Search This Blog

Thursday, March 13, 2008

Where are the photos?

Photos of Anchor Stone construction look much better than the CAD images, so why do I use CAD images?

To be honest, it's easier for me to create a CAD image of these models than to take a photo. The purpose of this blog is to share the designs with others, and I felt the CAD images do a better job at presentation. As you browse these images, know that the actual construction look's better (at least I think so).

So why do I think CAD images are easier?

All these images are created using a free CAD program called AnkerCAD. It's a great little program. If you care to download it and give it a try, I highly recommend getting it from Burkhard Shultz's web site. He created all the block definitions used in this program and his website contains the latest version. The site is in German, but you can grab English instructions.


NoKas said...

I am interested in finding out more about how you create the renderings you use for the models.

I just started using AnchorCad and have begun to reproduce some of your renderings. I find AnchorCad very awkward to use and have not yet found an easy way to build up the models. I find that in some views the stones don't line up or pop behind others. I generally have to go into the "top view" to make stone alignment work properly. I'll keep experimenting. Unfortunately, documentation is scarce.

I would like to know how you assemble your layer and "build up" diagrams. Do you lay them all out in one AnchorCad model or do you assemble them using another tool?

Also, how do you do the animation?

If you could share some of your techniques on the blog it would be much appreciated.

Thanks. ;-norm

releppes said...

Yeah sure, no problem.

I agree, the lack of documentation is a deterrent to using AnkerCAD. But hopefully I can clear a few things up.

One of the first tricks I picked up on is holding the shift key and right mouse button at the same time. It's a quick way to rotate a model around.

As for creating a model, you're exactly right. I use the top down view for everything. The only time I switch to another view is to check my work. But when laying blocks, I always use the top down view. Build the complete model first, then save.

I usually take a snapshot of the model at this point too. I switch my view to "Front right" and I select the camera "Save picture". Something else I do, which is complete optional, is that I select "Rendering options" and choose "Edge darking and set it to zero. I've experimented with all the image settings (*/A/2/4/8) and I found the lowest quality '*' works best for me.

Doing the layers is easier than you think. In the lower right corner of the AnkerCAD screen, you'll see a red box. It's call the "Cutting dialog". A coordinate window will pop up. I only use the Z coordinate when cutting. Leave the Low coordinate at '0' and set the High coordinate to '0' too. You'll see your model disappear. Don't worry, it's still there, you just can't see it. Now press the Up arrow next to the High box (remember, we're only working with the Z coordinate.) Click it about 4 times. When the number hits 80, you'll see the first layer of your model appear. Just keep clicking and you'll slowly see the model appear layer by layer. As I expose each layer, I click the '*' in my photo window and name the file 'a,b,c,...' until I expose the whole model. And that's it, I now have a PNG file for each layer of my model.

Now if you want to make building plans just like what's in the book, you have a little more work. In the Cutting dialog, instead of using the up arrow in the High box of the Z coordinate, you'd use the Both arrows to "only" show the blocks within a specific range (ie: cutting top and bottom). As you expose new layers, you'll want to "save the model" instead of just taking a photo. When saving the model, use a new filename and choose "Displayed stones only". Then you need to doctor up each file you saved by marking blocks with hash symbols and the like. I've done this before and it's just more work and tedious. It's why I've become lazy in making my building plans. I just take a photo as I reveal each layer.

Now comes the post processing, and it's optional. When you take photos in AnkerCAD, the background is always white. I like to make it transparent so it shows better on the screen. I use Gimp for this. I open the image in Gimp and do a select by color. I click on the background, then I choose color to alpha. That makes the background transparent. As a last step, I do an auto crop of the image. Then I save the file and I'm done.

As for doing to layer plans, using Gimp again, I create a new image (File->New). Make it huge (1600x1200) to give yourself some room to work. Then in that new image window, I do a File->"Open as Layers". I select all my layer images at once. Then I just move them around the screen until things look good. I flatten the image, make background transparent, auto crop, then save.

The animation, in Gimp, is easier than creating the layer plans. Do the same steps as above but don't move the images around the screen. Just leave them on top of each other in the center of the screen. Do not flatten the image. Keep all the layers as they are. Each layer will become a frame in the animation. Choose Filters->Animation->Optimize(GIF). And you're done. Do a file save and be sure to put .gif as an extension. During the save process, select animation. You'll also have an opportunity to set the frame rate. I just go with 1000ms.

As for making the animation work with Blogger, I need to do a little HTML hacking. If you're curious about that, just ask.

So there you go, my process for creating a model. It can be a lot of work, but I hope it's appreciated. Sometimes I get lazy too (hence the way I do layer plans). So enjoy! I hope my instructions didn't scare you off.

And by all means, if you want my models, just ask. I can email you the AnkerCAD files if you'd like. In fact, if I had an easy way to post them on the blog, I'd do that too.

NoKas said...

Your comment on modeling is very much appreciated. I have many questions but will wait until I try some things out. I would love to see your models and hear more about your html "hacking". For now let me say a quick thanks. Stay tuned!


releppes said...

To make it quick, the HTML hacking isn't much. It's more a trick I discovered while trying to get a GIF animation to work on a blog posting.

When you create a new blog post and choose to upload an image, switch to HTML view. You'll see one big long anchor tag. In the tag, you'll see an href and src attribute. Both attributes appear identical in that they point to the same location. Identical except for one small difference. The href attribute contains an 's1600' in it's path, and the src attribute contains an 's200' in it's path. Change the 's200' to 's1600' in the src attribute and you're all done. The GIF image should now animate when you view it on the blog.

I found this trick via trial and error. If Google moves things around, I'm sure this trick would break, but it works for now.