@pr3stl1@ned_poreyra As Eric and Axel have mentioned, the issues with both models can be solved by using the same topology strategy:
Block out the shapes so the segment counts match and use the existing geometry of the primary shapes as support with the secondary shapes intersecting between the edge loops of the primary shapes.
In the first example: the segment counts on both the rounded end of the subtracted shape and the wall of the truncated cone are adjusted to match where they intersect. All intersecting geometry also lands between the edge loops on the wall of the truncated cone. This provides support and a place to run parallel support loops without disrupting the overall flow of the shape.
The additional support loops are added with a chamfer / bevel operation and the two perpendicular loops are slid along the edges that make up the wall of the cone to even out the smoothing stress near the corner. There's some minor undulation in this area but it's only visible when viewed up close, at an extreme glancing angle with a highly reflective material. The subdivision previews show that it's unlikely to be a major issue but it could be resolved by adjusting the mesh along the edge normals to give the artifact a larger area to run out on or by increasing the segment counts to reduce the size of the artifact. It's a case of close enough is good enough.
In the second example: the same principles apply. Block out the primary shapes and intersect any secondary shapes between the existing edge loops. Inset the subtracted area and use a chamfer / bevel operation to add support loops. Shown are three different strategies for connecting up the corners. Each produces a unique visual artifact. Whether or not this is acceptable depends on the size of the object, normal texture details and desired quality level.
Increasing the segment count on the primary shape provides more support geometry and a better result. As Eric and Axel mentioned: use the existing geometry in the shape. Since it's a sphere it should be possible to rotate it into a position so the geometry matches the angle of the subtracted segment. If there's an edge case where this isn't possible then the same strategy of blocking out the shape, matching the segment counts and using existing geometry for support still applies.
It may be counter intuitive to leave space between the edge loops of intersecting shapes but for round object it's often the correct answer. Connecting directly to the grid of edge loops only seems to be correct because it's convenient.
Fight the urge to be lazy. Skipping the block out phase just ends up costing more time and frustration. The block out mesh isn't a throw away item. It's a jumping off point for adding details and working into the subdivision cage mesh. A good block out is the cage mesh without the support loops and a starting point for the low poly mesh.
How would you increase the sharpness of the corners on the sphere?
Use a narrower edge width on the highlighted support loops or use creases to reduce the effect of subdivision smoothing. The width of the edge loops can be adjusted independently or in combination.
If pinching occurs then it's likely the sphere will need additional segments to hold the shape. How sharp the edges needs to be will depend on the scale of the object. Overly sharp edges can make individual surface elements hard to read and can cause baking issues so there's a balance between replicating the exact shape and creating a mesh that bakes well.
Here's some examples of creasing, narrowing the edge width on the support loop around the perimeter of the cut out, narrowing the edge width on the inside corners and a combination of all three.
The center vertices are evenly spaced but appear to pinch outwards because of the subdivision preview. The mesh shading is acceptable so this could be left as is or the center quads could be triangulated to add an edge that will pull the center vertices back into line in the wire frame preview.
@wirrexx@Blaizer Thank you both so much, your instructions and especially the visual showcase have been super helpful! Thanks again guys, really appreciate it!
@Ridergraal Try selecting the edge loops around the shape's perimeter and use a bevel operation to add the support loops. Bevel operations will produce support loops with a consistent edge width and it's less work than manually adding geometry. Check that the object's scale is 1:1:1. Unequal object scale values can impact the consistency of tool operations.
Here's some Blender documentation that explains object scale:
Here's an example of a few different topology layouts. Keep in mind that, when the perimeter edges are properly supported, the topology on the flat surfaces has little impact on the overall shape. There's a lot of different topology strategies and each one is generated
by the differences in the starting mesh topology and the order in which
subsequent support loops are added. It's something you'll have to
experiment with to find the right topology strategy for what you're
trying to do.
This thread has a lot of information about the fundamentals of subdivision modeling and there's a few relevant discussions a couple of posts up and a couple of pages back so it's worth taking the time to skim through these posts.
To recap:
Experiment with different topology layouts and order of operations.
Use tools and modifiers like bevel to quickly add support loops with consistent edge width.
Research and verify existing information on modeling and topology strategies.
I am having significant trouble modeling this door, especially the corner, is there a different method I should employ when adding my edge loops
You are making it too complicated. The door is not just one piece. Try to seperate it into its parts. The frame consists of 4 elements. The hole in the middle is another element. Break it down into its parts. This makes life much easier. A carpenter wouldnt try to create a door out of a single piece of wood.
@Mir76 That video is a great showcase of what the speed flow add-on can do. Modifiers are definitely powerful tools that can be used to quickly add supporting geometry that remains editable throughout the project.
However, having to manually adjust the topology to reroute the edge flow can erase the speed advantage and having to apply the modifiers to merge parts of the final geometry negates the editability. Adding major support loops before adding the modifier allows the modifier to remain editable but the extra support loops will complicate the mesh and lock in the overall edge width in those areas and all of this still requires a significant amount of manual work.
Is there a way to apply the workflow in the video to the base mesh shown below and generate the results shown in the subdivision previews, without needing to apply the bevel modifier and manually add support loops or merge geometry around the intersection?
That's also a pretty strong and sweeping closing statement about Price. If that's an
indictment of destructive modeling techniques or tools then
the legitimacy of a number of other professionals would also appear to be on the line.
Part of subdivision modeling is learning to use existing geometry as
support and deciding where and how to make trade-offs between efficiency and shape accuracy. Manual work and rework can be a serious efficiency issue
in production but it's also important to look at the wider context of how a specific
tool or technique fits into a process.
Some processes and concepts are difficult to show without applying the
modifier stack and other processes (such as Boolean re-meshing and
detail sculpting) just happen to become irreversible after a certain
point. Not sure that would be a valid reason to make such broad statements or write off one technique in favor of another without adding sufficient context.
Access to automated tools and reversible process aren't a shortcut past
learning the fundamentals and learning the fundamentals sometimes means
getting dirty hands. When artists are learning it can be helpful to see a direct cause and effect between a tool and the subdivision preview. The transition to automated tools and processes can come after learning the fundamentals of how topology behaves when subdivision is applied.
Access to automated tools and reversible process aren't a shortcut past
learning the fundamentals and learning the fundamentals sometimes means
getting dirty hands. When artists are learning it can be helpful to see a
direct cause and effect between a tool and the subdivision preview. The
transition automated tools and processes can come after learning the
fundamentals of how topology behaves when subdivision is applied.
Rather than rely on third party solutions, I've only recently begun personally too delve deeper into native implementation of Blender's modifier stack for a non destructive approach NITROX3D which seems too me at least to progressively attain an aligned finesse not dissimilar to a more robust 3ds Max alternative much less seamless integration within a wholly subd workflow.
Andrew Price's tutorials are zero in modeling.
Also unsure what was inferred here, can you elaborate?
I jumped straight to 3:10 (soundlessly) and didn't watch any further, so let me know if I'm showing a solution to the wrong part or something.
1. Set the cursor to the center of the sphere and add an empty. 2. Add a Mirror Modifier to the sphere, enable Bisect X and set the empty as the Mirror Object. 3. Rotate empty.
