This post is part 2 in my series on Layer Theory.
Click here for part 1: LEGO and Layers

Last year I got into an argument with some non-ArchiCAD BIM users—this happens more than it should. They laughed at us for having Layers. They mocked “layers are so CAD! BIM should only have real things in it. Layers are fake. Layers aren’t worthy of BIM.” Okay I’m getting a little hyperbolic. They probably weren’t that bad; and we all actually do get along very well. But they did get me thinking about Layers. It turns out those grumpy BIMmers raised some good questions, although the most salient point was something unintentional. They equated ArchiCAD Layers to AutoCAD layers. Which in turn were just a replication of the old pin-bar drafting technique of laying different sheets of Mylar over each other. Each layer of a drawing was literally a different sheet, which when overlaid and pinned in different combinations eliminated some repetition of work (only draw the walls once!).

This is before my time, thus I don’t know the exact details. It’s hard to find discussions on the Internet about this archaic ritual, so I might have some details wrong. Nevertheless, the day I learned about the origin of layers from an older coworker, my mind was blown. Layers used to be physical things! Wow.

Anyways, back to the story. The grumpy BIMmers think Layers and think AutoCAD. They think digital approximation of an ancient method (whether they know it or not). And to be honest that’s what we all see Layers as when we jump to ArchiCAD from CAD. We understand layers because we used them before. We view Layers like simple bins of 1980s LEGO bricks. But (and you should all see this coming) just because we did something one way in one program doesn’t mean we need to do it that way in another.

What is the value of a Layer in BIM?

Well first of all, there’s the concept of graphic data, digital approximations, and metadata. Layers are metadata. They have value and purpose. Layers are an essential metadata tag on elements to help classify and organize them into groups. Layers allow us to turn elements into wireframe in 3D. Layers determine whether elements interact with other elements (Layer Intersection Groups). Layers control visibility. Layers determine whether an element is editable or not (locked/unlocked). And most importantly Layers allow us to take advantage of the automation that Layer Combinations provide us (spoiler alert: that’s Part 3 of this series). But in ArchiCAD we have a ton of ways of sorting and displaying data: renovation status, IFC properties, Structural Function, Element Classification, Position, linking elements to a Change, Model View Options, etc. So while Layers do act like those old pinboard layers, Layers can function differently in ArchiCAD. Or more importantly they can be parred down to their essence to better take advantage of all their other features and potential. While Layers can handle massive amounts of organizational options, they don’t have to. We can let other aspects of the program do some of the work we’ve been slavishly making Layers do. And when we cut down the number of Layers from 100 to 50 to 10 or even 1, then we can really start making them work for us. After all, managing 10 layers is much easier than keeping 100 organized.

Imagine if you had one layer per design scheme. Or just two layers: on plan and not on plan. How would those two methodologies affect how you work in ArchiCAD?

Value of LayersLayers aren’t always the answer.

Think back to my articles on Find and Select (part 1 and part 2). If elements on a given Layer can be divided easily using Find and Select, then you might not need to separate those elements out into two Layers. In ArchiCAD 9 or 10, the way to hide elements was always through Layers (essentially). Now in ArchiCAD 18, we have so many other ways: Renovation Filters, Partial Structural Display, Pens, etc. Here’s a great thread I learned about while on Twitter: A collection of reasons why an element might not be visible. The points Link and Laszlo make are about visibility, but they are also about organization. Each of those reasons could theoretically remove Layers from your file. We’ll look at many of them more closely in the future, but for now I want to stay focused on the humble Layer.

Ken Huggins. Ken Huggins. Ken Huggins.

Did I write Ken Huggins’ name enough time? I can never say enough good things about Ken’s articles on regarding Layers and data organization. Read them. You’ll realize you have too many Layers. I have a lot to say about Layers, obviously, but for a Masters class on Layer simplification, read Ken’s articles. I seriously reread them once a quarter, if not more often. I also pester him to write more just about as often.

Do you really need that Layer?

One Lonely Layer

Whenever you rebuild your template, or are thinking critically about the project you’re working on, there are a number of questions you should ask yourself about your Layering system. They all essentially ask the same thing, but from different vantage points: do you really need that Layer?

