Posts made by ledisnomad

This is really helpful. Thanks Free Agent for your process. To summarize so far...

Possible entourage workflows:
I. Eye-level
a. Post-production. Web resources include:
i. vyonyx.com
ii. archiforge.com
iii. kropped.com
b. 2D silhouettes as reference (and shadows), then post-production
c. 2D photo face-me components (potentially with long-exposure style blur)
d. NOT low-poly 3D
II. Overhead
a. Post-production
b. Low-poly 3d people if distant (NOT 2D people)

Possible vegetation workflows:
I. Post-production
II. A few 3D for lighting reference, then post-production
III. 2D face-me components as reference and shadows, then post-production

Free Agent, could you explain the shadow trick in a little more detail? If you apply a 100% transparent material in SU, doesn't that affect Vray's (or other renderer's) material?

It's not a one-size-fits-all solution (doesn't work for overhead at all), but I like the long exposure style photographs where the people are blurred. I've had good success with this technique because it requires very little accurate detail on the people. I use some of the cut-out photo people from the 3D Warehouse, copy them onto a new layer in Photoshop and use the motion blur filter. Like I said, not a universal solution.

Speaking of no universal solution, maybe a list of techniques and/or do's and don'ts for different situations would be helpful. Here's a very quick start. Please edit:

I. Eye-level:
a. Close-up
- NOT low-poly 3D
- Long-exposure style blurred 2D photo people
b. Distant
- 2D photo people (there are some in the 3D Warehouse)
- 2D sillouettes as reference to post-pro
- Low-poly 3D
c. Crowded
- 2D sillouettes as reference to post-pro
- Post-pro in foreground, low-poly 3D in background?
d. Sparse
- Very, very careful Photoshop?
II. Overhead:
- Low-poly 3D people
- NOT 2D people
- Post-pro only may be too time consuming?

Although it's not a plugin for SketchUp, so therefore may not be what you're looking for, but... I've used a program called Pepakura Designer (US$38), typically used for paper craft, to unfold or even just organize a cut sheet.

Cool idea! When I have a moment, I'd like to be part of this experiment... maybe over this weekend. Have you heard of Studio Wikitecture? http://studiowikitecture.wordpress.com/ They set up an interesting system within Second Life that allowed controlled collaboration, voting, and more. Perhaps it might serve as a model for this project, at least at a basic level.

Hey, Tobobo.

Interesting idea to work from a poem. Is it an assignment or did you come up with the idea yourself?

Three thoughts I have after reading the poem and looking at your renderings:

1. The poem is quite dark and cynical, right? I think you could portray that in the renderings with some moodier lighting.
2. Have you seen the paper relief models that Zaha Hadid maked/s? It could be interesting to combine the flatness of shadowboxes with a flat-ish architetural representation. Even just scale everything to flatten it and play with the field of view.
3. The stair are the most architectural element and could use some development. The texture on the treads doesn't seem right. And the form could either be more real (stringers holding thin treads) or conceptual like a physical model (a little overlap of each stair like they are glued or just floating treads like it is purely computer graphics without gravity).

Look forward to seeing where this goes.

Oh look at that... Dave R beat me to the punch. Always so quick here at SketchUcation.

I'm quite familiar with this problem. My solution is certainly not unique or instantaneous, but it doesn't require a plugin and doesn't require zooming in and out madly:

1. Make as many continuous, joined polylines. This assures minimum gaps. A few things I've found helpful: With nothing selected, enter pedit command, then "m" for multiple (personally, I didn't know about this option until recently), select everything, convert to a polyline, then choose join. Then you can check what joined and what didn't. Anything that should be joined, use fillet with 0 radius to join. The more time you spend in AutoCAD, the easier it is in SketchUp.

2. After I import and use MakeFaces, I start connecting edges with diagonals. Anywhere that isn't a face that should be probably has a gap or there is a tiny extra line somewhere that SU has problems with. Connecting diagonals will often may a face on half of the problem area, the second diagonal will fill in half the remaining area, then you can keep narrowing it down to the problem spot. This is pretty quick, really.

That's the best method sans plugin that I know. Hope it helps. And if I've been unclear about anything, feel free to ask for clarifications.

This is great. Thanks everybody! Definitely a start that I can work with. At the moment, I don't need to worry about scaled instances of components, so I'll be able to use this and expand it's functionality as I need it. Thank you everybody.

The select all with same material only works in the current context, not in components and groups, but thanks for the suggestion.