You can can stack multiple mirror modifiers targeting new empties to get extra bisections, and apply and symmetrize in Edit Mode to mirror the changes to the other axes.
Hi! This is my second time posting here. I've run into a major roadblock with this space station hatch. The two spherical pieces are a nightmare to model. I tried breaking it down into pieces by just isolating the the spherical sections first. But i have no idea how to get them to subdivide properly.
Hi! This is my second time posting here. I've run into a major roadblock with this space station hatch. The two spherical pieces are a nightmare to model. I tried breaking it down into pieces by just isolating the the spherical sections first. But i have no idea how to get them to subdivide properly.
@navneethdodla94 I would start with booleans to create the base shape and clean it up. You have big flat surfaces here, don't be afraid to use ngons. Here is a crude example, should be cleaner in some areas. 1. bool, clean up and add some edges to help the bevel modifier doing it's job. 2. bevel 3. subdiv
@Prime8 Thats a really neat looking result. I've been doing quads only sub division practices and it's been really painful and it does not make sense in so many cases because of how slow it has been. Kept going under the assumption that film and game studios don't hire people who use booleans and messy ngons for production. Would you consider that to be true?
@navneethdodla94 When comparing modeling workflows for film and games there's some overlap but there's also some significant differences in the technical requirements for each. Within each industry there's also a range of acceptable quality levels. Much of this variance is based on how the models are used and what the budgets are. What's acceptable for one company or project may not be acceptable somewhere else.
If the goal is to become part of a particular industry, specialization, company or team then it's important to research who is leading in that field and emulate what they are doing. Comparing and contrasting the processes and work of artists in each field can highlight where and how the technical requirements of each discipline are different. As an example, compare the wire frames of film and game models:
Andrew Hodgson is an artist that works in film and shares a lot of his modeling process and philosophy:
Matthias Develtere's subdivision modeling work for Wolfenstein II is an example of how n-gons and triangles can be used to speed up the production process with a minimal impact on the overall shape accuracy and surface quality:
For high poly game models in general: as long as there aren't any specific technical requirements for all
quads and as long as the mesh is easy to work with and subdivides
cleanly then a base mesh or cage mesh with n-gons is passable. Creating
the high poly model is just a part of the asset creation process. It's
not the entire process itself and it's unlikely the player will ever see
the high poly model. Other parts of the process (the low poly model,
normal bakes, textures, lighting, animation and presentation) will end
up directly in front of the player and are (as a whole) arguably more
important.
There's a few discussions on n-gons in this thread and one thing that's mentioned in a lot of these discussions is that a lot of the misconceptions about subdivision modeling are based on the abstraction and oversimplification of specific and contextual technical issues, limitations and requirements. Often the nuanced context of these situations is stripped away and this can lead to the perpetuation of nonsensical and counterproductive rules. This is why it's important for artists who are learning this skill to take the time to research and verify what's being said.
Another issue is that time, tools and topology are relatively easy to quantify and it can be attractive to look at these factors as a primary benchmark for judging quality. In theory this is fine for process improvement but it can also become a trap where an artist will judge the result of someone else's work based entirely on how well the rules were followed while excusing deficiencies in their own results solely because they followed the rules they made up.
For the shape question: Jan has pretty much covered it all but it's also worth mentioning that it's important to match the segments counts of the adjacent shapes. This will help reduce the chance of smoothing artifacts appearing on more complex shapes.
Thank you @FrankPolygon Your answers are always very insightful!
I've been bouncing all over the place looking for the fastest and the most intuitive workflow for modelling from concept to texture. Came all the way from CAD modelling from Fusion 360 to Subdivision and i'm still not able to properly dig in. It was becoming very counterproductive to keep bouncing between so many tools so i decided to stick with the basics of Sub in programs like maya and blender because i understood that adoption for cad based workflows is very rare among studios and also because understanding subD will be helpful for anyone in this industry.
I started watching andrew's twitch and took classes from Mario Brajdich who have extremely similar workflows. On the CAD side i'm always blown away by people like Alex Senechal and Edon Guraziu's intricate sci fi forms. All of them developed their own understanding and aesthetic of hard surface modelling based on their preferred toolkit. But in the end their design principles speak louder than the tools they use. So i guess i should focus on just the design aspect of models first instead of the toolkit.
One last question. With Blender's adoption rate going up do you think it would be wise to rely on it's famous third party add ons like Boxcutter, Speedflow etc or are they just temporary solutions in an ever evolving field? They all feel extremely intuitive but also carry a lot of limitations. But at the same time they're so helpful during the concept phase. Just wondering if i should give into the temptation and buy them
The best advice I can give is to look at what you like to do, figure out where you want to go with it and reverse engineer everything from where you want to be to where you're currently at. As an example: Start by deciding film or games. Then look at which studios, what roles, what tools, etc. Build up a workflow around that, start running projects through it and modify that workflow based on the results that come out of it.
Sampling different tools and workflows is a good way to figure out what
you like but different tools are geared for different roles. Most tools
come with some sort of trade-off. With a defined role and an established
process or workflow it becomes easier to judge which tool best fits
what you're trying to do and how you like to work. Knowing more than one piece of software is good but knowing when and how
to maximize its use is better. Plugins and add-ons are great. Just be
prepared for when they aren't available on a job or when they fall out
of development or become obsolete
Establishing a workflow and producing some kind of
art should be the primary focus. The tools and technical stuff should
be there to support the creation of the art. If a process or workflow isn't
established then it can't produce results and without results it will be
difficult to evaluate and grow. It's almost impossible to over stress the importance of outlining, starting, completing and evaluating projects. Starting with an idea and bringing it through the complete production pipeline will shake out a lot stuff and will make it much easier to figure out what tools you need and what techniques work the best for you.
I jumped straight to 3:10 (soundlessly) and didn't watch any further, so let me know if I'm showing a solution to the wrong part or something.
1. Set the cursor to the center of the sphere and add an empty. 2. Add a Mirror Modifier to the sphere, enable Bisect X and set the empty as the Mirror Object. 3. Rotate empty.
You can can stack multiple mirror modifiers targeting new empties to get extra bisections, and apply and symmetrize in Edit Mode to mirror the changes to the other axes.
OH MY GOD, thank you so muchhhh. This method is new to me. Kudo to you I now know a new technique
Hi Again! I continued practicing with some photobash reference images and came across a landing gear. This one part is turning out to be really difficult. I tried two methods. The first was making a side view first and then manually moving edges to create the curvature and the second time i tried to do it with boolean. But i'm not able to get the curvature of the horseshoe bit and the cylindrical cut right.
I'm trying to model this, the circular holes part. i want it to be sub dividable quad topology, and the tricky part is the angle at which the holes are oriented, i don't know it. Even after doing the cylinders i need to connect the space between them so it end up all quads, the only way i know is manual.
@Mohamed_Salah_Bchir mate you need to show your attempt first so we can give feedback to better approach it. from the ref photo you can go with symmetry too and collapse it all to work on unique areas.