  • Open your most recent (or best) project that utilizes your current template. Open up the Attribute Manager and look at what Layers aren’t being used (in the Attribute Manager, Attributes that are in use will have a check mark next to them). If there are any unused ones, can they be purged from the next iteration of your template? With an increasing number of caveats, this same mentality can be applied to other basic Attributes in your template (Surfaces, Building Materials, Composites, Pen Sets, etc.). Remember, there is only one Layer that HAS to be in your file: only the ArchiCAD Layer can’t be deleted.
  • This isn’t AutoCAD. It’s been years or even decades since many of us had to regularly use a 2D CAD program. But I bet those early Layer lessons we got in the 80s and 90s still unconsciously haunt us. Again check out Ken’s articles on You wouldn’t for a second make a window or door Layer in ArchiCAD, so don’t take for granted that we need any other specific Layer.
  • Look at the list of Layers in your template. What holdovers are there from old workflows? In ArchiCAD 11—which is where my modeling really took off—I had a number of Layers that were the result of two converging issues: model complexity and the difficulty of working in 3D. I had Layers for trim at various elevation heights (crown and baseboard). I also had Layers for vertical trim elements and horizontal elements (I was doing my window trim separate from the window objects, so there was some logic to this). But over time, while I maintained those Layers, my workflow changed. My projects weren’t requiring that much trim and more importantly what I was once doing in 2D was now being done in 3D. The need for all those Layers was so that I could isolate information to work on certain parts. Working in 3D in those much older versions of ArchiCAD was just too difficult. For little tweaks, sure. But for serious work, not really. Of course that’s not the case anymore. 3D is as easy or sometimes easier to work in. So whereas before I’d hide everything to work on only a few elements in 2D, now I can use Find & Select, grab my elements, then work on them isolated in 3D. The result: more fluid workflow and less Layers. Another reason I don’t need copious trim Layers anymore: we can drag or mirror a copy in Section and Elevation. That feature alone can cut out a few Layers in many people’s templates, I’m sure of it.
  • Do you really need those two Layers? Speaking of holdovers, remember to think about unnecessary complexity. Having a separate Layer for text and dimensions makes sense (maybe) for architectural plans. You have lots of views, lots of elements. It’s nice to separate them out. But for electrical or structural plans, can text and dimensions be lumped together in one annotation Layer? And furthermore, when was the last time you really needed to separate out dimensions from general text? Maybe it’s regularly, maybe it’s never. If it’s never, think about consolidating down to one Annotation Layer.

Be deliberate. Be bold. Go through your Layers and make a convincing (and simple) argument for why each should stay or go.

  • Do you have specific Layers for certain Tools? Why? Tool Type and Object Type do not equal Layer. You do not need to make specific Layers for specific Tools or Objects. A Layer for all Morphs makes no sense. A Layer for all Schematic Design elements might. A Layer for all Roof elements is wrong. A Layer for all elements that act as roofs (which might be Roofs, Slabs, Shells, Morphs, etc.) and need to display together might be more appropriate.
  • You do not need special Layers for different stories or special ones for section views, elevation views, interior elevation views, or detail views. The Dimension or Text Layer in plan should be the one you use everywhere else. The only reason you’d need a separate Layer is if you have multiple Views of the same Viewpoint that need information shown or hidden differently (and which can’t be achieved with non-Layer solutions). For instance hiding architectural text and dimensions, but needing different text and dimensions on a structural or electrical plan. Or showing Section Markers and Elevation Markers on one set of plans, but only Section Markers on another.
  • Minimize Graphic layers. Long ago I used to have a masking Layer for line work, a masking Layer for fills, a Layer for 2d elements representing things above, one for 2d elements representing things below, a general 2D detail Layer, and more. Now I have one Layer: Graphics and Masking.2D. All of that stuff now goes on one layer. Of course my workflow also is more 3D based, so I need less of that stuff, but still. It now all goes to that one layer. In the LEGO analogy I have fewer pieces to manage, so they don’t need separate bins.
  • Don’t just minimize the complexity of how you manage graphic data; use your Layering system to downplay 2D. One of the reasons I continually delete Layers that primarily contain 2D information is to make 2D less important. The overarching structure of your layering system will affect how you and your coworkers use ArchiCAD. Do what you can to nudge people towards 3D, IFC, and higher BIM principles. Make your Layering system about the model and visibility.

I’ll talk about naming conventions in a post soon (I promise soon, it’s mostly written), but I do want to mention this here. Make your Layer names clear. Minimize or outlaw abbreviations. Make them specific enough for someone joining your project to instantly know where an element should go, but generic enough to handle a wide range of elements (does it need to be A_Walls.Exterior or can it just be Walls?). Remember to use both the Layer Name and Extension (pro tip: if you add a period (.) in a Layer name, whatever comes after the period will be the Extension). I personally use the Extension to give additional qualifiers to the layer: Annotation, Hidden on Plan, 2D, 3D, On Plan, etc. The goal is to help another user understand my logic (for more on the specifics, download my template; run through the Layers and call me out on the ones I should cut, I know there are a few).

What am I forgetting? What other questions are worth asking when deciding if a Layer (and its added complexity) deserves to be in your file? One topic I haven’t covered yet is Layer Combinations. That’s up next in Part 3. Another concept I’m ignoring for now is various external standards. It’s been years since I dealt with any of these and unless they are required (or you are chasing work that will require them) I’m skeptical of building your workflows to meet standards that are designed to be general enough to cover everything.


IDelete a Layerf after reading through the above list you have some Layers you want to delete, here’s what to do once you open up the Layer Settings:

  1. Select the Layer you want to Delete.
  2. Click the Delete… button.
  3. ArchiCAD will warn you if deleting the Layer will result in the loss of elements and/or missing attributes. If that’s the case, you can click on details to learn a little more. If there are no elements, delete away. If there are:
  4. After clicking the Delete… button, select “Move Elements onto:” and pick the Layer you want the elements to move to. Then hit the Delete button in the bottom right.
  5. Repeat until you have as few Layers as possible.

