Colossus - some 3D models

thread on tubes

here are my first 3D models for the Colossus project
still need some work but getting there
may ahve to do some lower res tubes for doing whole bay with 100’s of tubes!

not certain yet of all the real tubes that ahve been use on the rebuilt Colossus
some might be Metal or Glass tube

also did the photocell Black mask

thanks
happycl

Attachments



1 Like

did a quick test for how amny tubes we can put on a bay

so use instancing and got 420 tubes with about 4 K verts each
for a total around 16. millions verts on that bay

and no problems to work with very fast in viewport
and I got only a small machine

so I think should be possible to make one bay back and front,

not certain if possible for all the bays !

thanks
happy bl

Attachments


here is a test done with PBR bluish glass and HDRI

thanks
happy cl

Attachments


You could cut down on the verts by using group instancing for those parts. Make that tube into its own group and use group instance. The down side is that you can’t array group instances, but the upside is that they don’t count as verts, edges or polys, only as objects. Or you can make the panels into groups and instance them as you need.

that might be another good idea
will test that later this week

thanks
happy bl

block diagram

happy cl

Attachments


Duplicate group instances using additional geometry - subdivided plane e.g. - which you set to Duplicate from Faces.
Downside - you would increase vertex count by some verts and would need to manage positioning.

The tubes look great and I dig the shaders, especially of the glass tubes! :slight_smile:

Like you I start to worry a bit about the polycount of the final model. Instancing would reduce the load of course and dupli faces or dupli verts would be an option, like ajm and eppo said.

Maybe the vertex count of each of the tubes themselves could be reduced a bit? 4K verts is more than I expected.

But I know that you spoke about low poly versions yourself.

for long camera shot I mean don’t really need most of the inside verts
so low poly that would bring down the verts count a lot less then 40 % !

but using group instance it does not really matter it takes no verts at all !

happy cl

some of the basic circuits for COLOSSUS
I was unable to read all the symbols !

thanks
happy cl

Attachments



other circuits

thanks
happy cl

Attachments


Biquinary counter

Attachments



here is the encoding for the Lorenz cipher machine

happy cl

Attachments


These charts are looking great, Ricky. It appears you are more of an engineer than I am :slight_smile:

Where did you find all these information? Are there any publications by the rebuilding team?

I did an approximation for a Lorenz wheel
still missing how the inside is made up !

the blue pins are used to activate numbers

happy cl

Attachments



here is first test for the paper reader photocells

still need the real model !

happy cl

Attachments


Looking good! The photocells were needed to read the perforated tape from the bedstead, right?

yes but not certain what model was used in the rebuilt !
that will be one question to ask !

thanks
happy cl

here is beginning of one small panel with switches

happy cl

Attachments


This is incredible work you’ve done! Those tubes, dials, and spindles look very real.