Mohamed_Salah_Bchi , I thought about it a lot, did some tests and i think the best/faster way to do it is, by hand place cubes for Boolean on a plane, make an inset then do your symmetry and shell it
Mohamed_Salah_Bchi , I thought about it a lot, did some tests and i think the best/faster way to do it is, by hand place cubes for Boolean on a plane, make an inset then do your symmetry and shell it
- hand placing is the problem because there is a certain pattern that i can't figure out, there is no blueprint, i have one like it at home but i'm not a mathematician, i don't know how to calculate angles lol it looks intuitive but it's not, like the degrees from one hole to another between each ring, varies, i don't know how to explain it. - boolean doesn't generate quad topology, so still manual work.
This is my try I started with this
I think it looks believable, but i would still want to make it like the manufacturer does it. for the topology, i created the cylinders and i started rotating them into rings, and i inset them and manually and connected between them, about 85% quads, and i used 1/4 for symmetry anything less wouldn't work. then there is this
It's hard to explain in words but im sure there is some math behind this
- hand placing is the problem because there is a certain pattern that i can't figure out, there is no blueprint, i have one like it at home but i'm not a mathematician, i don't know how to calculate angles lol it looks intuitive but it's not, like the degrees from one hole to another between each ring, varies, i don't know how to explain it.
This part at least can be done with less manual labor by using arrays.
There's a point in which the holes of all rows are organized in a straight line, thus it's possible to use an array to evenly place the "origin" mesh of each row.
You'll need to create a circular array for each row, and how you go about it depends on the software used. To control the pattern change the number of copies in each row, and create the gaps where a row skips a hole by applying the array modifier and deleting the meshes where there should be no gap.
- hand placing is the problem because there is a certain pattern that i can't figure out, there is no blueprint, i have one like it at home but i'm not a mathematician, i don't know how to calculate angles lol it looks intuitive but it's not, like the degrees from one hole to another between each ring, varies, i don't know how to explain it.
This part at least can be done with less manual labor by using arrays.
There's a point in which the holes of all rows are organized in a straight line, thus it's possible to use an array to evenly place the "origin" mesh of each row.
You'll need to create a circular array for each row, and how you go about it depends on the software used. To control the pattern change the number of copies in each row, and create the gaps where a row skips a hole by applying the array modifier and deleting the meshes where there should be no gap.
I don't know much about arrays, i'll check it out.
Hi all. I'm trying to tackle a tricky piece of geometry and after many attempts, I've been unsuccessful. This is my attempt (without smooth preview): It might not be so visible in the images but I am getting some strange pinching in certain places which leads me to believe my topology isn't ideal. How would you guys go about tackling a shape like this? Thanks.
@CodeferBlue In general: flat surfaces are largely uneffected by messy topology so
the accuracy, consistency and flow of the topology around the curved
shapes and shape transitions is arguably the most important.
An overly
rigid grid topology can interrupt the edge flow around the circular features so it's best to have a topology layout that flows
around them instead of across them. Manually
extruding and placing segments for curved surfaces tends to
generate inconsistent results so it's best to rely on tools that
mathematically generate geometry that's both accurate and consistent.
It may be helpful to start by analyzing the reference images to identify and categorize the object's surface shapes. Look for things like contiguous flat areas, continuous shape profiles and the transitional areas between each surface or shape. Visualize how the major surface features interact with and transition into the surrounding shapes. Plan out the basic topology flow needed to support these shapes and the transition areas. Block out all of the major forms, including the negative spaces, adding additional geometry to support shape transitions and other shape details. Work through the block out in stages and focus on creating accurate
shapes with consistent segment spacing, edge flow that follows the major
forms and support loops that hold the shape transitions.
Here's an example of this process using a grid topology that follows the curves of the shapes and relies on tools and modifiers to quickly generate additional geometry for the round over details and support loops.
Here's a second example of the same process only this time n-gons and triangles have been used to simplify and speed up the block out and detail passes. This demonstrates the importance of using tools to generate accurate and consistent geometry that defines the profile shapes and shape transitions.
It's also worth noting that the n-gons and triangles on the flat areas have a minimal impact on the viability of the subdivision model. If the mesh subdivides cleanly and there isn't a valid technical exclusion then it's fine to use n-gons and triangles in subdivision models. The last two rows compare the effect of a support loop around the top of the flat area.
For most shapes there's going to be multiple topology and modeling strategies. What's going to work best depends entirely on the project. The important thing is to work though the entire process. Plan out the topology. Create accurate shapes that are easy to work with. Use the least amount of resources possible while still hitting all of the goals for the project.
To recap:
Study the shapes in the reference images and identify key surface features and shape transitions.
Visualize the topology flow needed to support the major shapes and shape transitions.
Create an accurate block out with clean topology by working through the model in stages.
Rely on tools to create curved geometry that's both accurate and consistent.
@FrankPolygon thanks, I massively appreciate the detailed response. After reading your post a good few times, there are still some things which remain unclear to me. Would you be able to elaborate on:
"It may be helpful to start by analyzing the reference images to identify and categorize the object's surface shapes. Look for things like contiguous flat areas, continuous shape profiles and the transitional areas between each surface or shape. Visualize how the major surface features interact with and transition into the surrounding shapes. Plan out the basic topology flow needed to support these shapes and the transition areas"
This has been really helpful but as a subd noob I'm finding it difficult wrapping my head around why you've positioned your edge flow in the way that you have in your first image. I could just copy your topology and be on my way but I really want to understand how I can identify where edges need to flow regardless of what shape the surface is. Thanks
@CodeferBlue In general: flat surfaces are largely uneffected by messy topology so
the accuracy, consistency and flow of the topology around the curved
shapes and shape transitions is arguably the most important.
An overly
rigid grid topology can interrupt the edge flow around the circular features so it's best to have a topology layout that flows
around them instead of across them. Manually
extruding and placing segments for curved surfaces tends to
generate inconsistent results so it's best to rely on tools that
mathematically generate geometry that's both accurate and consistent.
[...]
To recap:
Study the shapes in the reference images and identify key surface features and shape transitions.
Visualize the topology flow needed to support the major shapes and shape transitions.
Create an accurate block out with clean topology by working through the model in stages.
Rely on tools to create curved geometry that's both accurate and consistent.
This, in answer too your question.
And to elaborate, it's really a fundamental tenet of subd modeling, that essentially an optimized blocked out base mesh from the very outset will to a large extent, mitigate unnecessary editing as the given object is progressively defined whether hard surface or indeed for that matter organic.
I'm finding it difficult wrapping my head around why you've positioned your edge flow in the way that you have in your
first image. I could just copy your topology and be on my way but I really want to understand how I can identify where edges need to flow regardless of what shape the surface is.
In addition, also enable close approximation of the shape/object's silhouette which in turn is when a well thought out topology comes into it's own, supporting an optimal edgeflow.
Like Sacboi mentioned: evaluating the overall form before blocking out the shapes will help inform which topology layouts, modeling tools and order of operations will be most effective.
Looking at the silhouette of this object, the negative space above the grip is a major feature and since this shape interacts with the shape of the stock and the grip it should be part of the basic block out. Leaving out major features and trying to add them in later can cause issues because it often requires significantly reworking the surrounding geometry to make everything fit and flow together.
The object's surface can be divided into two basic categories: flat areas (lighter overlay) and rounded transitional areas (darker overlay). Comparing the overall shape of the silhouette, surface types, shape profiles and shape transitions will help inform how the topology flow should be structured.