BONUS ROUND II: Rules of Thumb

If 2 elements will always be seen together on the drawings, then they should (probably) be on the same Layer.

If 2 Layers are always shown together in the Layer Combinations, then they should (probably) be merged into one Layer.

If 2D elements need to be seen separately to work on but always show up together in Views, then they might need to be on different Layers.

If 3D elements need to be seen separately to work on but always show up together in Views, then they probably don’t need different Layers. Show Selection in 3D and Filter and Cut Elements are two simple and great ways to work on 3D elements in isolation without the need for excess Layers.

BONUS ROUND III: Breaking the Rules

Of course there are exceptions to everything. But following the above guidelines is a great place to start. First you learn the rules. Then you learn to be creative within the rules. Then you learn how to break and ignore the rules.

 Are you following Graphisoft North America on Twitter? Click Here to keep track of all the latest ArchiCAD news in North America (and beyond).


  1. D.G.Drews

    Excellent information and concepts. I just read Theory1 & 2, and Ken Huggins piece.Thanks for describing more clearly how elements’ visibility can be controlled. Differentiating between the use of layers and using the identity and/or attributes of the various elements to control visibility makes better use of Archicad’s abilities.

  2. Link

    Hey Jared

    Thought-provoking article indeed. I’ve always found it interesting that people treat the word layer like it’s a four letter word. And I’m not talking about grumpy BIM aficionados looking to pick a fight, but even the most conscientious ArchiCAD users.

    So in the layer’s defense I’d like to pose the question: what is it to ‘manage layers’ and is it worth all the effort to strip the layer list down to a low number, just to have a low number of layers?

    With the millions of pixels on our screen these days and free spinning scroll wheels it shouldn’t be a matter of saving real estate. So I guess that basically leaves creating/updating layer combinations. But in any template worth it’s salt, this is typically a one time deal and there are easy ways to copy layer settings between combos. To me, the management is minimal.

    On the flip side, if you have plenty of layers you have plenty of flexibility, predictability and consistency. I’m not talking about creating layers willy nilly or having a new layer for every conceivable element. I’m talking about having just what is needed to utilize everything you mentioned; visibility, lockability, intersectability, find and selectability, etc. And let’s not forget reservability. In a teamwork project reserving by layer should never be forgotten.

    WITH ALL DUE RESPECT, I question whether a template that has just seven layers, based on cut plane methodology which produces nine rather major limitations (including plans from a worksheet?!) is rather counterproductive. For the sake of a small layer list we are now working around the Graphisoft intended workflows which are designed to make the whole process easy. I do believe it could work, but I don’t believe it is ideal, especially when a large team is involved. Perhaps I’m missing something? Either way I’d need some convincing.

    Personally I’ve always considered a good benchmark to be 100. If a template is under a hundred layers then it is manageable. That way you can have floor plan specific notation, layers for all typical element types both real-world and ArchiCAD specific, residential and commercial, etc, that are all intuitive, consistent and predictable. And reservable:-)

    Don’t fear the layer.

    Just my two cents. Please forgive me if I took this a place you didn’t want it to go.


    • Jared Banks

      Link, this is exactly where I want this conversation to go. I think a template done right could have a small, medium, or large amount of Layers. Mostly what I’m interested at the moment is a small or medium amount because that fits my work. But you’re right there is a sweet spot for a “large” amount. But as you also say, it’s about right sizing that. So if 100 is a good level, then make sure you’re not at 90 or 110 if they aren’t optimized.

      Someone was sharing with me privately about using Omniclass for layers to facilitate IFC export. I think that’s great and would create a different set of Layers and a different number. And if one were to go that way, then the Layers better fit that perfectly and not have any of my plain english ones.

      In the end it comes down to what and why are we modeling? I think it’s legit to say my SD model has less than 10 layers because those layers function differently than a file with 100 layers and a different purpose.

      I have a post brewing about sketching in ArchiCAD. That’ll defend the low layer proposition more.

      Can you tell me more about “reversable”? I’m not sure I understand what you mean. And by the way, thanks for the awesome comment and taking the discussion in this direction.

    • Jason Smith

      Hi Jared & Link

      I totally agree with Link. Flexibility is what I’m after. You can add another layer combo and get a completely different output (provide additional information from the same model).

      I have just checked my V18 Template and it has 99 layers. I haven’t been to concerned about the number of layers as long as the template is set up to provide the required outputs we need for the types of projects we have.
      There are some of those layers that may only be used for 1 thing, e.g. area calculations for a particular schedule of areas.

      On Managing Layers, thats not hard. Layer Combos handles that.
      I do remember at one office having minimum layers (had a variety of different experience with AC). We used stories as sheets and had layers for each tool. Made sense back then, no 3D model though. Would never go there again. But that does make me think that users who don’t like layers probably don’t understand how to manage them. They of course need good template.

      If GS adds more filters then layer numbers could be reduce to a minimal amount.

      • Jared Banks

        For everyone’s reference, my personal template (Shoegnome Open Template v18) has 60 Layers and 10 Layer Combinations. So not 1 Layer, but also not 100 or 500. It is also designed for flexibility and is very flexible. I do dream about getting it to 50 Layers, but perhaps I’m already pushing the lower limit for what we all are arguing for.

