Monday, February 29, 2016


Just for a change, nothing deep or historical, just some simple Revit tips. 

Have you ever had that thing where you want to right-click on the end point of a wall so you can do a "disallow join", but there's too much other stuff in the way and the wall just deselects.  This used to happen to me fairly regularly, but there's a very simple workaround.  Simply select the whole wall and uses the sunglasses to temporarily "isolate element".  Now that there's nothing else to interfere, disallow join will work just fine.  Reset the sunglasses, and you're done.

While we have the sunglasses handy, try this one.  Select a bunch of stuff in a 3d view.  Isolate elements.  Deselect.  In the properties of the view, activate the Section Box.  Reset the sunglasses.  This is so easy to do, it makes me wonder why people went to so much trouble to write a special command.  The great thing is that it also works in camera views, which is handy because it can be quite frustrating trying to drag a section box into the field of view while working in perspective.  The most accurate way to do this is to go to a plan view, find the camera view in the project browser, right click and "show section box".  But the sunglasses trick can alse be useful as it offers a very fast way of bringing all 4 boundaries into view at the same time.

Another thing that used to really frustrate me.  Sometimes when you reload a family into a project nothing happens.  For some reason Revit doesn't recognise the family as being a more recent version.  Most often this is a door or a window, and the workaround is childishly simple.  Within family editor, stretch the end point of the host wall very slightly.  This won't have any impact on the geometry that loads into your project, but it does seem to rouse Revit from its sleepy state.  "What? oh, you mean this is a new family? Fair enough, I'll swap it out then."  If it's not a hosted family you can make a slight change to some insignificant text field.  The principle holds.

This is a strange one really, and I don't think it is widely known.  How small can a room be before a door refuses to recognise its existence ?  Does that sound like a silly question ?  The answer is very straightforward.  Anything less than 13 inches is too small.  It seems a bit arbitrary.  Maybe there is a technical reason.  I only discovered this by accident quite recently.  If you have a door that provides service access to a shallow MEP shaft, the room name won't show up in your schedule.  The workaround is easy enough.  Open your door in Family Editor and activate the Room Calculation Point.  You can adjust the position of this point from within Family Editor, but not within the project, which is quite frustrating.  In the snapshot below, you can see the room calc point in the highlighted hatch family, which recognises the room.  The door on the right can't see the room.  Hence the blank field in the schedule.

One more for the road.  If you want to change a "baked-in" parameter from type to instance, you can do it by selecting a labelled dimension and ticking "instance parameter" on the options bar.  It's crazy how many things in Revit work easily be one method, while another approach seems to be telling you it's impossible.  But my tip has to do with "what if no labelled dimension exists?"  For example, it could be that the thickness of a nested door panel is controlled by a linked parameter in the parent family. 

My workaround is to draw two symbolic lines in the current view.  Dimension them.  Assign the label.  Select and change to instance.  Delete the lines.  Fast and effective.


I don't often do these "handy little tips" types of post.  They are well covered by others and I have no wish to emulate anyone else's style.  Which brings me to the Revit Kid, who has been blogging for much longer than me and who excels at offering hand little tips and practical Revit tutorials.  For some time now he has been developing his Revit After Dark series and volume 3 is about to come out. 

He was kind enough to give me a sneak preview and I really enjoyed looking through it.  There's a lot there, and I think it will repay more visits when I can find the time, but I really enjoyed his approach.  The videos have a very relaxed style and he's opted for a "warts and all" live cut, which makes it very fresh and easy to identify with.  All the videos are about Family Editor, and if there's one thing for sure, it's that you are going to be faced with a fair amount of trial and error and backtracking when ever you set out to make a bunch of new families, so you can learn something from watching how he handles those situations "live".

I won't say much more, but there's all kinds of bonus goodies in the package also so it's well worth the investment.  Take a look.

The Revit Kid

Tuesday, February 23, 2016


This is my first attempt at a slightly different format for my posts.  The aim is to group the "stuff about buildings" and the "stuff about Revit" into two parts.  Readers who don't use the software can skip the second part.  Software addicts who don't see the value in my musings of "the way we build" can focus on the "tips and tricks".  Hopefully most of you will gain a clearer picture of the synergy between these two aspects of my work.  The starting point today is one of Palladio's lesser known Villas.


According to the pictures on the web, this is still a semi-derelict villa.  Most of the stucco has fallen off, and there are flowers growing out of the joints in the brick steps leading up to the front porch.  It seems that the villa hasn't had the happiest of histories.  Palladio was modifying an existing structure for Gazotti who had to sell it on to a Mr Grimani before it was finished in order to solve his cash flow problems.  Perhaps he was trying a little too hard to compete with his neighbour whose grand estate dominates the village.