The topology layout should support the major features by providing an organized path for the shape transitions to flow around the shapes. Matching the segments of the adjacent shapes and providing enough space for the support loops and shape transitions will require some adjustments to the mesh but it's much easier to work through this now than it would be to try and graft in additional geometry later.
This example shows how the topology flows around the edges of the silhouette, major features and transitional shapes. Adding support loops (highlighted) will prevent any stray topology on the flat surfaces from effecting the smoothing of the rounded shapes and can also be used to sharpen the edge of the shape transition. Softer shapes that have matching geometry segments may not require the use of support loops.
The important thing is to generate accurate geometry and consistent topology flow in the areas that matter. In most cases there's minimal benefit to carrying excess edge loops across flat areas or spending a lot of time trying to maintain a perfect grid topology across the entire mesh. If the major shapes are accurate and the rest of the mesh subdivides cleanly without causing any major smoothing artifacts then it's passable.
@kuronekoshiii The sharper corners will need support loops to prevent the shapes from collapsing and the smoothing artifacts across the rest of the shape can be resolved by removing the excess geometry and edge loops that cross over into the curved shapes. Flat surfaces are largely uneffected by triangles and n-gons so they can be a good place to end the extra edge loops.
Try simplifying the starting geometry, matching the segment counts of adjacent shapes whenever practical and adding support loops around major shape transitions with a bevel / chamfer operation or modifier. Here's an example of what this could look like:
Hi, I'm trying to model the receiver of a shotgun and I'm having some problems:
this little bumps are tricky to model, and make them look good with the superior part of the model.
First I tried modeling the superior part(I was having some problems with the plane shape, but I achieved to make it. But before that, I don't know how to create that shape in the lateral zone. I tried using booleans but it looks bad when connecting with the superior zone.
How do you think that could be the best way to approach this model?
Hi, I'm trying to model the receiver of a shotgun and I'm having some problems:
this little bumps are tricky to model, and make them look good with the superior part of the model.
First I tried modeling the superior part(I was having some problems with the plane shape, but I achieved to make it. But before that, I don't know how to create that shape in the lateral zone. I tried using booleans but it looks bad when connecting with the superior zone.
How do you think that could be the best way to approach this model?
Using Blender: sorry don't have an attempt, I have no idea how to achieve this - coming from Max to Blender I'm at a loss substituting curves for splines. I did try manually extruding verts but it wasn't pretty. I'm hoping there is a better solution - I'd like to get a tight weave and windings (2nd photo).
Any suggestions on getting the wire loop and braid like the first image?
And on the other end of the string ...
I'm finding the twist and screw modifier, and the lack of a helix function, somewhat challenging.
Using Blender: sorry don't have an attempt, I have no idea how to achieve this - coming from Max to Blender I'm at a loss substituting curves for splines. I did try manually extruding verts but it wasn't pretty. I'm hoping there is a better solution - I'd like to get a tight weave and windings (2nd photo).
Any suggestions on getting the wire loop and braid like the first image?
(snip)
And on the other end of the string ...
I'm finding the twist and screw modifier, and the lack of a helix function, somewhat challenging.
(snip)
You want the Screw modifier.
Here's an example using only vertices with Skin & Subsurf for thickness + Screw for twisting. You'll need to do a quick reading on Skin in case you're going this route but are unfamiliar with it.
@G0056 There's a few different ways to approach modeling this shape but with most complex shape intersections it's best practice to block out the major forms and match the segment counts of the adjacent shapes so the intersecting geometry matches the base geometry reasonably well. How accurate this segment matching needs to be will depend entirely on the overall goals for the project.
Here's an example that shows how matching the segments of the rounded fillet on the side to the segments of the round over on the top allows the support loop to flow around the shapes without causing any major smoothing artifacts.
If the overall shape accuracy is a concern then it's important to block out all of the major forms before adding support loops or surface details. It's also probably worth finding some additional reference images that have less lens distortion and using those as a guide to blocking out the overall forms.
Boolean operations are a great option for building up the block out in stages (similar to how the actual object was produced) and can make it easier to adjust the individual mesh elements when matching the segment counts of adjacent shapes. Adding the support loops to the final cage mesh with a bevel / chamfer modifier can make it easier to add details or adjust the edge width during baking tests.
Provided the geometry subdivides cleanly, without causing any smoothing artifacts and provided there isn't a valid technical restraint then it's also acceptable to use triangles and n-gons in a subdivision cage mesh.
Here's an example of this process with triangle and n-gon topology.
To recap:
Block out the major forms before adding support loops and details.
Match the segments of intersecting geometry whenever it's reasonable to do so.
Use modifiers to automate tasks and enhance edit ability whenever it makes sense.
I've spent quite a bit of time getting the following form correct:
I'm talking specifically about the black bolt in the middle. It seems like a normal round bolt but with a number of cuts made out and a slanting edge upwards, which ends in a curve. After trying many different approaches, this is the best I've managed to do so far:
The idea here was to make sure the supporting geometry for the vertical edge loops were there from the start. So they could be edge slided closer to the cut-out to make those edges sharp, without having the pinching effect they create seep through to the top of the screw. The result is the best I've managed so far. Some remarks:
The surfaces at the top going from the straight cut out the rounded shape upwards are non-planar unfortunately. This might be the reason for the weird shading in the corners, below
The corner part right now is a quad, I've tried it with a triangle before but that created a 5-pole in the corner which lead to a lot of pinching from the corner outwards (on the non-cutout circular shape in the middle). (See blue arrows)
It doesn't look to bad. And it its a very small part in the total render. But I would really like to know how to solve a problem like this "The Right Way" (TM).
Anybody could point me in the right direction here please? Thanks in advance!
@tatertots In general: it's best practice to maintain the concentricity and consistent spacing of the edge segments that make up the walls of cylindrical shapes. Moving the edge segments too far out of position will cause the subdivision smoothing to either pinch or pull the mesh around the shape intersection and this can cause significant smoothing artifacts.
There's best practices for how to approach the topology and modeling but in the end what's right for the project depends entirely on what the goals are. Evaluate how the model will be used, how accurate the shape needs to be, how much time can be spent, what tools are available, etc. and use those parameters to decide which way to approach the modeling and topology layout.
It all comes down to what's an acceptable trade-off between time and accuracy. If it's a small part that will have a lot of grime or high frequency normal details and the smoothing artifacts can be contained in a small area then it's probably not worth spending a lot of time on trying to get a near perfect result.
As for what's causing the smoothing artifacts: working off of the existing edge segments of cylindrical shapes will generally go in three directions:
Increasing the segment count to generate support geometry. (Not shown)
Placing support loops on the flat side of the shape intersection. (First row)
Placing support loops on both sides of the shape intersection. (Second row)
In general, increasing the segment count does increase the accuracy of the shape but it's important to balance shape accuracy with working efficiency. Throwing an excessive amount of geometry at a topology problem generally results in an overly dense cage mesh that's harder to work with and this will reduce overall efficiency. Instead the goal should be to use the absolute minimum amount of geometry required to accurately hold the shapes and provide a smooth transition between surfaces.
