That was really helpful on many levels, so thanks for responding in such a nuanced way.
My question about “workflow” was not only about the order in which you do things, but also “If I knew what Sketchup was best at, I’d know where to start and what to expect.” It was my impression that after designing a structure (in my case), I could transfer the model over into Layout and end up with a stack of 2D plans by just clicking a couple of buttons. Not so easy, apparently. Because I found no way to do that (the minimal interface craze of the last few years can leave a lot undiscovered), I reasoned that the fault was in the haphazard way I was building my model in the first place. I hope to be able to use a more orthodox Sketchup-as-it-was-intended method to rebuild my model, and this forum has been very helpful for steering me in that direction. But I still don’t know if I’ll ever manage to figure out what the program’s designers had in mind without further input. (Yes, I’ve watched videos. But their projects are obviously never just like mine, and the makers of the videos don’t realize how stupid I am. Basic things like groups might never be mentioned at all because, after all, who wouldn’t know about something so basic?)
I’ve been working with Sketchup on and off for years, actually, but more like a toy. Now I’m trying to apply it in a serious manner, and it feels like trying to get a Nintendo to do astrophysics when all I know how to do is Angry Birds.
@Sdotkling
As I use sketchup daily for residential additions, I would recommend a different work flow entirely. If you try and learn sketchup followed by layout as well as learning about home construction, drafting, permitting, you will never get to your end result. I do applaud your wanting to learn these things, but if your goal is have an addition built, I would recommend finding a good architect or design / build group. Once you have a working concept of what is being built then I recommend you model it into sketchup so you can do walk-throughs. Hell reach out once you have CDs and I’m sure users in this forum would be willing to model the addition for you, if not for free then at least at a fair compensation.
No argument that an architect would be the proper way to do this. That’s what my wife says. But I think it’s too much fun to design my addition to not try. In fact, we have an architect, who is very sloooooowwww and apparently too busy with bigger projects. I thought I’d design something, get him to point out its obvious flaws, stamp it, and then hire a good contractor I know to carry it through. Life is too short!
I sort of like doing small projects sometimes, but they are somewhat frustrating as they require proportionally much more work than bigger ones. You can get a long way into designing a whole house for the amount of time it takes to take a garage or porch through all the required hoops and somersaults, at least with the bureaucracy that we here have to face.
I have transitioned to using Sketchup from decades of 3DS Max and using 2D vector programs like Freehand, Illustrator etc.
The main difference in operation I notice is the absence of the MEMORY of previously created shapes as isolated objects… ie a Rectangle remains a rectangle, and a cube remains a cube. It wouldn’t matter of another shape intersected it. But Sketchup primarily behaves in a sketchy way where everything is part of one great big creative much like working with several overlapping brush strokes in a bitmap program like Photoshop which blend into one another.
In other 3D or 2D vector programs when you revisit an earlier created shape it remains intact unless specifically ‘joined’ or a Boolean operation is used on them.
Having to get used to having everything merging together in one unifying blob will be something that would be part of the initial Sketchup learning curve… till you start GROUPING everything you need separate instinctively as you go along.
Of course, there would be the ‘normal’ learning curve for getting used to the Sketchup UI
But overall, I am rather pleased that I am able to achieve a design idea goal through Sketchup in a far faster and intuitive ‘sketchy’ way, than through using other more ‘formal’ 3D software.
You will find it essential to keep groups (or preferably components) as your ‘memory’ of previously created shapes. Without that, as you say, you get a connected network of edges and faces which can quickly become impossible to edit usefully.
And do try Layers - they control visibility of groups and components. Used with Scenes, and combined with a mixture of orthogonal (Parallel projection) views and Perspective views, they allow for a very varied presentation of different views and selections of your model.
You might find the video here helpful. I assisted in an engineering course in which students learned a bit of SketchUp. One example we made was creating a model of a Tumbleweed Tiny House. Here’s the starting video, and other videos followed
Great feedback Tommy - esp about starting w/ pencil (or marker) on trace. So much faster to get a plan and key elevations worked out by hand and then copy into SketchUp to work out the details. You can even import the drawing as a graphic onto an appropriately sized blank rectangle and then work on a transparent rectangle six inches above it to get the basic model rolling. (Painting it with the materials tool to a transparent glass color) For site plans I’ll import the survey on one layer then the GIS info above that with the topo lines and then work on a third layer to integrate all the elements to what I need as a site and landscape plan w/ bldg setbacks, cubic yards to be relocated, planting areas, key local plants to be conserved, etc.