        We need to remember that at some point flexibility stops and unnecessary separation begins—there is a level of flexibility that could be unnecessary. Maybe.

        I agree that managing all this isn’t hard, but as we add more users to the team, making sure they follow the rules does get more complex.

        Finally, perfect example of a horrible Layering scheme! Who else has an example of a Layering system gone wrong?

    • Tom Markunas

      I find it curious that this ‘appropriate number of layers’ discussion so blatantly omits any mention of context, ie. project type or size.
      A kitchen/bath renovation in Soho, a gut rehab of a brownstone in Harlem, a new house in the Hamptons, a corporate relocation on 6th Avenue, or a new office building at Hudson Yards (sorry, all NY references here) all have DRAMATICALLY DIFFERENT layering requirements.

      May I suggest further discussion about this topic be specific to project types?

      • Jared Banks

        Tom, good points. It could be argued though that it’s grain/complexity that matters just as much. If you do a bath renovation you might need multiple tile Layers so each grouping can be isolated. Or for a kitchen, there might be some weird multiple Layers for cabinets or equipment. Conversely for low detail skyscrapers, maybe 3 Layers is enough. Just a thought exercise and coming up with weird examples. You are absolutely right though that all this also hinges on project type and scale. What project types are we doing with these number of layers? I personally am doing primarily residential projects. So stuff under 10,000 sq ft.

        I also want to say I’m loving everyone’s comments. It’s showing that Layers aren’t just something we can all assume every ArchiCAD user treats the same way—or even thinks about!

  3. Link

    I actually meant reservability – as in the ability for elements to be reserved by layer in a teamwork project. Much like Find & Select in that you can use almost identical criteria, including layers, to reserve elements.Great for users who need to only deal with specific items like interiors, structural, or updating drawings, etc.

    I definitely agree about the sweet spot. I just don’t want users to get the impression that the absolute minimum is the holy grail. Most people will probably find it shocking that my template not only has 85 layers but also 66 layer combos. But the way the template is set up, they will almost never have to bother with them. In fact there is only one ‘working’ layer combo that ever needs updating, so management is minimized but flexibility is maximized. And it’s easy as pie to take in.

    Don’t fear the layer. Or the layer combo:-)


    • Jared Banks

      Doh! Reservability. That makes way more sense! Oops. Right on. I haven’t bothered reserving my Layer in Teamwork since TW1. But I do see the value—also for some templates/teams to have Layer Combos for reserving different groups of Layers.

      As we discuss more, if I had to make a recommendation I’d say 50-100 Layers is probably right. Less than that may approach restriction, but above that might start having unnecessary bulk. I don’t know. I reserve the right to change that opinion latter! What do other people have for numbers? I’m looking forward to checking out your template Link. I think we have some similar views, but also some very different ones. I can’t fathom 66 Layer Combinations. But then again most of my sets rarely have 66 saved Views or placed Drawings.

      Minimal Management, Maximum Flexibility

      Don’t fear the Layer. Don’t Fear the Layer Combination.
      Don’t worship at the altar of the Layer and Layer Combination either. ???

  4. Larry

    Hey Jared,

    Great article as usual. I find myself always looking for ways to trim down and simplify my layer list. You definitely, gave me some new ideas of where to trim the list. When we started with ArchiCAD 8.1 many years ago, we brought over most of our old 2D CAD layers with us. The first thing I did was try to make them plain english, since I didn’t know what many of the abbreviations even meant. Since then, we have continued to pare them down while at the same time adding 3D specific layers. In the end, I still have 99 layers.

    I wonder if many people are like me in that they like using layers rather than other parameters because they are just so easy to use. Sometimes digging into an object to make it more smart, is more effort than its worth. Especially, if the recipients of your drawings aren’t really using your BIM model at all.

    But, the best take away I got from your message this time is a reminder of the importance of constantly looking at ways to improve your workflow.

    Thanks for the inspiration.

    • Jared Banks

      Larry, you’re welcome! Thanks for sharing your story. I think what Link would say is cut the fat, not the muscle and bone. Which again is the concept we’re all circling around.

      Old gibberish CAD layers are fat, new 3D specific Layers are probably muscle and bone.

  5. Wolfgang Fitz

    Hi Jared
    For us it’s impossible to drawn without Layers and Layer-Combinations. Normaly we have nearly 100 Layers and 50 Layer-Combos per Project. They are really important to manage all our different Layout.

    by the way: i did this oldschool, physical “Layerthing” too and i’m happy this Time is over … 😉

    (sry about my english)

    • Jared Banks

      You did the old physical layers! 🙂 Could you share some of your less typical Layer Combinations? Beyond plan, section, elevation, detail, 3D model, electrical, other consultants, what sort of other combos do you (and Link) have?

  6. Florian Techel

    The layers vs. no-layers debate always appeared to me as this “invented here”-“not invented here” phenomenon. If my software contains layers this has to be good, and if my software does not support layers they must be of the thorny one. This makes it odd for market-leader aficionados as their one application does it and the other one does not. 🙁

    I always found layers as a means of filtering/structuring/viewing a document. The more of these tools the more different views it permits, the happier I am.