I'm always amazed at how quickly questions and requests get answered here at SketchUcation!

This script, modified a little, is serving me well as a way to get the areas of my master plan models. However, not being familiar with Ruby, yet, I wonder what it would take to do the following: sum the area each material.

To put it another way, I'd like a report that reads:
There is 460.33sf of material 01.
There is 230.15sf of material 02.
etc.

And it would have to work with groups and components and exclude back-faces... that's the part that makes me hesitate to try to do it myself, not knowing Ruby, yet. What do y'all think?

Alienware is a subsidiary of Dell. I don't know if it really makes a difference, though, one way or the other. If you like Dell, maybe you get their customer service when you buy an Alienware computer. Don't know. Good luck with your purchase.

I'm no expert, but I try to keep up on exactly what you're looking for: bang-for-the-buck hardware. My office just bought a new laptop for presentations, but it needed to be robust enough to support real-time fly-through and design charrettes. We got an Alienware laptop from Dell. Although the Alienware is great, the 15" uncustomized model is $1200, perhaps a little more than your budget. However, the uncustomized Dell XPS 17" with similar power, maybe even a little more, is listed as $949. The only thing I'd upgrade on that is from Windows 7 Home to Pro for $130, so $1079, just over your budget.

OK, Vray folks, I'd love help figuring out how to set up a scalable, cost-effective render farm. thomthom has had some good suggestions scattered about the forum, but he suggested I start a new topic to brainstorm. Here are some starting questions/topics:

1. Hardware vs. Finances: What is the best bang-for-your buck hardware?
2. Legal: How many nodes are you allowed to install using a single Vray license?
3. Financial vs. Design Process: In this economy, what is a good balance between cost and added render speed?

A few thoughts I've read and will repeat here:

• It may help cut costs to use all-in-one motherboards with video and network on-board, since the quality of those items are much less important.
• The Intel i7 950 may be a good bang-for-your-buck CPU. What do you think? Can we do better?
• Does doubling the number of cores always halve the render time? The whole core-to-speed calculation is important to optimizing my setup economically because if I typically keep my renderings under 2 hours and let's say a 15 minute rendering is plenty fast, then it would be a waste of money to get more than 32 cores... which is a bit of a silly calculation because I won't have the money to get 8 quad core CPUs. But you get the point.
• Are multi-threaded processers financially helpful?

So tell me what you think.

Hey, dybi. Thanks for the response. I suppose I could make it a material, but I'd really like to model it as separate surfaces for each stone. This is just a small part of the pattern and it never repeats, so a texture map doesn't seem like the best way to go. I also need to render it, so I'm going to have to apply actual materials to each stone.

Here's the image of the DWG file.

I have a large, random pattern that I've developed and drawn in Autocad as a floor pattern (three types of stone, so you understand the context). See the image and the dwg, both small portions of the whole thing. I need to bring this in to SketchUp to model and render, but don't want to redraw it. Does anybody have a way to bring in this kind of thing? I have the MakeFaces plugin, but I don't want to have to repaint everything. Thoughts? Suggestions? Thanks.

Drawing1.dwg

The user would chose the materials to consolidate. When I orginally posted, I wasn't even considering an automated solution to determine duplicates. TIG's plugin works wonderfully for that and is very helpful. I guess what I was originally thinking was an updated/new Global Material Changer plugin to select multiple materials and consolidate them.

Here's how I did it. The proportions are quite right, yet, but I did it quickly as a proof of concept: no plugins and only a few minutes since my last post.

lamp_top.skp

Hey anstvam,

I don't think you need a plugin. I'd make the general shape as with the follow-me tool (circle as a path, squished semi-circle as a profile), then unsmooth all the lines. Then you can pick a segment and trace a single, curved element, mirror it on that segment, then rotate array it around. Making the curved element will take a few minutes, but no longer than finding the right plugin and using that.

Cheers,
Damon

I just got back and found all these responses to my question/request. Fantastic! TIG, thank you so much for the Material Consolidator plugin! It works beautifully and rids me of all duplicate materials!

Now... how difficult would it be to take this one step further? In addition to all the duplicates I get, I also get SIMILAR materials that should be consolidated. For example: "Glass-blue", "Glass-blue1", "Glass-blue2", etc. are all meant to a glass material in my model. Different people in my company tweaked them to suit their own asthetic so they have different properties. I'd love to be able to manually choose these materials and consolidate them.

What to you think? Easy? Hard?