Working off the existing geometry can be be appealing because it's the path of least resistance. This is viable for smaller, less important parts but it can often lead to significant issues with shape accuracy and smoothing artifacts. Slicing up primitive shapes and placing the support loops on the flat side of the shape intersections can be a quick way to generate complex forms but it tends to produce rounded shape transitions that generate noticeable smoothing artifacts when the edge width of the support loops is reduced in an attempt to sharpen up the transition areas. Following this logic the next step is to add support loops on both sides of the shape transitions. Though this can work for some shapes it generally produces sharp pinching artifacts near the corners and also tends to deform the curvature of adjacent surfaces.
The first two examples below show this approach and although it's definitely viable when speed is more important than accuracy the prominence of the smoothing artifacts will be a major drawback. Instead (as shown in the third example) treating the existing geometry as the support loops and adding or subtracting additional geometry between the existing edges is a much better alternative.
The example below shows how routing the edge flow around the shape intersection and using the existing geometry as support loops produces a simple mesh that subdivides cleanly. Increasing the segment count of the larger cylinder shape increases the overall shape accuracy but it also increases the mesh complexity. Maintaining the flow of the existing geometry while also using it to support the intersecting geometry will tend to reduce the amount of geometry needed and make the mesh easier to work with. When deciding how much geometry to use the size and prominence of the shape will be a major factor. Try to balance shape accuracy with editability and efficiency.
Subdivision modeling is an established discipline so when it comes to solving modeling and topology challenges it's important to research, test and evaluate existing strategies. A couple of post above there's a detailed discussion about a similar shape transition and a couple of pages back there's several discussions about adding or subtracting shapes from cylinders by using existing geometry as support loops. With practice it will become easier to identify which strategies will work best for a given situation.
To recap:
Try to maintain the curvature and segment spacing of the cylinder walls when adding shape intersections.
Use the existing geometry as support loops for shape intersections.
Research, test and evaluate different modeling and topology strategies.
Hey Frank, thanks for the extensive answer. I'm pretty new to modelling and I have to admit that after reading the answer 5 times or so, I still don't quite understand everything.
While I have managed to reconstruct something that leads to similar results, it feels like a trial-and-error approach to get there. I have difficulty interpreting the screenshots you posted. You always post a sequence of three images. I understand the last version is the subdivided/smoothed version. However the step between the first and the second image is unclear. In the images you posted the first one of each series contains a pretty big ngon, which does pretty bad in subdivision. So I assume you start of with a simple version with ngons and then flesh out the geometry in step 2 and have the mesh in that step show the deformation of the subdivision modifier.
Getting from step 1 to step 2 however is a difficult process though. I don't quite get the steps to do this in a straightforward way where I understand why each thing is happening. It's probably a bit much to explain in a single post. I found this resource here: http://wiki.polycount.com/wiki/Subdivision_Surface_Modeling
I'll look at the video's and examples in the above link. Do you happen to have any other resources to recommend for the theory behind topology when subdivision modelling?
I've gone a few more times through the process, and made some screenshots of the different steps I took and the reasoning between things. If you could look over it and correct me where I'm wrong?
From top-bottom, left-right:
Add edge loop where cut-out needs to be. Make it the right shape
Remove all faces, add edges that make the flat area's and fill with ngons
The supporting lines for this shape will be the vertical lines already available. So in order to make the cut-out without altering the supporting lines, I've inset the n-gons (and tweaked a bit). This creates a face loop loop around the detail which can be used to help sharpen the detail.
Remove top ngon and add 3 edge-loops in the middle. Creating 4 faces in the middle
In red I've marked a pole (3-pole), though I missed the pole next to the blue arrows (5-pole). The combination of these two poles bend the face-loops in the right directions. They cause the blue loop to circle the inner side of the detail, and the black loop to circle back up. This is also the reason we only need 4 faces (3 edge-loops) in the middle, because there's only 4 faces (purple lines) that run top to bottom.
Result is all quads
Adding two extra horizontal edge loops to sharpen up the detail at the top of the curve, and the flat bend in the center.
Smooth subdividing.
I hope with the logic above - if correct - that it'll help create a mental model on good topology and how to reason about it. I think the next time I run into a tricky situation like this (though the result seems so eas
@tatertots You're welcome and thank you for sharing your results and followup questions.
Example images in my previous post only show the initial block out, the final topology with a smoothing preview and the final subdivision model with a shiny material to highlight any artifacts. Only showing the starting and ending points of the different topology strategies provides a comparison of the results without removing the challenge of modeling the shape.
Referencing the previous discussion: the important thing about working with curved surfaces is to maintain relatively accurate and consistent edge segments. Any shapes that interact with these edge segments tend to disrupt the topology and smoothing flow so it's generally considered best practice to have intersecting geometry end between existing segments rather than on them.
Comparing the results in the first example in my previous post: all of the shaded models are passable but there's some slight differences in the smoothing artifacts and under the right conditions some of these smoothing artifacts can become major issues.
Deformation of the existing vertical edges that make up the cylinder wall segments can be problematic on larger objects or when the cylinder is locked into using a lower number of segments to match the adjacent geometry.
The edge highlight and highlight roll-off around the shape transitions on the shaded model is a good indicator of the potential severity of the underlying topology issues.
The first example has a relatively soft corner where the flat area transitions into the slope. That's fine for that level of edge softness and view distance but if the model needs to have sharper edges then that topology is going to be somewhat problematic and can cause significant smoothing artifacts.
The second example has sharper edges but it also has a slight pucker under the corner because of the support loop that falls along the outside edge of the existing cylinder segment. This isn't a major issue because the edge width is relatively narrow and there's a sufficient number of cylinder wall segments to provide support. Widening this support loop to soften the edges or reducing the segment count of the cylinder would cause severe smoothing artifacts.
The third example has the most consistent edge highlight and there's no visible artifacts in the corner when the mesh is displayed at a normal viewing angle. There is a slight distortion when the mesh is viewed at an extreme glancing angle but it has a very soft roll-off and it's likely this artifact wouldn't be obvious in the baked textures. This is why this sort of topology strategy is generally considered best practice.
Part of the original question was what is the right
way to model this and the answer to that question is it depends entirely
on what parameters are used to judge the results. All of the results are viable but the third example will tend to be the most flexible under a wider variety of conditions.
Evaluating topology: Overall you're moving in the right direction and your recent results are
mostly identical to the topology layout in the first example. As for
whether or not this is the correct topology strategy and modelling
process, that's something you'll have to evaluate based on the goals for
the piece. If you're satisfied with the process and the results (assuming the mesh subdivides without causing any major smoothing artifacts) then it's passable since this part of the model is small and out of the way.
Evaluating learning: You've been able to work through a lot of the issues with outer walls of the cylinder and have established a clear path for the topology to flow around the shapes. From here it would be beneficial to try replicating the two other topology strategies and comparing the results. This way you'll be able to gain first hand experience with how each topology layout performs.
Self directed learning is a very powerful skill and it's worth
cultivating. That's why it's important to develop a personal learning
process
that includes research, testing and evaluation. Looking over the past
8-10 pages of this thread, there's several discussions about topology
and modeling strategies for similar shape intersections. It's definitely
worth going back and reading through these posts.