    Seriously, the mess with layers in LaLaCAD stems, in my view, from the incomprehension of the topic by most users and preciously few layer-management tools in the software (I believe version 2015 made some moves in that direction). As a result drawing elements usually are all over the place (layers) and Stop Making Sense (Talking Heads).

    The one thing that truly annoyed me about ArchiCAD at first (I reviewed it in version 4.55 for the first time) were the allocations of layers to tools (or the other way around?). I felt incredibly limited in my abilities. It took a while for me to realize that these allocations were not chiseled in stone but actually created “some” organization that I was free to change but until I did that the file would be at least organized in some way.

    A friend of mine runs a rendering service for architects and gets his files from many applications but agreed that the ArchiCAD files are among the least chaotic.

  7. Nathan Hildebrandt

    Jared layers being a piece of metadata and as a sorting tool can always lead to many different approaches and philosophies as can be seen by the previous comments. We have always taken a reasonably logical approach with our Layers / Layer Combinations based on documentation output, purely due to the deliverables being drawings and in many instances .dwg files. The only way in which we could control output and allow our collaborators to understand what lines meant in .dwg was to apply a logical layer to it.

    I began to undertake some research over the Christmas break because I want to ensure that any future direction that I take follows a standard. Now I need to point out that in Australia there is currently no formal standard for layers or BIM Standards. So with Australia being a member of the Commonwealth and with the UK being the leader in BIM Standards I consulted with the number 1 BIM guy in the UK – Rob Jackson. He pointed me to this document on the AEC UK Website – It is the standard for Layers when collaborating in BIM. One read of this document and it is very confronting and you immediately want to put it down. But if you take a step back for a minute and think about it, it is really logical. How in the world could this be logical you ask, with potentially 1000 layers? I will explain why I think there is great value in what the AEC UK guys have done.

    Ok the first thing you need to do is disengage the whole layer thought process. Then you also need to understand that the naming of the layers are a conforming BIM Classification system. If you want to collaborate with someone in BIM you will need to add this data to each element somehow so that they can identify what the element is. If you decide to stick with your normal layers you will have to add this data manually as part of your IFC Data. Not only is this time consuming and painful but if you open up an IFC file say in Solibri Model Viewer you will see Layer and it means nothing. So imagine your user needs to select a layer for an element, why shouldn’t that layer be the IFC Classification that you would need to add to the element anyway. You are removing the need to have to add it manually elsewhere and your layers when exported will have more meaning than ever. Also imagine the possibilities of layer combinations and exporting the elements that you need too.

    The thing to note is that it is pretty complex and Rob’s comments back to me was that a practice would start with all layers, a BIM Manager would delete the irrelevant ones to their practice and then save a template from it. So it would take a little bit of getting used too but I see it as an easier step than making people add more manual BIM Data to their models.

    I can imagine that a sole practitioner is reading this and thinking I am only doing houses what does it have to do with me. I would pay a lot of attention to the Local Authorities and their plans to essentially capture the whole built world in the virtual. There is plenty of this work happening in Singapore right now and New Zealand are also leading the charge. So BIM will need to be embraced by all in the Built Environment and the sooner you make the adjustments the easier it will be once the data requirements become greater.

    So some food for thought.


    • Jared Banks

      Nathan, I just scanned through that document. Very interesting. I think to go that route it takes a psychological leap to see Layers as something different, as not just CAD in ArchiCAD—which was my hope with this discussion and these articles. To follow that IFC data concept for Layers as proposed by your comments and that document is intriguing and for anyone pushing IFC, it’s probably hard to argue against. And for anyone interested in learning about IFC, this might also be a great starting point. It requires a different segmentation of data via Layers because then it’s not about simplicity of viewing data (for saved Views and Layouts) but really about data organization for BIM focused sharing and collaboration.

      My biggest question though, is will it also force a shift in how one models, if an element belongs in two classifications and thus needs to be broken up? For those that are IFC curious, but not IFC focused or sharing IFC yet, might this be a step too far? Or perhaps they shouldn’t worry about that yet, and just start learning to live in an IFC layer based world…much to think about.

      I should talk with Rob and do a post that explores this IFC-Layer concept. It might be too much for people to grasp at the moment, but then again perhaps it shouldn’t be. And it isn’t that hard. All it takes is thinking about each element in the IFC spectrum, which is valuable. It helps classify and organize elements as digital approximations, as things to be built. Which is pretty cool.

      Combining your comments with Ben’s below, it makes me wonder: what if Layers just merged with IFC classification and just became the same thing. Or two views of the same thing. I guess with IFC mapping this is 100% possible today (and I assume what Rob sets up).

      • Nathan Hildebrandt


        It will be a big shift for people but once you get your mind focused on data deliverables you will want to find the quickest and easiest way to apply the data. But it will be possible and it will be easy after a transition.

        Regards an elements Classification, the system is set up so it will only belong to one, it is that clear it will never belong to two and people with a little guidance would be able to understand how the system works. It just needs to be a global standard not per country, we all work around the world now, there doesn’t need to be heroes in each country creating new standards, just adopt and move on.