Palladio had two main basic ideas for the central element of his façade: the triple arched porch, and the classical temple portico.  This one has the arch motif, but he throws in a pediment for good luck, and an array of 8 pilasters that divide the whole frontage into 7 equal bays.  I think it's the only one of his country villas to set up a consistent rhythm right across the full width.

Before I started this exercise, I hadn't really given much thought to the importance of "outdoor rooms" in Palladio's villa plans.  This one is a classic front porch, overlooking the street and large enough to lay out a large table on a Sunday afternoon for guests coming from a service at the church next door.  There are 3 imposing doorways, leading to the 3 main rooms of the house.  It seems to me that these four spaces, (one outdoor and 3 indoors) define the house.  The rest is a bunch of much smaller storage spaces and hay lofts stacked up at the back, cramming in 3 floors to the one lofty storey of the master's domain.

The principal room is cruciform, with an impressive brick groin vault.  There is a fascinating photo online of the roof void over these vaults showing the ribbing of the vaults and the massive timber truss and rafters that support the clay tiled roof. 

One thing that doesn't come out very clearly from my sheet at present is the contrast between front and rear elevations.  From the back it looks pretty much like a storage barn and I am starting to wonder if Palladio's main contribution here was to remodel the façade, basically decorating the shed.  The ends of the front wall actually stick out at the sides rather than wrapping around to form a proper side elevation.

 Just to recap, what I have built here is a simplified massing model that aims to put Palladio's Villas in context.  There is a large square of topography, with subregions for rivers and fields; a few floors and walls, some "lollipop trees, and families to represent buildings. 


Like anything we do using a BIM approach, deciding how much detail to show is critical to success.  It's not just a question of imitating reality.  We have to consider our goals and develop the model accordingly.  Hence the simplified trees.  In a similar way, we are always faced with the question "how parametric?"  Most of this is directly modelled, much of it judged "by eye".  I've bitten off a rather large task here, and can't afford to get too bogged down with accuracy.  In any case we don't really know what the Veneto looked like more than 400 years ago at this level of detail.  But I am assuming that there were rows of cypress trees, as there are now, and for these I can take advantage of Revit's dynamic array feature, which really comes into its own at masterplanning level.  By choosing the "last" option I can nudge around the tree at either end to fine tune both length and alignment of a row.  If the density of spacing doesn't look right, just type in a new number.  This is parametrics built in to a system tool.

The contextual buildings are loadable families. Inside you have a box (rectangular extrusion) and a nested Roof family which is set up so that the ends can flip between gables or hips.  Gables need to be wall colour and hips to be roof colour so you end up with two pairs of parameters: a yes/no and a material for left, then the same again for right. 

The switching between hip and gable is achieved by a retractable void extrusion and a simple "IF" formula.  If the box is ticked, the extension is set to full depth, which cuts a hip slope.  If the box is not ticked the extension is set to zero, and you have a vertical gable.  The pitch angle controls the height of the roof via a simple trig formula using the tangent of the angle.

The end result is a family which can be easily switched between 3 conditions: two hips, two gables, one of each.  Opinions may well differ on the balance of type and instance parameters for a family like this.  In my view there should be a maximum of 4 or 5 types, preferably with simple, intuitive names.  Currently I have them set to combinations of roof pitch and material. That way it's easy to select a bunch of and switch types without disturbing the geometry of the layout.  It's just a way of introducing variety.  Meanwhile, in plan view there are shape handles to make it easy to stretch things around.

So the contextual buildings are simple, variable, highly parametric.  By contrast, Palladio's villas are modelled in much more detail but have no built in variability at all.  I've tried to gauge the minimum level of detail needed to convey the essentials of each design.  The basic room layout is revealed depending on the height of the cut plane, or the clipping of a section box.  The principle openings are cut out by voids, and there are simplified mouldings to break up the elevation and reflect Palladio's basic proportions.

It's all done with simple solids and voids, mostly extrusions, a few sweeps, the occasional revolve.  But the starting point is a drafting view where I have assembled my source material and scaled it up to ... now there's a question.  What size to use ?  In a few cases there are dimensions in known units.  Then of course there are estimated measurements from Google earth, and finally we have the numbers written on Palladio's own plans.  I am quite certain these are Venetian feet (348mm)  So I've used a module of 350.  External walls are commonly 700mm thick for example, and most rooms are whole multiples. 

