematic if there's a lot of sharp, complex shape intersections on compound curves. Working an edge extrusion modeling strategy off a single feature, without considering the topology requirements of adjacent shapes, can seriously compound these issues and tends to generate a mesh that's either too dense or not dense enough.

    This is why it's generally considered best practice to block out the shapes of complex objects and match the geometry density of adjacent shapes whenever possible. Starting with the basic shapes and adjusting the number of mesh segments in each feature to match the adjacent shapes can make it a lot easier to plan out the topology flow and ensure there's enough geometry to support all of the major shapes. From there additional details can be added with a variety of different modeling strategies.

    Here's an example of how the shape block out process can be used to quickly create a base mesh that has accurate shapes and consistent geometry density. Start by matching the basic shapes in the reference image. Adjust the segment count of each key feature to match the mesh density of the surrounding geometry. Try to position intersecting shapes where the geometry of the adjacent shapes provides support around the base of the shape intersection. Use tools and modifiers that add clean geometry without requiring a lot of manual input or re-work.

    In this example the base geometry is mirrored and most of the major support loops are generated by a bevel / chamfer modifier. This makes it relatively easy to adjust the width of the support loops or edit the underlying mesh without having to manually adjust the support loops.

    Depending on what the model will be used for, some of the smaller surface details could be added with floating geometry and this would simplify the geometry requirements for the base mesh. The last half dozen pages have a number of great discussions about modeling strategies and topology layouts that will be relevant to the shapes on this mask. Definitely worth the time to skim through and get a few different ideas about how to approach the shapes.


    @PixelMagic To add to what Sacboi has already covered and recommended: there's been a few recent discussions about subdivision topology and adding details with booleans.

    In most cases, it's neither right nor wrong to use booleans on subdivision models. Mesh smoothing behavior around intersecting shapes is influenced by a variety of factors that include the underlying shapes, surrounding geometry density, topology routing, boolean cleanup strategy, etc. Whether or not it's a viable approach often depends on what the mesh will be used for, how the topology has been cleaned up and the surface quality level needed to support the materials that will be applied to the final model.

    Often it just comes down to two questions: how good does the model need to be and how much time can be spent on hitting the quality target? More often then not it doesn't need to be perfect, it just needs to be good enough. With game art there's often marginal returns on doing things the hard way or trying to improve on something that players won't ever see or notice.

    A lot depends on how players will interact with the model. Assuming it's a situation where it won't be easy to change the density and topology of the underlying geometry then it becomes an exercise in working with what's already there.

    Here's a couple examples of the same basic shape subtracted from a compound curve with a boolean operation. First row there's no cleanup around the boolean operation. Second row the surrounding geometry has been made all quads. Third row the perimeter of the shape is simplified but the supporting geometry is left in.



    All that said: if that's the royal J-type 327 star ship and it's going to be viewed up close then it's likely going to require a higher quality mesh to support the highly mirrored material on the exterior. In the past, @Sacboi has shared several links to resources that cover creating car models with a very high quality surface finish. Maybe he would have some additional recommendations on resources for this specific case where the model has materials that replicate a mirror like finish.
  • PixelMagic
    sacboi said:
    @PixelMagic  I'd recommend looking this, it's honestly the best solution for your issue.
    https://polycount.com/discussion/comment/2724988#Comment_2724988 
    @PixelMagic To add to what Sacboi has already covered and recommended: there's been a few recent discussions about subdivision topology and adding details with booleans...



      Thank you for that detailed explanation. The effort is appreciated. Having said that, even after looking at your examples, I simply cannot do it. I think I'm going to have to just accept that doing proper Sub-Ds on the Queen's ship is simply beyond my abilities right now. Nothing I do, nothing I try, nothing I think of, will avoid pinching in one way or another. Especially once I get to triple edging to hold the edges of the window frames sharp. There are simply not enough spans to run those to. I get overwhelmed with n-gons or tris. Maybe one day I'll be good enough to be able to do it.
  • PixelMagic
    sacboi said:
    @PixelMagic  I'd recommend looking this, it's honestly the best solution for your issue.
    https://polycount.com/discussion/comment/2724988#Comment_2724988
    @PixelMagic To add to what Sacboi has already covered and recommended: there's been a few recent discussions about subdivision topology and adding details with booleans. ..
    Thank you for that detailed explanation. The effort is appreciated. Having said that, even after looking at your examples, I simply cannot do it. I think I'm going to have to just accept that doing proper Sub-Ds on the Queen's ship is simply beyond my abilities right now. Nothing I do, nothing I try, nothing I think of, will avoid pinching in one way or another. Especially once I get to triple edging to hold the edges of the window frames sharp. There are simply not enough spans to run those to. I get overwhelmed with n-gons or tris. Maybe one day I'll be good enough to be able to do it.
  • Udjani
    Offline / Send Message
    Udjani interpolator
    @PixelMagic Even if the way Frank show would be doable, if you have more details in other directions and other sizes it will become such a huge pain to do it that is not worthed. You can see this youtube video on sacboi link to and older discussion in this thread. And is possible the best solution for this problem, Also you can do a retopogy before you project it to the base surface so you have loops running in the directions that you need.   

    https://www.youtube.com/watch?v=3rlMzsBWtPY
  • Udjani
    Offline / Send Message
    Udjani interpolator
    Trying to get make a nut that dosen't have any bending, those first 4 you can see the problem I am talking about, by connecting the 6 sides with the round one you end up with polygons that are not flat with results in some bend.



    This one I made using loft/bridge and adjusting how the 2 surfaces will blend, also had do to some manual adjustments on the tips of the 6 sides. I can't see any bending in this one but it's a bit of cumbersome way to do it. So what technique do you guys use for it, show your nut!