  8. Ben Frost

    Organisation, organisation, organisation. Did I say that word enough? 😉

    I have advocated many times to Graphisoft in the past (on the ‘TALK forum and in the VDT/GDC), for the ability to have hierarchical layers: not because I like layers but because that’s how building elements are organised in most classification systems the world over. Imagine being able to control complete classes of your model elements by showing/hiding a parent layer?

    Layers – well organised layers – go a long way in aiding in good ArchiCAD practice as well: most ArchiCAD-using Architects know more about buildings and how they go together than they do about ArchiCAD, (we geeks are the exception, not the rule!). A familiar layer name in a list of other building-centric names – and in an expected place in that list – means a user gets the element in the right place more often than not.

    You may not believe it, but your non-ArchiCAD Revit-using friends already do this, they just don’t know it. Their ‘layers’ are represented by a classification system hard-coded by the big A. Immovable, un-editable, inflexible and well over a 100 of them. They do have hierarchy going for them though…

    One of my directors controls the number of layers in our practice template by the age-old, tried and tested ‘if the layer dialogue box when fully stretched from the top of my screen to the bottom cannot display all of the layers at once then there are too many’ method. I found this out by adding a couple of essential layers and being barked at the next time he was in the project. My response? Get a bigger monitor – they all fit on mine!

    I’m with Link – don’t fear the Layer!

    • Jared Banks

      I like the bigger monitor solution! I have to admit, I am a fan of having everything in a list (or info box) visible on the screen with little to no scrolling. That is probably one of my semi-conscious reason for paring down like I do.

      Also I couldn’t agree more. Hierarchical Layers would be amazing. I could do so much with that. Wouldn’t it also (potentially) provide the answer to versioning and schemes? Turn on/off the parent where the parent is Scheme A or B or C. We can do this a bit with renovation filters, but this would be nicer. Perhaps. But again that goes to my point of what if it’s a non-layer solution…

      • Ruben Van de Walle

        Great article Jared! – And a lot of interesting comments…

        The idea of hierarchical layers is somewhat used in the way HLM’s have a Master Layer. Actually resembling the way Autocad blocks work in a big way…

        It’s also how I personally use layer-extensions often. In combination with HLM’s this is a very powerfull use of layers in my opinion.

  9. Jared Banks

    Can I just reiterate that I am loving all these comments! Keep them coming. So many great, viable, and high quality Layer Concepts here. Thank you all.

  10. Link

    I am envious of you Nathan (and Rob) being in a place where BIM is being pushed hard enough that you can entertain the idea of using IFC layers in your template.

    Just as a background, our ArchiCAD ‘territory’ consist of 8 (gorgeous) states in the US: Idaho, Montana, Utah, Wyoming, North Dakota, South Dakota, Colorado, and Nevada. This is an area of over two million square kms (over 8 times the size of the UK), yet with only a population of 15.7 million. In this part of the Intermountain West, BIM uptake is astonishingly slow. Our inside joke is that IFC stands for I’m Freakin’ Clueless. This is a part of the country that is probably most resistant to change apart from maybe the deep South. All this in a country still using imperial measurements. American the superpower indeed. BIM seems more popular on the coasts, and there are certainly people implementing it, but across the board the US is way behind Europe and Australasia.

    We do what we can to push BIM, but we also know our market very well and realize we need to balance the old with the new, so baby steps are in order. Incorporating data rich elements into our template that produce amazing schedules that can also be referenced by labels and exported with an IFC model are a good start. Incorporating COBie is noble at best, as most people start looking around for a dog when you say the name. Jokes aside, this is the reality. Sadly incorporating any of a thousand IFC based layer names is out of the question, for now at least. If I renamed my A-CEILINGS.M layer to A-G5411-M-LightingInternal-Rfl, or A-FURNITURE.M to I-G42-M-FittingFurniture-Fwd I would be out of a job. Users generally think in architectural terms not IFC code.

    I am not intimate with how ArchiCAD layers translate into IFC layers and how much of a part they play in the resulting data exchange (hopefully this is something that will be covered in an upcoming blog post), but it seems that until BIM/IFC is more mainstream here that some kind of as-needed layer translation would be more realistic.

    So soldier on, pave the way, share your ideas and I’ll bring up the rear. Ugh.

    In the meantime, I have always liked Ben’s idea of hierarchal layers. And I’ve supported him over the years pushing Graphisoft for them. Renovation status reduced the layer list quite considerably. Now the biggest duplication of layers I come across is for 2D and 3D versions of the same element type. Hierarchal layers could certainly help with that, plus a tonne more.


    • Jared Banks

      Well said. I’ve already warned (threatened?) Rob that I’ll be pestering him about the whole Layer as IFC data concept. I agree, it feels far from my everyday work, but I want to understand it nonetheless. Perhaps it’s not as far off as we assume. Or perhaps there’s a way we can prepare our users by a Layering system that secretly supports the shift. So perhaps the first step is to understand the IFC as Layer so that you can make sure your Layer creation methodology could be mapped easily in some future iteration of your template to the land of IFC. So by your example, you don’t need to rename your Layers to the IFC lingo, but knowing your Layers have the proper analog in IFCland means you’re doing it right and future proofing your work—thus when a client calls in a panic you can say “MAGIC! Let’s use this .aat file to rename your Layers, add this mapping, and BAM! IFCmagic”.

