The Raymond Lemaire International Centre for Conservation (RLICC) at Leuven University, Belgium offers a masters in conservation and this includes documentation techniques. Almost uniquely RLICC offers practical as well as theoretical teaching in applied metric survey for architectural conservation. Archdoc is the practical training module in which students are required to conduct survey tasks by TST, laser scanner and photogrammetry. At the heart of the process is the understanding that record of condition and significance is key to meeting information need with metric tools. In the end this means generating drawings that meet a good architectural standard with adequate metric performance given the time and resource available.
From Cambridge to Leuven by train always pleases me, I didn’t mind the long wait at midnight at Brussels Zuid, my Eurostar was late and I missed the 2nd last train to Leuven.
I like to think the two cites are joined by rails even when the trains don’t run.
This year I was privileged once again to be a guest of the University to teach TST/ CAD survey, as ever the challenge is to translate a building, via 3D wire-frame from TheoLt into drawings in short order, a task which is never simple as the translation from reality to line drawing is a complex and personal process, year on year the tools and their facility improve but there is no avoiding the necessities of orientation, line selection and feature interpretation-always difficult for those who have not done surveying before.
I bring to Archdoc a pencil, a camera, a TST, a tablet with CAD and, hopefully, an experienced eye. The forecast is too awful to bring a kite and rig. The teaching resource comprises an elite international team of heritage documentation expertise equipped with laser scanner, thermal imaging camera, TST, a drone and some serious software.
The devil is in the detail.
Getting a grip on geometry is one thing, getting the delineation of key architectural detail is another, with a project scale of 1:50 we can expect to see the organic forms depicted on the final drawing in a stylistically recognisable way. Drawing 1:50 scale fruit and veg means abstracting the overall shapes and edges but avoiding the use of stipple or shade, in cartographic terms this is generalisation, the character needs to be shown but not the entirety. In this example things are made harder by the inaccessibility of the detail, they are 11m above floor level.
After worrying about this for a while and gently suggesting to the students they make a sketch I sensed the panic : this is no easy thing to draw and the ‘how do we know which edges to pick out and which to ignore?’ question loomed.
A ‘cartoon’ sketch is needed, this is not going to be a plate from Normand’s Parallels of the Orders. At 1:50 and with time tight the sketch is as stripped down as at can be. Measurements can come from the point cloud, photogrammetry or TST but the feature selection is best done by eye and hand: so let’s draw! I pick the easiest of the 2 capital types that will be in the section and dive in. 1st find the centre line, then the symmetry and gradually I can break down the apparent jumble of lines in to clear leaf, petal ‘bat wing’ and ‘feather’ shapes. 1744 Rococo/Baroque style is all about flowing opulence and that’s what I try to catch in the sketch. I end up with a clunky caricature, it’s no masterpiece but it will do and it gives the team a leg up in sketching the remaining detail: the fear of the pencil has gone.I try to avoid shade as I know I will simply depict with line in CAD but given the depth of the modelling here it’s not easy, taking the lead from my sketch the student sketch of the 2nd capital is perfect for the task, edges are defined and an approximation of the organic forms achieved. I love it when I see confidence build in students, drawing is often about knowing where to start and I pointed the way here- job done?
The next step is to spoon the thing into CAD. It’s hard to keep the flowing lines smooth but at least the selection is resolved. Traced in polyline and then splined, the trick is to make sure the sharp points are the ends of the lines. A continuous pline would ‘melt’ when splined. The symmetry can be exploited to mirror left and right hand details.
The CAD trace ready for fitting to the section drawing…
…2.5D fit to wireframe. In this case the model is purely a means to an end, a sectional elevation drawing. I use solid geometry to simply complete and close lines from the TST trace.
The precision is in the line work and the backfaces of the model are irrelevant. The traced capital and medailion details are backed by thin sheet surfaces to exploit the ‘hide’ viewstyle.
Architectural surveying is best done with the final drawing in mind at all stages of the process, the convention of plan, section and elevation are made real as you go. A live trace of the TST work is a real boon. This is what TheoLt is all about: surveying in CAD by TST confident that line selection is clear:
A TST trace can be revealing: an entirely new ceiling profile was plotted (superimposed over previous survey in red above) only a few points are needed, each point selected for its utility at capture. Previous survey, despite strong draughtsmanship, relied on estimation for ceiling and roof profiles.
A vintage ’50s diagonal eyepiece on a 21st century Leica telescope, Heinrich Wild‘s foresight never fails to amaze!
In the search for low cost solutions I used a Linx10 tablet PC for site CAD with TheoLt. This is a remarkable, if fragile, device: a Windows 10 touch screen tablet with a basic RAM and processor specification. I bought it along with my groceries from my local supermarket; something that would be unthinkable until the current boom in tablet devices. It’s cheap (€ 200) and runs for about 6 hours before it needs charging. It has drawbacks, the touch interface is tough in CAD and I needed to buy a micro SD card to expand its tiny memory but for wire-frame capture in CAD it works well, to save power (Bluetooth is available) I ran the TST by USB cable via a micro USB adaptor.
Having lived with it for a week (including hauling it along with a TST and a weeks worth of laundry on and off the trains of Europe) there are things I really like about it:
- Battery life
- Screen size
- Screen clarity: no problem seeing what you are tracing indoors.
- A4 form, fits in a small bag
- Light weight
- Light weight, generic 5v 2A charger
- Runs TheoLt & CAD
- Touch ‘pen’ cheap to replace (I lost one on the way home, they are 90p each)
- Generic slipcase easy to find for a 10″ tablet
and some I don’t:
- Funny power switch (press- hold- release to start awkward)
- Fragile, it feels like it could snap like a biscuit!
- Touch in CAD is a bit tricky
- Tricky to see the screen in direct sunlight
- Fiddly USB adaptor
- Tiny SSHD/RAM spec: can’t expect image processing.
- Seems to forget some settings if you run the battery flat
- Hibernate/sleep modes can be tricky
- Touch keyboard dosn’t always ‘snap’ to CAD screen (I think this is a Win 10 issue)
On balance the cost wins out, a Motion CL900 type is a much more durable device but costs almost ten times as much and is expensive to look after (replacing a failed battery is €200 alone). I’m investigating ways and means of fitting it securely on the TST tripod: the first step in preserving it. The Linx10 is already updated to Linx10.1 so I expect these devices to improve in time and, provided you don’t expect to much from them, are a great tool for the money. An alternative is the Microsoft Surface family which are much more powerful and a good device to run 3D CAD and image processing as well as TheoLt at around 5x the cost of the Linx.
With much thanks to Luigi Barazzetti, Mario Santana and all at RLICC.