  • FrankPolygon
    Offline / Send Message
    FrankPolygon grand marshal polycounter
    @PixelMagic It's difficult to provide more focused feedback without seeing any specific examples of the topology layout's you've tried and the smoothing errors you've encountered. Most of the advice that's been provided should get you close to some kind of usable result but there's no real way around the fact that it will take some effort to re-work the topology to accommodate the windows.

    It may be helpful to focus on experimenting with different topology solutions to see what works and asking for feedback on that rather than focusing on coming up with the perfect topology layout that meets some fuzzy idea of pure subdivision modeling. It's important to avoid conflating specific smoothing issues on this model with the idea that the only way to do subdivision modeling is to have an all quad mesh with a perfect grid layout. Any mesh that subdivides cleanly without causing any major smoothing errors is passable.

    Looking at additional reference images: it seems like the hull of the ship is broken up into several major pieces with distinct panel lines separating each section. The panel lines that run around the window section could be used to break up the mesh into more manageable chunks. This would allow each panel to have it's own topology layout and mesh density to support adding specific details.

    Here's a very basic example of how this could be done with existing topology. Break the mesh up into individual panels as shown in the references. Separate the panels at the center of the panel lines and adjust the topology and mesh density of each panel. Sections that require significant re-work could be done with face snapping, mesh projection, re-meshing or the shrink wrap modifier.

    Having an all quad mesh doesn't always guarantee there won't be any smoothing issues. Subdivision smoothing is still susceptible to artifacts caused by shape and topology flow issues. Likewise, as shown in this example, having n-gons or triangles doesn't always mean the mesh will have major smoothing issues.



    @Udjani The key to these type of shape problems is to remove the human factor and rely more heavily on the raw geometry generated by tools and pure shapes. Using an approach that mimics the operations and constraints of most CAD programs tends to increase the overall shape accuracy and reduce the amount of manual editing required.

    Below is a basic example of how this could be done. Start with the basic shape. Subdivide the upper and lower rings to support the intersecting geometry. Create the angled slice and spin it into shape, matching the number of segments around the top of the nut. Run a boolean operation and then cleanup the mesh using merge by distance. Mark the appropriate edge bevel weights around the key features then add a bevel and subdivision modifier. This way the base mesh remains easy to edit and the support loops can be adjusted by changing the modifier parameters.


    All of this assumes the model replicates the profile of common nuts and that the model will be subdivided. If there's a nonstandard shoulder profile or if this is a smooth shading issue on an un-subdivided model then that's a slightly different discussion.

    Some other thoughts: the angle and height of the intersecting cone shape will influence the width, profile and perceived flatness of the chamfer. Here's an additional example of the same nut with a flatter chamfer profile.



    Keep in mind that mass produced, commercial hardware can have a very rough finish and some actually do have an uneven profile with random imperfections caused by the speed of the manufacturing machines and tooling issues. The final coating process can also impart some roughness into the micro-surface which could cover any minor smoothing issues. A lot depends on the scale of the nut and the backstory behind how it was made.

    If it's some kind of large precision fastener that was hand made in a tool room then it's likely that there's a lot of consistency in the surface features and sharp edges between the shape transitions.

    If it's a small nut that's just slammed out in a factory machine shop to fix a broken part and get something running again then it's likely the corners were just broken with a file or whatever tool was in the machine and it was sent out the door.
  • sacboi
    Offline / Send Message
    sacboi high dynamic range
    sacboi said:
    @PixelMagic  I'd recommend looking this, it's honestly the best solution for your issue.
    https://polycount.com/discussion/comment/2724988#Comment_2724988
    @PixelMagic To add to what Sacboi has already covered and recommended: there's been a few recent discussions about subdivision topology and adding details with booleans. ..
    Thank you for that detailed explanation. The effort is appreciated. Having said that, even after looking at your examples, I simply cannot do it. I think I'm going to have to just accept that doing proper Sub-Ds on the Queen's ship is simply beyond my abilities right now. Nothing I do, nothing I try, nothing I think of, will avoid pinching in one way or another. Especially once I get to triple edging to hold the edges of the window frames sharp. There are simply not enough spans to run those to. I get overwhelmed with n-gons or tris. Maybe one day I'll be good enough to be able to do it.
    Hi there,
    I can relate to exactly what you're going through, subd modeling may at the best of times seem like a 'science' when an apparently unsolvable problem crops up but in the main I personally find a fairly straightforward method to implement an envisioned goal, especially for most hard surface objects. I mean, what mostly works for me at least, was asking myself *why* did this issue arise in the first place?! and what could I've done better and/or might have somewhat prevented occurring?!

    So at a glance, just using your model as an example, any number of techniques we've shared with some forethought on your behalf will provide a workable solution. Now the only other advice I'd suggest, is reviewing Ali Ismail's work generating automotive content via polygonal workflows which may be of interest in particular - Modeling Cars in Polygons  over at Ebal Studios. There's a ton of insight that helped me resolve similar problems in the past.      
  • Blaizer
    Offline / Send Message
    Blaizer interpolator
    @Udjani the second model is the best for me. That's how i have modelled the nuts i use as floaters.

    This is the way i solve too many things instead of using booleans, with mesh constraints or mesh projections. I just need to move the nut base mesh to obtain the slope angle desired. Here's the example:



    And different approaches using different level of details. With more polygons we have better results, but i prefer a smoother look, more natural and with less polygons. I optimize the mesh using pentagons and triangles. The shading on the slope is pretty nice, and the cap is cleaner.




Sign In or Register to comment.