When it comes to
modeling the actual shapes there's a wide variety of approaches
available and what tools and what order of operations makes the most
sense will depend entirely on the project. Trial and error
(experimentation) is an important part of the
learning process because it tests the validity of found information and
helps highlight any knowledge gaps, which better informs what questions
need to be asked.
Here's the modeling steps I used to generate each topology layout:
Example 1.
Cut in the edge loop and shear the front segment to match the profile of the shape intersection.
Dissolve the excess geometry to complete the shape profile.
Inset the flat areas that are adjacent to one another.
Select the bottom edge loop and use a bevel / chamfer operation to add support loops on both sides.
Add support loops around the top and sloped shape transition.
Add support loops across the middle of the flat area to support the outer edges and resolve the area to quads.
This approach to modeling and topology is relatively fast but it's also fairly inaccurate and will struggle to hold sharper edges in some critical areas. Note the deflection of the vertical edge segment on the cylinder wall near the perimeter of the flat area and how this generates a minor artifact and highlight hot spot near the corner of the shapes.
Example 2.
Cut in the edge loop and shear the front segment to match the profile of the shape intersection.
Dissolve the excess geometry to complete the shape profile.
Select the outer edges and use a bevel / chamfer operation to add the support loops.
This approach to modeling and topology is extremely fast and fairly accurate when used with narrow support loops and higher segment count cylinders but it will struggle with wider support loops (softer edges) and lower segment count cylinders. Note the slight pucker and pinching artifacts under the corner and along the perimeter of the flat area. It's also worth pointing out that the n-gon on the sloped area has minimal impact on the mesh because the support loops and adjacent geometry hold the shapes.
Example 3:
Cut in the edge loop and shear the front segment to match the profile of the shape intersection.
Cut a flat in the cylinder using a Boolean shape positioned between the vertical segments on the cylinder walls.
Select the flat and sloped areas and use inset to create the inside support loop.
Select the top and bottom of the cylinder and use bevel / chamfer to add support loops.
Cut in the support loop for top of the sloped area.
This approach to modeling and topology is relatively fast and fairly accurate. Positioning the shape intersections between the existing segments on the cylinder allows these existing segments to act as support loops and reduces the overall deflection of the cylinder walls when subdivision smoothing is applied. Any difference in the shapes is take up by the faces between the existing geometry and the added geometry. This helps reduce smoothing artifacts and allows the support loops to flow around both shapes without interruption. Note how the support loops flow cleanly around all of the shapes and the edge highlight travels around the entire shape without distorting in the corners.
Maintaining some offset between the existing geometry and the intersecting geometry is considered best practice when adding or subtracting shapes from cylinder walls. It's worth noting that this topology does require a little extra planning during the block out stage but in most situations it tends to be the best balance between accuracy and efficiency. This topology also tends to produce the least amount of smoothing artifacts when viewed from most angles.
Here's what that process looks like without using Booleans and n-gons: (It's a lot of extra work with minimal improvement.)
When it comes to Booleans, triangles and n-gons: as long as there isn't a
hard technical issue that precludes their use and as long as the final
cage mesh subdivides cleanly there's nothing wrong with using these
geometry elements in a mesh. Used correctly these elements can offer a
significant speed advantage without any significant drawbacks. The
important thing is to block out the shapes accurately and cleanly so the
topology flows without any major disruptions.
Recap:
Extruding or subtracting from the existing geometry on cylinders and curved shapes is attractive because it's easy and feels intuitive. Subdivision meshes need support loops near shape intersections and working off of existing geometry generally means either having to go without complete support loops or having to push support loops out into adjacent shapes. This is what causes a lot of smoothing artifacts on curved surfaces. Sometimes it makes sense to use this strategy on small parts or shapes with denser geometry because of the speed advantage but most of the time it will produce results that are less than optimal.
Blocking out the major forms, matching the segment counts of adjacent shapes and offsetting intersecting geometry from existing geometry helps ensure that all of the surfaces have support loops that don't push into each other. Applying these principles will help solve most of the problems with subdivision artifacts that are mentioned in this thread.
Researching different approaches, making test models and evaluating the results will help weed out inaccurate topology layouts and inefficient modeling strategies.
No questions, just wanted to thank you all who are answering questions and giving out advice. I've got so much to go through!
And @FrankPolygon if you decide to put out a book I'd snap that up in a heartbeat, just picked up all your pdf tutorials today and will go through them tonight.
Thank you so much for this very detailed explanation. This was very clear. I've tried and tested all the different approaches you described and this has given me a whole new view on how to use the bevel/chamfer tools and the inset tool. Insetting n-gons to create edge loops around the details you wish. Using the bevel tool to sharpen up details on both sides of loops. Really great. And less of a fear of "the n-gon". Also, dissolving the edges in the middle is much faster than whatever I was doing... I deleted the faces and then manually started connecting the edges and faces again to make the flat. :P Also the workflow(?) that just starts with the ngrons and then creates more agreeable geometry from that starting point will be helpful in many more cases I believe.
This is absolutely super valuable help, so thank you for taking time out of your day and helping out less experienced (=total noobie in my case) people.
Of the all the approaches, the third approach with the boolean suited me least of all. I found it very difficult to get the boolean well positioned and found the manual approach to be faster in my case and more accurate.
And thanks to Yogifi for mentioning the pdf tutorials, got curious and found them. I'll buy them as well. Any good information on topology and workflow is very helpful right now. I'll also be looking to some of the older pages of this thread. Around 150 the images still mostly seem to work. Though the issue is often very similar where a subdivided mesh is shown (optimal display) and there's no clear idea of what the underlying topology actually looks like.
For instance, the first picture in your previous post, showing the 3 different cylinders. I have no clue how the topolgy of those meshes relates to the methods and images with the descriptions of the 3 methods later in your post. It is hard to find the resemblance.
Again, thank you so much for taking your time to answer these questions.
EDIT: Quick additional question. Is there any particular order you would recommend for your 4 pdf's?
I was planning on going through the weekly hard surface modeling challenge from 2014 (though I'm no where near prepared and instead decided to first go through all of Frank's detailed posts here). What you talk about with N-Gons and bevels to create the topology and join them up, I'm getting a good understand of myself just now after following this from a few pages back:
Outside of specific project requirements, stock 3D certification programs and technical edge cases, there really isn't anything wrong with using triangles and n-gons in subdivision modeling. Flat surfaces are arguably the least effected by messy topology. As long as the corners are supported and the surfaces are co-planar then it should subdivide without causing any major problems. If a mesh is easy to edit and subdivides cleanly, without any major smoothing errors, then it's passable. There's a point where done is better than perfect.
As an example: here's four subdivision previews (left column) and four topology strategies (right column) can anyone spot the subtle differences and match up which subdivision preview belongs to which mesh?
The four topology samples (right column) are:
Triangles and quads.
N-gons, triangles and quads.
Quads only (manual cleanup)
Quads only (tool generated)
The matching subdivision previews are directly across from the mesh samples.
For most high poly baking models, what happens on the flat areas between the edge loops doesn't make much difference. As long as the shape intersections and the support loops are properly structured the flat areas will remain flat. They generally subdivide and bake without causing any major issues.
There are edge cases like highly reflective surfaces where the quality of the mesh does effect some things but whether or not this is relevant for a project is something that can be validated with some test bakes.