Then I traced over the source images, creating filled regions for the floor plans so I can copy-paste the sketch into an extrusion in the family. 

Elevations are a bunch of detail lines, grouped.  A second copy of this group can be swapped out in a clean drafting view and exported to CAD.  That's the best method I have for bringing the front elevation into my family.   One elevation usually provides enough guidance to create an acceptable massing model.

Maybe you don't see yourself studying the works of Palladio using BIM tools.  But why not apply these same methods to an urban design study as part of a concept design report for one of your current projects?  Adjacent buildings could be modelled from source material to a similar level of detail as Villa Gazotti here.  The more distant cityscape could be assembled from parametric families of various shapes and sizes. 

Take the principles.  Apply them to your situation.

Tuesday, February 16, 2016


There must be some way outta here, said the joker ... aaah Jim lad, you've been reading too many error messages.  Take this one for example. 

Actually the element may well still be there.  The little robot that sits inside Revit can't possibly be programmed for everything that might conceivably go wrong.  Maybe it would be better if it just said, "can't edit that until you synchronise".  Then you would know what to do, instead of going "who deleted my stuff !!"  I've only come across this when I open a workshared file and (very naughtily) try to do something before creating a local.  Or if I open from a local, trying to do something before reloading latest.  Either way, I'm taking a risk in the first place, but then, life is full of risks.

For example, I took the risk of upgrading to Windows 10 a few weeks ago.  The good news is that all my versions of Revit (and everything else) are running ... no problem.  Well, just one or two slight irritations.  I think I have to update my video driver.  Sometimes views appear really strange when you first open them ... all pixelated.  It goes away as soon as you pan or zoom or do anything really, so it's not even annoying enough to persuade me to look for the driver.  I'll get around to it.

Somewhat stranger, and more disruptive, the Quick Access Toolbar disappears when placed above the ribbon.  Not a train smash, because I can place it below, and I'm slowly adapting to this position.  I originally placed it up there some years ago to maximise on my limited laptop screen space, but these days I have a second monitor at work and at home, a couple of millimetres doesn't bother me.  And of course there's room for a lot more commands when you put it below, just in case I ever want them.

That's all I've noticed after 3 or 4 weeks of daily working, but there is an unrelated graphics issue I want to mention.  The background shading we can now add to columns in a schedule is a wonderful thing.  It can turn an intimidating mass of data into something quite presentable and relatively easy to read.  Some additional formatting in the horizontal direction would be even better of course, but what we have is a good step forward.

One unfortunate side-effect of solid fill though is that when you select a schedule on a sheet to adjust the column widths, it all turns blue, and you can't read the text.  The main reason for making this kind of adjustment is to make sure the longest item will fit, so inability to see that item is rather unfortunate.  To be honest, it would be far better if the background fill just disappeared when the schedule is selected.  I assume that would be easier to achieve than having it turn a very pale shade of blue "behind" the text.

and that's all I have to say for now.  Must be one of my shortest posts ever.

Monday, February 8, 2016


Right at the end of Project Soane, (Take 1) I got this idea about Soane's 45 year epic tour of duty lying at a fundamental turning point in our history as a species.  There are lots of graphs floating around that show a massive swing from almost horizontal to almost vertical taking place at that very point in time.

I decided to draw my own version of this, in a Revit drafting view.  You can graph almost anything: population, urbanisation, energy use and the results come out much the same.  Several people have invented combined indices of "Social Development" or some such similar idea.  Don't get the wrong idea, higher numbers don't have to mean "better".  Lots of societies in the past achieved high numbers on the back of slave labour and warfare.  Here's an overview.

The basic story is of painfully slow rising levels.  At this scale you can't see the ups and downs in the early part, but there were lots.  Civilisations rose and fell: ecological collapse, plagues, civil unrest and all that good stuff that comes with cramming ourselves into cities.  But there is a visible hump that coincides with the Mediterranean boom of the Greeks & Romans (don't forget the Phoenicians)  In fact the levels of industrial production, mining etc that the Roman Empire generated were not to be equalled again until around the time that Soane was growing up.

The bar chart in orange and blue shows book publishing in Western Europe.  It's on a logarithmic scale so you get a fairly straight line, but if it was plotted normally there would be an elbow bend in that one too.  Maybe the upturn would start a bit earlier, but the mid-point of the turn would still be in a similar place.  Why?  You can give complicated explanations if you want, but the simplest answer is fossil fuels.  For whatever reasons, Western Europe succeeded in shifting to a predominantly fossil fuel powered economy around that time.  In my view, a whole range of factors happened to come together at the right time, motive and opportunity if you like (capital, skills, markets, knowledge ...) England was at the heart of this, but the interaction with other European states was also crucial.

