http://en.wikipedia.org/wiki/Lamborghini_Revent%C3%B3n

$1,000,000.00 buys you a new look for your car's dashboard -
                                                                  or, you can print this out and glue it onto your Pinto.

I posted a few days ago about the Zygote/3D Science Dataset that arrived the other day. The quality of their data is quite amazing.
Here's a small sample of their offering rendered in Windows Presentation Foundation (WPF).

Muscular System      Skull   Female Organs    Lymphatic System    Circulatory System    Male Skeleton

I thought this was a good example of context (overview) and detail with text.

http://www.fibredesign.co.uk/index.php?/case_study/motorola_75th_anniversary_poster/

As people who follow this blog know, I've been working on an overall idea that marries 3D models and 2D images, documents, data visualizations, etc. into one big thing.  A major part of this is is factoring things like tagging and annotation across the *overall* capability, instead of just doing it in all the individual tools (like it's done now)...
Here's an example of using a ScreenSpaceLine3d class instance with the 3dsreader...next up: tying the right side of the annotation to a point that can anchor linkages to any kind of element.

ScreenSpaceLines3D mline = new ScreenSpaceLines3D();

mline.Color = Colors.Black;

mline.Thickness = 1.5;

Point3D onjectCenter;

onjectCenter = _mainReader3ds.Transformer.GetObjectCenter(_mainModel3DGroup);

Point3D start = new Point3D(onjectCenter.X, onjectCenter.Y, onjectCenter.Z);

Point3D ender = new Point3D(100,100,40);

mline.Points.Add(start);

mline.Points.Add(ender);

Viewport1.Children.Add(mline);

Here are some images I built today using the great library available from WPF Graphics. It renders 3ds files in WPF 3D viewports. The source data is from www.zygote.com, specifically, the Male/Female Premier Anatomy Collection, which we recently purchased. The quality of the datasets and the texture maps really speak for themselves.

fastcompany.tv/video/exclusive-first-look-worldwide-telescope

http://naimoli.spaces.live.com/Blog/cns!AD540FB65800D945!165.entry

Periodically on this blog I will comment about how bad the default colors are in most BI/Visualization Tools.  The easiest target is, of course, Excel, but Reporting Services and other tools have the same gaudy, default color schemes: fire engine red, grass green, mustard yellow, royal blue, etc. 

Is color important? Can it actually help readability of data?

The answer is yes, but this can be a bit subjective.  I include two images here from my friend, Michael Peters, a visualization expert working at Oculus in Toronto.

These are visualizations of financial data across something like 610 columns and hundreds of rows.  The data undulates, fluctuates and bends at various points across time in a rather - OK, I'll go ahead and say it- it's beautiful - it's pleasing to the eye. Did this happen by accident?  Nope - it's very deliberate and I'm guessing here that getting those colors to look like that - beautiful to the eye - probably took more time to achieve than the rest of the visual combined.

Next time you build a visual, take some time and work on the colors, don't just take the defaults - those extra minutes can make a big difference.

Visit Oculus.

Recently, one of my colleagues (Gabor Fari) sent me a bit of a challenge.  Displaying Metabolic Pathways in a Resolution Independent manner.  Naturally, I immediately thought of WPF, but Silverlight would also be a good choice for this particular visualization.

Gabor sent me this link: http://www.g-language.org/GenomeProjector/# which is basically the Google map engine applied to massive scanned images.  You could show anything with this technology that is image based: architectural drawings, floor plans, even artwork.

Here's what the link generates:

I really like the bookmark concept (which is really just masking x,y coordinates with an English description). There's also a zoomed map on the right (cut-off on this view - sorry about that) that provides classic Overview & Detail based awareness.

So, I went to the source of the scanned data (yes, someone really drew this map - regardless of whether it was a tool specifically designed for Metabolic mapping, or something like Visio, you have to give the author credit - I'm sure they are wearing bi-focal's now).

Here's the source: http://www.expasy.ch/cgi-bin/show_thumbnails.pl

Each panel is a separate .png image that can be tiled out. So I downloaded each tile (each one has a tile coordinate based scheme - like tileA10.png - like a Battleship game layout).

Then I created a WPF canvas and tiled them across - generating images dynamically and mapping them across x and y.  Each image is 720 by 600 pixels, so my images are 72 x 60 - more on this in a second.

Here's the first show - the overall canvas is blue and you can just make out the tiling in between the tiles.

Now, to zoom in on the Canvas...

and again,

and since we are here, let's ink on the tiles using an InkCanvas overlay...

I'm particularly happy with the readability of each tile - amazing work by the WPF guys in sampling and scaling these things.

This particular challenge only took about 30 minutes to build, but more to the point, adding things like annotation, connecting it to literature, tagging, breaking out each tile to work with them independently--you get it all for free.

I'm still working on it, but wanted to post this out...

http://www.nytimes.com/interactive/2008/02/23/movies/20080223_REVENUE_GRAPHIC.html#