The n-gon mesh is easier to edit but there are some cases where curves need the stress from triangles to help pull the edge loops into shape.
Manually cleaning up a mesh to make it all quads is a major pain and often a waste of time. Avoid throwing more work into bad geometry. If it's broken enough that it needs a lot of manual cleanup then it's probably worth rebuilding correctly.
Take the time to plan out the shapes and match the segment counts between adjacent shapes. Plan out the edge flow so edge loops can be added without effecting critical shapes.
Here's an example of how planning out the edge flow and working through the shapes will result in all quads with minimal cleanup. The segment count is a little higher than I would generally recommend for a game model of this scale but the assumption is that the corners on the USB ports need to be curved. This could have been done using a wider support loop on the corners but leaving it in shows how to deal with similar shapes that require more geometry.
Start out with the basic shapes and define a clear path for the support loops to run out on. Match segment counts to minimize the amount of geometry required to support the shape. Rely on tools for creating basic geometry, curves and edge loops whenever possible.
The support loops around the Micro USB port are added with a chamfer operation and run out between the two USB ports. The support loops inside the USB port are added with a loop cut and run out to the sides of the case and around the Micro USB port topology.
Use a chamfer operation to add the support loops around the face of the USB port. Adjust the segment count on the circular case geometry and bridge the edge loops so all of the segments are connected. This side is now complete. The other side will have the same number of segments.
Use the same segment count for the opposite side of the case and the button cut out. There's a support loop that holds the shape for the outside of the case and it eats up one segment of the button so an edge loop needs to be added between the case and the button to equalize the segment count. Additional support loops could be added to the right side of the button if necessary. Bridge the faces to close the geometry. The basic shape is now complete.
Add additional support loops around the button and outside of the case. Move the center loops back to set the button depth and the seam depth between the case and face plate. Extrude the rest of the shape out to form the rest of the case. All quads and no manual cleanup work to remove triangles or n-gons. How much geometry is required will depend on the desired shape accuracy.
Subdivision modeling is about planning out the edge flow and shape intersections while choosing the right trade-offs between efficiency and shape accuracy. If there's no hard requirement for all quads then use the minimum amount of geometry required to hold each shape. Triangles and n-gons that aren't causing smoothing problems are fine. In most cases (if they aren't causing smoothing errors) it's not worth the time to edit them out.
Here's a comparison of two extremes: a mesh that leverages a lot of n-gons and a mesh that has been made all quads for compliance. The n-gon strategy is easy to work with but may not pass some strict QA / QC processes for stock 3D. The all quads compliance strategy hits this metric but the over optimized edge flow makes it difficult to add or remove loops. The geometry is locked in.
What's right depends on the project and part of the process is balancing technical demands with the art process and time management. A product visualization model is going to have different requirements than a background prop for a mobile game. Narrowing down the scope of the project will help determine which strategies will work best.
If a project genuinely requires all quad geometry: Take the time to plan the edge flow and block out the shapes ahead of time. Match each shape's segment count with surrounding geometry and run out edge loops where they won't disturb adjacent shapes.
The mesh you already have shows a good compromise between efficiency and accuracy. There's only a couple spots that could cause problems with subdivision and even then it might not be a problem if the smoothing errors aren't visible when baked out and textured.
Add the rest of the support loops and subdivide the mesh to see if there are any major smoothing problems. If the mesh looks good with subdivision then it's fine.
I was always at a fork between should I be modeling this all quads, or follow what the youtubers are doing with booleans, bevels and creases - especially as I'm just starting out. I guess I still am a bit unsure but I will keep the advice from above in mind. Flat surfaces should be okay...I'll see about the curves.
He doesn't provide a complete step-by-step video but I find I learn better when I have to overcome challenges myself and really studying the info to understand what's going on though it does take longer, especially when trying to figure out why he's done it that way.
With the above example I was also having difficulty projecting the shapes onto the n-gon to have the edges connect where I wanted them to, but after some experimenting I found that when adding the loops for the connections, snapping to the vertex of the shape helps a lot.
Anyway, I prepared the images of the first week from the 2014 challenge when
a lot of the images have gone offline, but you can use archive.org and not
always but quite often it will have a copy of the images - the linked ones anyway. Good tool to
know about.
- I can't answer your question on the best order to go through them as just following along with the above took me a while.
Another great resource is Arrimus 3D's youtube playlist called 3ds max (I'm using blender and it's mostly quite easy to follow along). I'd start with the first 4 topology videos and skip along that playlist until he talks more about topology. He shows a really cool thought process to getting a circular object to blend smoothly into a square one in the second topology video. It's a very...very...very long playlist!
_
His request playlist is also awesome.
At this rate I'm not going to get to the fun stuff like sculpting till next year
Hey Frank, could you please tell me how you do the automatic chamfer in model number 3(middle left) here? Is it as simple as selecting to 2 edges in model 2(top right) and hitting the chamfer function in Blender(I'm assuming that is what you are using) and messing around with the edge count and falloff options? Thanks in advance!
@tatertots Thank you for taking the time to follow up your initial question with a detailed explanation of what you tried and what you were looking to improve. Glad you found the posts helpful.
Cylinder intersections is a good place to start and from there the v
block project covers three different approaches and the key project has a
lot of cylinder intersections. The 123 block project relies heavily on
Booleans and modifiers so it's probably best to leave that for last.
In most cases the underlying geometry and topology layout is largely identical to what's shown with an optimized subdivision preview. This can often be useful because it shows how the subdivision smoothing is deforming the underlying topology but on the other hand it can also hide some subtle nuances. The feedback on the usefulness of seeing the different steps Vs just seeing the starting and final topology is worth noting. Something I'll keep in mind.
A lot of artist have contributed to this thread over the years and it's a great resource. It's a shame that some of the earlier images have been lost but the pages from the last couple of years still have a lot of information. Lots of answers to commonly asked questions there.
You are correct: select the two outside edges and use either edge weights and a modifier or the manual tool to develop the round over along the outside of the shape profile. Working through the model in stages and relying on tools like bevel /
chamfer to add these kind of details helps keep everything accurate,
easy and simple. If the geometry needs support loops then they can be added by using a wider square chamfer operation before adding the round over or by using an inset operation on the flat area after the round over is added along the edges.
Replies
How would you increase the sharpness of the corners on the sphere?
I am having significant trouble modeling this door, especially the corner, is there a different method I should employ when adding my edge loops
You are making it too complicated. The door is not just one piece. Try to seperate it into its parts. The frame consists of 4 elements. The hole in the middle is another element. Break it down into its parts. This makes life much easier. A carpenter wouldnt try to create a door out of a single piece of wood.
Andrew Price's tutorials are zero in modeling.
can someone please help me at 3:10, Arrimus is using symmetry to do that but I couldn't find any similar solution inside Blender. I try to follow this tutorial with Blender but now I am stuck.
1. Set the cursor to the center of the sphere and add an empty.
2. Add a Mirror Modifier to the sphere, enable Bisect X and set the empty as the Mirror Object.
3. Rotate empty.
You can can stack multiple mirror modifiers targeting new empties to get extra bisections, and apply and symmetrize in Edit Mode to mirror the changes to the other axes.
I would start with booleans to create the base shape and clean it up. You have big flat surfaces here, don't be afraid to use ngons.