So my idea was to make an abstraction of the bank in plan, and use it to represent that hinge point.  Part of the inspiration is the way shape formed where the NW extension (Tivoli Corner etc) meets the rest of the bank.  The curve of Princes Street, before it was straightened is just about right for that dramatic change of direction.  In a way, abstraction and the LOD issue ask the same question.  What level of simplification is appropriate to my current purpose?  Rationalising the angle at Tivoli to 60 was a no brainer and leads to a triangular grid. How many divisions though ?  Splitting Threadneedle into 2 halves seems too crude.  I settled on fifths.  You get a narrow slice in the middle for Sampson's original courtyard block, with pink and blue squares to left and right for Taylor's contribution.  I tried a curve where the green meets the orange and pink, but that was distracting.  Better to limit the curves to the perimeter.  All this is old-school drawing, but done in Revit.

Nothing wrong with that.  Visual thinking goes way back: 30k ya in caves where Spain and France were to meet up in more recent times.  That's 2d abstractions on stone surfaces: the beginnings of reality capture perhaps.  Then 9/10 of the way to year zero (and Rome) we get 1d, captured on clay with wedge shaped sticks.  That's the roots of digital I guess (mark or no mark) a single stream of linear meaning representing abstract ideas (food, life, five ... me Tarzan, you Jane)  Actually the me Tarzan thing is spoken language which goes back a bit further (or maybe much, much further)  But let's stick to the 1d, 2d, 3d idea: visual thinking and abstraction are the current topics.  We could choose to pinpoint 3d as coming of age in the Renaissance with Brunelleschi and the theory of perspective.  Soane used those 3 modes of abstraction habitually: orthographic views drawn to scale (2d) written annotations (1d) and watercolour sketches (3d perspective)  BIM just links those 3 modes together so that they talk to each other.  Back to my diagram

Perspective kicks in shortly before Guttenberg.  Portugal discovers Madiera around that time and starts growing sugar to further undercut Venice's monopoly on trade with the East (they also grew sugar on Cyprus)  That got Palladio building villas for Venetian gentry going back to the land as the bottom dropped out of their merchant adventuring. He and others started to use Guttenberg's technology to immortalise their achievements (Serlio, Scamozzi, Vasari)  Columbus found another good place for growing sugar (Carribean)  Portugal already had a good place to buy the slaves that you needed to make sugar plantations viable (Europeans couldn't hack it in the heat) So the graph is climbing, people are making money (and babies) right across western europe, but especially on the Atlantic rim as open sea sailing started to pay off big time.

Zooming in a bit (and using the mouse-driven pen on the windows snipping tool to make my red worm nice and wiggly)  we can see some of the spin-offs.  England used up most of their forests building ships, which pushed them into coal mining.  This came in handy around the time Soane was born when demand for iron and steel was driving that technology forward (Benjamin Huntsman, Abraham Darby etc)  Indian cotton imports had been making big profits for the London merchants who ploughed some of this back into financing warfare with their French and Dutch rivals.  That's the birth of the BoE, crudely pinpointed above ... with the six main phases of its building programme shown below as a series of abstract snapshots, like a video editing timeline.  One thing that strikes me about this "Hinge Point" period is that people had hardly built something before they were knocking it down.  Soane's retirement is closer to Baker's demolition than it is to Sampson's foundation stone.

This chronic addiction to change is still with us of course, and I am pointing the finger at fossil fuels as the critical enabling mechanism.  But the switch to fossil fuels was driven by the cotton industry.  That's what the blue line is trying to say (which is just my invention, no hard numbers to back it up)  The India trade makes lots of cash.  Wool merchants in London get miffed about the competition, persuade parliament to pass the Calico Acts (Calicut Trade) cotton profits drop, English weavers grasp the opportunity, artisan inventors like John Kay clear away choke points ...  Now you have capital available, a proven market for cotton textiles, a labour pool, coal mining & steel production already growing steadily. 