      Mostly for my future reference, but it applies to the comments above:

      known knowns, known unknowns, and unknown unknowns:
      future proofing:

    • Nathan Hildebrandt


      Australia is as far down the track as you can imagine. There are only a few pushing the boundaries and the reality is that the boundaries are not far enough as far as I ma concerned. I started looking into the UK Standards as a bit of a research project to see what a country that has its s*&t together is doing. Australia is a mess with no standards in place and a government that thinks that industry should lead where BIM goes and we all know what that means.

      I have to admit that if I rolled out the UK BIM Standard Layers tomorrow in my practice it would be super disruptive. With a lot of explanation and guidance I think that people will see the value and change their mind and see that there are benefits to recognising that there are benefits to making a minor adjustment.

      It is something we should definitely talk more about in the next few months. I am going to experiment on myself and see how I go with it, then test on a team in the office. It will open the door to 4D and 5D BIM processes without having to add extra data or have a lot of favourites. I see the time saving benefits in dealing with data that way.

      People will get there they just need to see the benefits. If a few of us lead the way and demonstrate how they need to make a minor shift to make a huge leap in BIM Collaboration we will see the wave and it will be a big thing for all ArchiCAD users.

  11. Damian

    Revit works without layers, so obviously is doable.

  12. bruce

    Having used archicad since v4.1 I can attest to the variations of default layering (among other attributes) over the years, ironically seemingly now abandoning the NCS in favour of the simpler ‘Queen’s English’ of simple terms used originally.

    The number of options to control imaging and the variations possible suggest a couple of critical practical concerns:
    1) cumulative assembly & variations of past ‘I’ (information in BIM) & the risk of file corruption
    2) need for comprehensive technical knowledge & strategic zeitgeist for staffing up if using this increasingly complex tool with potentially disastrous consequences & liabilities

    I have long contemplated (wishlisted) an holistic ‘SANDBOX’ import capability for more than dwg (esp objects) into worksheets, and wonder how others may do this. If anyone has ever inadvertently ‘polluted’ a layer structure accidentally dropping in a legacy object (custom or otherwise) they may understand one can potentially make a project unusable in a few seconds.

    To this end I try to retain the base default archicad layers (and all attributes) as much as GS may make possible between versions, in the past using an attribute purged intermediary file for ‘cleansing’ & reuse of prior version building information model assets. I have felt this more of a ‘workaround’ than elegant efficient process. Any suggestions would be most welcome?

    Additionally the debate on layers, pens, etc has been long raging, with debatable efficacy in terms of clarity and efficiency of cost & time, and arguable empire building:

    I have been around long enough to have worn mylar smooth with changes in manual pin bar drafting. Every now & then I fire up the early archicad, to remind myself of how simple it used to be, even though at the time seeming complex in comparison to more universally understood manual drafting…

    I continue to ask, like with the most recent simplification of the layering default, if or when GS might focus less on adding yet more features & refine the software to make the toolkit more transparent & the work easier, as a means rather than feeling some times to have become, with a now ~3,500 page manual, an end in itself…?

    • Jared Banks

      Bruce, one thing that helps with DWGs (as I’m guessing you know) is to place them like PDFs, rather than merge/import them directly to the file (in other words just drag and drop the file from your desktop folder to the place you want it in ArchiCAD). You can then turn layers on/off from the placed DWG and get it to display jsut as you need. Of course if you need to delete elements, etc. that’s a whole other story. In that case one could edit the DWG in another instance of ArchiCAD and just update (automatically) the placed dwg.

      As for legacy objects…I’m not sure. Hopefully others have advice!

      This is a good topic that deserves its own post.

      I have another post brewing about a great solution for keeping imported data from adding garbage to a file, I’ll try to get to that ASAP—the trick is to first import the data into a file with as little information in it as possible, so any garbage because apparent immediately.

      Thanks for the comments! Much appreciated.

  13. Chuck Kottka

    I continue to defend traditional layer sets (we still use the NCS/AIA CAD Layer Guidelines, like A-WALL-INTR), though I understand and applaud the effort. As much as I try to clean up the template, evolution is glacially slow. When you have a gaggle of employees (about 50 active designers in 4 different divisions), democracy actually restricts progress. These are some of the reasons that wholesale Layer migration isn’t yet viable:

    1. When exporting DWG files, Layers are THE metadata. There are very few other options. That is how elements are identified. Since there are still many AutoCAD users out there (still #1 design platform), including engineers, owners, facility managers, and (sub)contractors. It is the only format that can be opened and edited by almost ALL software (including free options like DraftSight). Even when I work with new consultants, I’m confident that they know what a rectangle with the E-LITE-EXST means. If smooth communication is a primary concern, then this is paramount! Perhaps if ArchiCAD’s DWG Export Translators had more options for constructing layers out of parameters, we could avoid these extra identifying layers.

