Same here. I just hope that there was a strong negative response to their questionnaire and have decided to leave things as they are for now.

So, it won't be long until we find out if Trimble have decided to shift to a subscription model.

Very nice. Reminds me of Jürgen Mayer's residential work.

.rar is just a compression format like zip.

I think the model is deceptively detailed...it only took a couple of minutes to create using the Line tool, Offset and the Split Tools plugin.

For more complex forms I use Quad Face Tool, Vertex Tools, Split Tools and Zorro extensively. Solid Inspector is also handy for finding unwanted faces and internal geometry.

If you do actually need a completely smooth transition, then you'll have to use SubD or Artisan.

Here's a quick basic shape to play with.

Wedge.skp

I agree, turning them on and off would be ideal; but since that isn't possible, you can't avoid destructive bevelling if you want that extra level of realism.

As for last minute changes, I don't run the final renders until the client has fully signed off on the design and they are aware that no major revisions are possible. Up to that point I present draft renders without the bevelling.

@jql said:

I remember Thomthom was talking with both Thomas (Jiminy) and Tomasz to make it happen. Maybe it's being developed as we speak, maybe it stalled.

Yes, some Skatter wisdom would probably help to move things in that direction.

@jql said:

Anyway what I wanted was simply to have raw geometry without bevels, so I could design easily with sketchup and some method to fake bevels so I could render realistically, but I don't want to have permanent bevels.

The way I envisioned that working would involve a new version of Round Corner that adds the loops needed for SubD bevelling and then assigning a SubD @ Render condition to the geometry.

Whilst it wouldn't reduce the geometry/calculations rendered, it would keep models manageable and be partly non-destructive.

Naturally the ideal would be completely non-destructive bevels, but that's not going to happen without major changes to SUp's core engine.

Personally I leave all bevelling until the last minute, when I know there won't be any design changes. Sometimes I'll save non-bevelled versions of components in my project file - just in case.

I do this sort of thing manually on a daily basis and it is frustrating that a number of plugins come very close to automating much of the process.

A quad version of Round Corner with a subdivision mode (just adding the required loops) would be awesome.

Having more control over the addition of loops in Quad Face Tools would be very helpful. I seem to recall that this is planned for the next version.

For subdivision at render, I think that could be possible with SubD at some point in the future. I imagine that getting it to play nice with all the different render engines will be quite a big ask.

Something like this might help with the conversion:

Loading...

(www.360toolkit.co)

Output options for panoramic and stereoscopic 2D images and animations.

Obviously we can do this using various render engines, or exporting to something like Unreal. However the option for a quick and dirty SUp pre-viz VR or panoramic would be incredibly useful.

@rich o brien said:

I made the liquid of the beer penetrate the glass.

If you mean that the inside surface of the glass has the beer material assigned, then that's the right way to do it. You do however need to then set the beer material's container to the glass material. Again, you can only do this in Studio.

You would then use a separate beer material for the surface and that will have the Global Medium as it's container.

I've got an illustration somewhere that explains all this, but can't seem to find it right now.

@rich o brien said:

The bubbles are little Chrome balls because I could not get Thea to render bubbles properly.

You need to set the beer as the Container for the bubbles material (I use the air from the glass of water example scene). Unfortunately you can only do this in Studio.

The poly count is complete overkill, but it's the only satisfactory alternative at the moment.

In theory, you could remove many of the loops with QFT, delete the faces and re-skin the frame, but that's a lot of work for something that should be simple.

Trying out BoolTools2 is n my to-do list.

As great as Round Corners is, you are right about it's limitations. It's the main reason I use SubD for filleting objects that aren't rectilinear or have complex junctions.

If you do find anyone willing to take Round Corners to the next level, how about starting a Kickstarter campaign to fund the exercise? I'd happily contribute to something like that.

@rich o brien said:

You're beyond help David....

Tell me about it

@thomthom said:

1000 polys will be an interesting restriction which should produce some interesting solutions to optimise the mesh.

That's me screwed then. I have a serious addiction to overly complex control meshes. I should probably seek professional help.

Looks like the update is being rolled out directly to SU now - I've just had a prompt on launch.

That second video just goes to show how misleading the overview video was. Workflow doesn't look much faster than using SU with SubD.

This looks a lot like the Morris Minor model:

https://3dwarehouse.sketchup.com/model.html?id=u1418a3cc-2432-4776-a76f-2ac696b7fa67