Richard Arkwright just had to light the fuse by proving that he could make a fortune almost overnight with the first mechanised cotton spinning mill in Cromford, Derbyshire.  And Bolton & Watt just happened to be on hand to offer Steam for sale when the water power maxed out.  The blue worm is saying that the India trade outstripped average growth, triggering restrictive regulations.  Blue worm dips below red, opportunity knocks and invention boosts the output of English weavers.  Greedy eyes make greedy hands get to work, the wig maker (Arkwright) borrows an idea and turns it into a reality, England sells cotton to the rest of the world in unheard of quantities and the blue line is now driving development.  Cotton mills become the market that Steam power needs to fund its R&D.  Without that roller-coaster ride of happy coincidences we might still be trundling along at a modest rate of development.  Fossil fuels might not be running out.  Population levels might be a fraction of what they now are.  Pollution, industrial waste, rain forest devastation ... maybe that would be better.  But slavery might not have been abolished, higher death rates, less democracy ... we just don't know.

All I'm saying is that Soane's Bank of England sits at a hinge point of History, it's a fascinating topic, and Project Soane really helped me to dive into it.  Speaking of which ...  Sean Young from HP, one of the major sponsors behind Project Soane, sent us a link to some interesting photos from the Bank's archives.  You can see them here.

I had seen some of them before, but at lower resolution.  Looking closer I started to make some interEsting observations, so I've done some markups to share with you here.  (done in Revit of course)

I'm guessing that the first one is early C20 Looks like the street lighting is still gas powered.  That's a technology that started to come in during Soane's era.  Seems hopelessly quaint to us, but what a huge impact it must have had.  You can also see that the urban transport situation is still in the balance.  Horses are a renewable resource, oil is a fossil fuel.  Steam powered railways would have transformed inter-city transport by then, but getting around in central London was still a mixed bag.  The first cut and cover, steam-powered underground would have been in place (metropolitan and circle line) but not the electric powered deep tube lines.  Buses of both types are visible.  Also notice the lighter colour of the upper portion of the screen wall.  That was a panic response to chartist riots which made a serious mess of Soane's carefully considered parapet details.  Proportions all to cock and seriously crude detailing at the back.

The next one is from almos the same place, but a decade or two later and looking to the left.  The railings have gone, melted down in the first world war I guess.  Lamps have changed.  New tops on the old poles I think.  That would be electric lighting replacing gas.  It must be early 1930s now and Herbert Baker is well into phase 2 of his reconstruction programme.  Interesting to see that the stone cladding is all hung off a steel frame.  So Herbert's work has something in common with Sir John's.  He had to work in phases, within spaces surrounded by existing structures, and he had to keep the weight down to minimise disturbance to adjacent foundations.  This photo is great for final confirmation of Soane's screen wall elevations by the way.  It would be wonderful if they were even sharper, but this is very valuable input.

Now we get to the guys in flat caps and waistcoats demolishing the rotunda dome.  No safety helmets or harnesses.  Pretty much the way things still are on many African building sites.  The decorative treatment of the rotunda interior is shown more clearly here than any other photos I've seen.  Typical Soane incised lines and geometrical patterns based on ancient Greek precedent.  But the most interesting part for me is the insight into the haphazard accretions that have colonised Soane's "battlements"  All along the watchtower, strange carbuncles grew.  I suppose this is a reflection of the chronic need for more square feet to accommodate the bank's activities: basically the pressures that led to the destruction of Soane's masterpiece.

Moving on, we have a caryatid being lowered from the lantern above the rotunda.  So this must be a couple of weeks earlier than the previous shot.  You can still see those caryatids by the way, if you visit the museum inside the current bank.  The other half of the picture is looking towards Bartholomew Lane from where the Dividend Office used to be.  You can see the last little bit of the Stock Office not quite full demolished.  The entrance to the Bartholomew Vestibule is a bit of a surprise.  It seems to have a half-round fanlight, which was unexpected.  I have never come across a drawing or a photograph that shows the treatment of this doorway clearly.  So maybe it was more similar to the two gateways along Lothbury than I had previously imagined.

And finally to the dividend office itself.  This is from a different source.  Clearly Baker reconstructed that space in almost the same place.  A homage to Soane I guess.  There are some changes.  The top part of the lantern is omitted, and the whole back wall opens up to the space beyond.  Not sure if this is part of Baker's scheme or from a later conversion to "open plan" office arrangements. Old questions answered and new ones posed.

These may seem strange uses of Revit to some, but I would rather do my quick 2d sketching in the same software as my full-on BIM work whenever possible.  It opens up the possibility of better continuity.  Personally I would really like to be able to hand sketch directly into a Revit drafting view using a touch screen device.  Hand sketching is still the fastest way to explore certain kinds of idea, to link together hand, eye and brain in an intuitive feedback loop.  That human capability should be better integrated into BIM workflows. 

If we are going to link BIM directly to robot arms, surely we should also link it directly to human hands, thinking visually as they have been doing for tens of thousands of years.  Don't you think ?