    2. If your firm wasn’t born yesterday, then you have a history. We’ve been using ArchiCAD since 1989, so we have a LOT of archives, many of which are still valuable assets that get copied into new projects. Though we’ve considered updating old archives on a regular basis, it amounts to thousands of files that may or may not ever be reused… and zero billable time to manage them. Add to that parts and pieces that are constantly copied into new files, and you have big legacy issues. With the advent of the Renovation Filter, and subsequent loss of -EXST and -DEMO layers, it has been painful enough to keep clean. A whole new layer system makes you incompatible with your own history. Not a deal breaker, but it has to bring a huge payback to be viable. OCD is not enough reason. Perhaps if ArchiCAD had a Layer Standards Conversion tool…?

    3. I have been fighting with users who demand we keep extensions. It’s not that I don’t like extensions; they are great for organizing the model. However, ArchiCAD has another limitation: The Layer module in the infinitely wonderful Info Box is limited to 11 characters, or else it is truncated, and made illegible. A-WALL-INTR fits nicely, but A-WALL-INTR.3D does not. This is also why I prefer the universal 4-letter abbreviations over whole english words. Many construction terms are simply too long. Again, a small tweak of ArchiCAD would eradicate this.

    4. Any Layer system has limits. There is always an exception to the rule (my firm excels at finding them), a special case where more layers are needed. So users must construct new Layers. Without some guidance, a structure that they have to follow, they can create utter Layer chaos. These are designers, each a delicate snowflake, no two alike, so they can each destructively overcomplicate a system in their own unique way if given too much freedom. So I limit them to X-XXXX-XXXX, preferably using the NCS list on our intranet. It keeps me from having to clean up Layers like “A-Stuff”, “D4010-084613-Dr_Hdwe.extr”, and “Jeffs Wall Below The Parapet But Above the Transom”, especially after they read that article from the UK… Yikes! A little consistency is comforting.

    5. As great a system as IFC has created, it is still hard syntax to understand for non-programmers. The vast majority of Architects spend their whole lives understanding construction and building terminology, not to mention the lingo of their respective markets. We can’t control those things, but we can control what software we use. We want applications that bend over backwards to understand OUR language, not the other way around. Until CAD/BIM can speak in natural Architectese, I don’t consider any of the alternatives to be better than NCS. IFC is a great database structure under the hood, but keep it there. We just want to see the beautiful dashboard!

    6. I’ll reiterate what some have already said before, in response to Revit needing no Layers… yes it does. It has Types and Instances and Filters, many of which you cannot edit. Everyone needs a system, at least I can adjust mine. Even if it is glacial.

    Perhaps in the afterlife, we will need no Layers. But for now… Cheers!

    • Jared Banks

      Chuck, fantastically wonderful comment and a great counterpoint to all of the rest of us. Managing that many users with that much history is definitely a challenge most of us don’t face. You make a compelling argument for the classic X-XXXX-XXXX naming system.

      I think what is clear from all of us is that we need a system for Layer naming, and it needs to be adhered to. And depending on the system and the firm, that suggests different systems and different management processes. Great to have you describe another solution.

      I am tempted to include “Jeffs Wall Below The Parapet But Above the Transom” as an Easter Egg layer for my template moving forward. That is just too perfect of a ridiculous Layer name. Unless it became: “Jeffs Wall Below The Parapet But Above the Transom_Current_Not Used-Active.2D 02042015″

      A layer Mapping Add-On would be great and super valuable for dealing with old projects. I’m going to e-mail a GDL wizard who’s been looking for a challenge. I will also aim to write about another solution next week.

  14. Scott Graham

    I transitioned my firm to using non-abbreviated layer names a few years ago. It’s still hard enough to just get our people remembering to keep layer assignments in mind without throwing arcane abbreviations at them.

    I do like the idea of IFC analogous layers, and will be carefully reviewing IFC classes before finally rolling out our new AC18 template (yes, I am woefully behind).

    I currently run 157 layers, and am going to shoot to get that down to 100. I think some careful thought will create a more flyweight, but just as efficient, layer list. The bulk of my layers are actually annotation though, not modeled elements.

    Great conversations and points all around though!

    • Jared Banks

      what sort of annotation layers do you have? It sounds like you have more than the typical dimensions, texts, and markers.

  15. Carmel

    It is impossible to produce a set of Architectural drawings from Archicad WITHOUT layers. Anyone who says that it is possible with only a few layers is not producing a typical set. I use different layer combinations for different types of plans from the same story…. such as floor plans, reflected ceiling plans, site plans (at various scales), enlarged plans etc…. Text needs to be often duplicated so that it looks good at different scales. Also… the IFC official layers is a complete non starter. It’s hard enough to try and have keynotes reflect the CSI format!
    Perhaps with future releases of Archicad we can rely on less layers… such as the introduction of the Renovation tool (love it!)


Submit a Comment

Your email address will not be published. Required fields are marked *

Looking to meet other ARCHICAD users? Why not come along to one of our upcoming User Groups!  VIEW DETAILS