Tuesday, 12 March 2013

Interface Design


Spent this morning working on some of the more superficial design elements of the interface - above in the dark maroon I've have highlighted potential candidates for the interface logo. I then made a study of potential interface windowletts.
Another issue for consideration is partner branding - I reworked the logos below as flash vector objects to conserve memory load and improve the fidelity - are DCAL & the Nerve Centre Brands needed too?


Monday, 11 March 2013

Update

I've updated my siege era Bishops gate reference model to alignment with another reference I've been sent from Mark, once we've agreed upon it's accuracy I can then commit the design to ink and integrate an illustration into the timeline


Friday, 8 March 2013

Interface Test 02

Update

To help with the reconstruction of the Siege era design of Bishops gate I've been looking at a variety of sources and focusing on 16th/17th era fortification design - from this I've created a 3d mock up to use as reference for the sketched renderings  -  I need to confirm with Mark Lusby & his associates the accuracy - below fig a. is one of the widely used images pertaining to the siege which I have been notified is inaccuate so I've been looking at proxy fortified settlements of the era and contemporary renditions from the period such in fig b.

fig a.
 fig b.
 

Monday, 4 March 2013

Progress Update

Having spent last week on the basic code framework I'd use, Anthony got in touch with me over the weekend with the scan data, the point data is huge! Initially I was working with the dataset in the freeware program MeshLab.
 
 Meshlab isnt massively stable but considering the file sizes this is understandable, it also wasnt supporting vertex colouring on the point cloud. To counter this I've been familiarising myself with point clout to mesh workflow.

A massive godsend has been the process of sampling - though which a cross section of that data can be sampled to create a decimated model, this makes it easier to view the data, but equally this means a loss of quality, so for qualitive purposes I shall cut up the cloud -> convert to a mesh export then decimate and then generate the textures/normal maps in Zbrush.



 I have since got my hands on Geomagic which is a v. impressive program, also it does mean I can use the original scan vertex colours!

 

Friday, 1 March 2013

[3] 3d loader Issues

Have encountered a number of issues with trying to get the 2d menus to work in 3d space en scene, also both the flash papervision and away3d platforms seems to have limitations as far the polycount/ detail of 3d models on screen.

Through my research I've came across two alternative platforms, webGL and unity - for the purposes of the project, creating the interface within Flash but using webGL or P3d for the 3d elements, longer term creating the interface entirely within webGL/ HTML 5 might be a more efficent approach.

 p3d is an external service provider much like youtube but using a webgl viewer means it can be used locally/offline ---------------------------------------------------------------------------------------------------------
http://www.jaccogoris.com/ViewerDownload.php
http://cgeverything.co.uk/2012/10/17/verold-webgl-viewer/
---------------------------------------------------------------------------------------------------------