Here is a crude example, should be cleaner in some areas.
1. bool, clean up and add some edges to help the bevel modifier doing it's job.
2. bevel
3. subdiv
Your answers are always very insightful!
I've been bouncing all over the place looking for the fastest and the most intuitive workflow for modelling from concept to texture. Came all the way from CAD modelling from Fusion 360 to Subdivision and i'm still not able to properly dig in. It was becoming very counterproductive to keep bouncing between so many tools so i decided to stick with the basics of Sub in programs like maya and blender because i understood that adoption for cad based workflows is very rare among studios and also because understanding subD will be helpful for anyone in this industry.
I started watching andrew's twitch and took classes from Mario Brajdich who have extremely similar workflows. On the CAD side i'm always blown away by people like Alex Senechal and Edon Guraziu's intricate sci fi forms. All of them developed their own understanding and aesthetic of hard surface modelling based on their preferred toolkit. But in the end their design principles speak louder than the tools they use.
So i guess i should focus on just the design aspect of models first instead of the toolkit.
One last question. With Blender's adoption rate going up do you think it would be wise to rely on it's famous third party add ons like Boxcutter, Speedflow etc or are they just temporary solutions in an ever evolving field? They all feel extremely intuitive but also carry a lot of limitations. But at the same time they're so helpful during the concept phase. Just wondering if i should give into the temptation and buy them
from the ref photo you can go with symmetry too and collapse it all to work on unique areas.
- boolean doesn't generate quad topology, so still manual work.
This is my try
I started with this
I think it looks believable, but i would still want to make it like the manufacturer does it.
for the topology, i created the cylinders and i started rotating them into rings, and i inset them and manually and connected between them, about 85% quads, and i used 1/4 for symmetry anything less wouldn't work.
then there is this
It's hard to explain in words but im sure there is some math behind this
There's a point in which the holes of all rows are organized in a straight line, thus it's possible to use an array to evenly place the "origin" mesh of each row.
You'll need to create a circular array for each row, and how you go about it depends on the software used. To control the pattern change the number of copies in each row, and create the gaps where a row skips a hole by applying the array modifier and deleting the meshes where there should be no gap.
This is my attempt (without smooth preview):
It might not be so visible in the images but I am getting some strange pinching in certain places which leads me to believe my topology isn't ideal. How would you guys go about tackling a shape like this? Thanks.
"It may be helpful to start by analyzing the reference images to identify and categorize the object's surface shapes. Look for things like contiguous flat areas, continuous shape profiles and the transitional areas between each surface or shape. Visualize how the major surface features interact with and transition into the surrounding shapes. Plan out the basic topology flow needed to support these shapes and the transition areas"
This has been really helpful but as a subd noob I'm finding it difficult wrapping my head around why you've positioned your edge flow in the way that you have in your first image. I could just copy your topology and be on my way but I really want to understand how I can identify where edges need to flow regardless of what shape the surface is. Thanks
This is how I always ended up.
I've done it! thank you!
this little bumps are tricky to model, and make them look good with the superior part of the model.
First I tried modeling the superior part(I was having some problems with the plane shape, but I achieved to make it. But before that, I don't know how to create that shape in the lateral zone. I tried using booleans but it looks bad when connecting with the superior zone.
How do you think that could be the best way to approach this model?
take this, i hope this help u
Here's an example using only vertices with Skin & Subsurf for thickness + Screw for twisting. You'll need to do a quick reading on Skin in case you're going this route but are unfamiliar with it.
Here's an example that shows how matching the segments of the rounded fillet on the side to the segments of the round over on the top allows the support loop to flow around the shapes without causing any major smoothing artifacts.
I've spent quite a bit of time getting the following form correct:
I'm talking specifically about the black bolt in the middle. It seems like a normal round bolt but with a number of cuts made out and a slanting edge upwards, which ends in a curve. After trying many different approaches, this is the best I've managed to do so far:
The idea here was to make sure the supporting geometry for the vertical edge loops were there from the start. So they could be edge slided closer to the cut-out to make those edges sharp, without having the pinching effect they create seep through to the top of the screw. The result is the best I've managed so far. Some remarks:
It doesn't look to bad. And it its a very small part in the total render. But I would really like to know how to solve a problem like this "The Right Way" (TM).
Anybody could point me in the right direction here please?
Thanks in advance!
-Tom
While I have managed to reconstruct something that leads to similar results, it feels like a trial-and-error approach to get there. I have difficulty interpreting the screenshots you posted. You always post a sequence of three images. I understand the last version is the subdivided/smoothed version. However the step between the first and the second image is unclear. In the images you posted the first one of each series contains a pretty big ngon, which does pretty bad in subdivision. So I assume you start of with a simple version with ngons and then flesh out the geometry in step 2 and have the mesh in that step show the deformation of the subdivision modifier.
Getting from step 1 to step 2 however is a difficult process though. I don't quite get the steps to do this in a straightforward way where I understand why each thing is happening. It's probably a bit much to explain in a single post.
I found this resource here: http://wiki.polycount.com/wiki/Subdivision_Surface_Modeling
I'll look at the video's and examples in the above link. Do you happen to have any other resources to recommend for the theory behind topology when subdivision modelling?
I've gone a few more times through the process, and made some screenshots of the different steps I took and the reasoning between things. If you could look over it and correct me where I'm wrong?
From top-bottom, left-right:
This is also the reason we only need 4 faces (3 edge-loops) in the middle, because there's only 4 faces (purple lines) that run top to bottom.
I hope with the logic above - if correct - that it'll help create a mental model on good topology and how to reason about it. I think the next time I run into a tricky situation like this (though the result seems so eas
Thank you so much for this very detailed explanation. This was very clear. I've tried and tested all the different approaches you described and this has given me a whole new view on how to use the bevel/chamfer tools and the inset tool. Insetting n-gons to create edge loops around the details you wish. Using the bevel tool to sharpen up details on both sides of loops. Really great. And less of a fear of "the n-gon". Also, dissolving the edges in the middle is much faster than whatever I was doing... I deleted the faces and then manually started connecting the edges and faces again to make the flat. :P
Also the workflow(?) that just starts with the ngrons and then creates more agreeable geometry from that starting point will be helpful in many more cases I believe.
This is absolutely super valuable help, so thank you for taking time out of your day and helping out less experienced (=total noobie in my case) people.
Of the all the approaches, the third approach with the boolean suited me least of all. I found it very difficult to get the boolean well positioned and found the manual approach to be faster in my case and more accurate.
And thanks to Yogifi for mentioning the pdf tutorials, got curious and found them. I'll buy them as well. Any good information on topology and workflow is very helpful right now. I'll also be looking to some of the older pages of this thread. Around 150 the images still mostly seem to work. Though the issue is often very similar where a subdivided mesh is shown (optimal display) and there's no clear idea of what the underlying topology actually looks like.
For instance, the first picture in your previous post, showing the 3 different cylinders. I have no clue how the topolgy of those meshes relates to the methods and images with the descriptions of the 3 methods later in your post. It is hard to find the resemblance.
Again, thank you so much for taking your time to answer these questions.
EDIT:
Quick additional question. Is there any particular order you would recommend for your 4 pdf's?