Sunday, January 21, 2018


I'm going to pause the 4 part Bank of England overview series to describe this weekends preparations for the next phase of collaboration. We are targeting the Waiting Rooms in the Director's Parlours, part of Soane's N.W. Extension. Three of us involved in the work at present

I am packaging up tasks into manageable, short, Family Editor exercises.  You get a starter file and a set of reference images.  The task involves increasing the Level of Detail.

There is a super-family in there that includes the vaults of the "starfish ceiling", the fireplace, skirting (base) board and a very simple family to represent the decorative arch that frames the barrel vaults at either side of the main space.  The fireplace needs quite a lot of work, and there's a chair rail to be added too.

I have taken ownership of all the door families for the moment because they follow quite a complex nested system of my own devising.  I spent some time enlarging and refining this door based on the reference material.  Needs some mouldings around the panels, but otherwise it's pretty much there.

The window family needs to be fleshed out with mullions and glazing bars.  Fairly good reference images for that.  The lantern at the centre, atop the vaults also needs an update and there's a survey drawing that provides a good reference for that.

The photos also show furniture and paintings which it would be nice to add.  So this is going to be an interesting experiment.  Two keen volunteers at present, so if you want to contribute you'd better jump in fast.

Meanwhile I will be working hard to extend this kind of exercise to adjacent spaces. So lots of scope for more people to join in the fun. 

Essentially, the aim is to bring more parts of the model up to the kind of detail level that Russell and Alberto achieved in the original modelling competition.  While I was putting my energies into understanding the way that the whole composition of the bank had developed over time they focused on the two major banking hall spaces that were part of that exercise.  I did also model the exterior screen wall up to quite a good level of detail, but most of my work is still quite roughly massed out.

With that in mind I decided to revisit the Consols Transfer Office model.  Actually I wanted to plunder Russell's work for some specific details that have parallels the Waiting Rooms: coffers on a curved surface, and rows of spherical beads.  So I extracted these families and put them into our reference database.

While I was doing that I decided to extract the two mesh families that Russell created: Lion's Head and Caryatid.  I wanted to try applying the edge-hiding technique that I described back in 2016


It turns out that his meshes are too complex to allow for the DXF export that this technique uses, so I hunted around for a way of reducing the polygon count.  This led my to install MeshLab, which is open-source software that I have never used before, but well enough documented for me to stumble through a successful simplification process.

I used 3dsMax 2017 to do the hiding, and it turns out that the interface has changed a bit. I wish I was an accomplished Max user, but sadly I'm a complete novice.  Still I managed to figure out the changes and jump back and forth between 4 programmes to create two new families, which turn out to be much smaller than Russell's originals (but lacking the fine detail of course) 

I'm not so worried about the detail because in any case these models differ from Soane's designs in a number of ways.  In other words they are placeholders that convey the design intent, which is what BIM is for the most part anyway.  Keep it Simple Sunshine, Less is More.

I really have no idea what Clustering Decimation is, or whether it's the best technique to use.  I just read a few sentences in the help, scrolled down a long list until I saw something that looked promising and decided the result was good enough.

There's something fascinating about the way that software and the human brain interact. On one level it's very mechanical, but on another level the intuitive leaps of faith are also really vital.

I do wish the lion's eyes were a little more sharply defined.  Actually I've seen some better heads on the web, but haven't figured out how to download them yet.  I cheated in front elevation by adding some symbolic lines, and the shadows help in 3d views, but you have to agree that the stone-carver's version is far more artistically satisfying than the digital modeller's.

Russell had disguised the mesh edges by setting the subcategory to a very pale grey, which also means that the outer contour loses definition.  So all round I'm quite proud of my new lightweight versions.  It would be great to find mesh sources for some of the other sculptural elements that we need.  Or maybe someone with mesh editing skills wants to have a go.

While thinking up a title for this post I discovered that my first two choices had already been used up.  These two posts from April and November 2016 reminded how far the model has come, working in bursts of enthusiasm punctuated by quieter periods.  


is the first post, but in fact it's mostly about the Waiting Room Court.  The Director's Parlours, where these two small waiting rooms are situated was just an amorphous flat-roofed space at that stage.  I hadn't yet acquired the survey drawings that helped me to understand the complexity of the roofscape with its multiple lanterns.


records that process of discovery.  When I first built the super-family with the starfish ceiling (although I wasn't yet using that terminology)  This post reminds me that although it all seems so simple and obvious now, I really was stumbling about in the dark for the longest time.

Come stumble with me.  .

No comments:

Post a Comment

I've been getting a lot of spam so had to tighten up comments permissions. Sorry for any inconvenience. I do like to hear from real people