Update: Thanks to all who responded. With the discovery of the importance of “groups” and “components,” I think I’ve taken a giant step toward controlling Sketchup. I’ve been rebuilding my architectural models a wall or section at a time, and now stuff doesn’t fly around and become off-axis with an errant touch of some tool.
Here’s another question: What’s with making sure an entity is a solid? I overlay two shapes that I’ve pulled into a 3D form, and try to combine them using Solid Tools, only to be told they are not solids at all. I must be doing something wrong, or missing a step. Does a solid only happen when you group all the facets of an object?
And how about “Intersect faces”? It then asks “with model” or “with selection,” but I can’t see any effect (or I’m told that no faces are available for intersection.)
Thanks again for showing me the path to Sketchup self-sufficiency.
A SketchUp solid must be a group or component. Loose geometry, no matter how it is arranged, won’t be considered to be a solid. But also a SketchUp “solid” has rules about how its edges and faces can be arranged. You can search help and the forum for any number of discussions about this. The simplest (though not perfect) version is that all the geometry must be connected and every edge must bound exactly two faces, no more and no less. That produces an air-tight surface that surrounds a non-zero volume.
There is a lot of room for confusion about how Intersect Faces works. Assuming you have read the help and viewed tutorials to get the basic idea, we could probably provide better focused help if you can upload a simple example skp file in which this happens and then the experts can tell you exactly what you are doing wrong.
To me the differences between intersect with selection, context, or model are relatively simple. In each case the fundamental operation is the same: the software considers some set of geometry and looks for intersections (where to non-coplanar faces cross through each other, where an edge passes through a non-coplanar face, etc.). New geometry is automatically created at planes of intersection. The difference between the three variants is how the set of geometry to be considered is chosen, and (amongst that set of geometry) which entities become eligible for new intersection geometry to be created.
With Selection is the most limited: only the geometry that is currently selected (with the Select tool) is considered. Intersections found amongst that geometry are used to create new geometry at the intersections.
With Context is more inclusive: it considers the geometry that is selected, along with all other geometry in the current context (which is typically an opened group or component, but would be everything in the model if no group or component is currently opened for editing). Intersections found between the selected geometry and any other part of the currently-open group or component are used to create new geometry at the intersections. Note that if there are intersections between entities in the current group or component that are not initially selected, no new intersection geometry is constructed for those theoretical intersections.
With Model is even more inclusive: it considers the geometry that is selected, along with all other geometry anywhere in the model (i.e., raw geometry at the top-level along with the contents of any groups, components, sub-groups, etc.). Intersections found between the selected geometry and any other part of the model are used to create new geometry at the intersections. Note that if there are intersections between entities anywhere in the model that are not initially selected, no new intersection geometry is constructed for those theoretical intersections.
I almost always use Intersect with Selection, just to have the most control over what happens (preventing SketchUp from “tampering” with entities I want left alone). This also improves performance because it causes the least amount of potential geometry to be considered.
The detail that causes confusion for many people is that intersect with always creates edges in the current edit context regardless of what is being considered for intersection. So, for example, if you select two groups in the model and do intersect with selection,the new edges will go in the model context, not into either of the groups. As a result they won’t cut any faces in either of the groups.
Does that mean if I am editing a component which includes other components,
the geometry will only go into the top-level component context, and not into any of the
sub-component contexts ?
So then one would need to find the “orphan” geometry and and either cut-paste-in-place into the sub-components, or edit each of the subcomponents and do a intersect with model for each one ?
The edges created by Intersect Faces With… always go into whatever is the currently open edit context (model, component, or group) regardless of what is chosen for the intersection. Nested groups or components are not considered to be “open” when you open their containing parent, otherwise they would not isolate their own contents. So if you want intersections inside them, you must get the new edges there one way or another.
If you really want to get in to the nitty gritty of SketchUp pro I suggest you read the book “SketchUp and Layout for Architecture” written by Matt Donley and Nick Sonders
