creative design, custom houses, and the limits of Chief?

[Lighthouse]

I'm interested in the insights of advanced CA users re their workflow in highly custom designs.   Part of my work is designing very modern houses that are somewhat sculptural in shape, with may different floor levels and unusual roof lines.  The interiors often feature clerestories and lofts.  Sometimes I'm starting with an existing house which is incorporated into a new building, sometimes it's from scratch.  The common demoninator is that I have no idea what the house is going to look like when I start designing.  My process is to iterate very quickly, essentially sketching with Chief like I used to do on trace years ago.  I want to pull and push on the walls and roofs while constantly viewing the model from various angles until the shape I'm looking for starts to emerge.  Typically I'm doing "save as" when a plan has a few elements I like but I want to try another direction on some part of the house, while retaining parts I like.  On average I usually go through 10-20 major iterations and maybe another 20 minor ones.

 

The problem with my method is that all this pushing and pulling on the model (manual roofs, manual wall edits) leaves me with a hopeless kluged-up model.  A couple of Chief wise men (I will let them identify themselves if they so choose) have told me that my method is fundamentally incompatible with Cheif, because I am not establishing defaults at the outset, and all the manual editing of roofs and walls creates screw-ups that can't be undone.  Accepting the wise mens' word as gospel (which I do), what is to be done?  I can only think of three options:

 

1) use my method to arrive at the design, then start from scratch and recreate the entire model with defaults and auto-roofs derived from the finished plan info

2) do my initial modeling in sketch up or another program, if it has to be recreated from scratch anyway

3) continue my method and accept the fact the my finished model is going to be full of patches and polyline solids masking the kluged up parts

 

Anyone have any suggestions or insights as to how to address this problem- in other words, how do you use Chief if the house is going to be uncoventional and you want to sculpt the shape as you go?

 

Thanks for your thoughts!

camera.pdf

[Rashid_Garuba]

David,

 

I asked the same question a while back... can't find the thread. Same answer... Chief doesn't like too much structure tinkering after defaults are set.... "Learn the Chief Way". It IS the truth though.. that is the way Chief is and it is not going to change as Chief is quite mature... Like marrying and expecting change.

 

It is a bit of a bother as I find it hard to discourage clients from revisions, especially early in the process.

 

Get good with Sketchup or do your design schematics on paper BEFORE getting into Chief.

[Lighthouse]

paper!    radical   Well, I don't know if I'm out of practice or was just never a great sketcher to begin with, but I find it much easier to visualize complex 3d with Chief than trying to draw it.  So maybe SU is the way to go.  My problem is that if I'm not using it (SU) every day, I forget how to do it.    Too bad Chief doesn't have a 3d modeling mode similar to the 2d room planning bubble diagram, where you could easily manipulate shapes and then turn them into more defined elements

[johnny]

I had the same issue when I started - only to come to the realization Chief simply doesn't work for this type of process.  You need to use something like Sketchup to start (I use Vectorworks).  Some people will say they can "design" fine using CA, but when I do I get too wrapped into the minutia of the details.  What I want is a free flowing method like Sketchup/Vectorworks has - but even they too are a bit restrictive compared to truly sketching.

[dshall]

........ Chief doesn't like too much structure tinkering after defaults are set.... .....

 

Nuts......  CA is very flexible when it comes to changing things after defaults are set...  and I should know because I don't set defaults until the project is done...... if I do set defaults.

 

I do know of one item that is not easy to change later on in the model building phase.  

 

So Rashid,  you made a statement,  do you want to back it up with examples of what defaults cannot be changed later on in the modeling process.  It is very easy to make a statement,  so if you are to make a statement,   please be prepared to back it up with examples.

[dshall]

I had the same issue when I started - only to come to the realization Chief simply doesn't work for this type of process.  You need to use something like Sketchup to start (I use Vectorworks).  Some people will say they can "design" fine using CA, but when I do I get too wrapped into the minutia of the details.  What I want is a free flowing method like Sketchup/Vectorworks has - but even they too are a bit restrictive compared to truly sketching.

 

Yep,  I get this Johnny. I do design in CA,  but it is not the best method....  I do not understand SU and VW when it comes to designing, but I bet they are better tools than CA when it comes to designing.  

 

Pen and paper is better for designing than CA,  and I can only assume that SU and VW are much better tools than CA for designing . 

[DavidJPotter]

Method and habits may become confused. I learned from Dan Baumann and I ascribe to his philosophy from the standpoint of planning out your design before you actually start to create it. By that I mean sketch out the floor levels using 2CAD and dimensions so I already have a good understanding of floor levels, floor platform thicknesses etc. Then I set ALL applicable defaults (it saves time and clicks later on). Then you start actually creating the 3D model. Often times as the design takes shape, certain fixed ideas I or the client had fall away and evolve. There is no system that would fit all users and clients. The key is taking the time to fully understand what you are doing and how to best use your tools and then enjoy the process. Let the nit-pickers waste their time worrying. There are some constructs that are more difficult than others (requiring thinking out of the box), a great example of this is the work of Yusuf in Ethiopia. He has figured out ways to create effects that others missed using Chief. DJP

[Rashid_Garuba]

So Rashid,  you made a statement,  do you want to back it up with examples of what defaults cannot be changed later on in the modeling process.  It is very easy to make a statement,  so if you are to make a statement,   please be prepared to back it up with examples.

 

Easy there Scottie... .. I did not kidnap you first born.. I did not say defaults cannot be changed either. I am referring to floor structure which is not straight-forward unless you know what you are doing.  Floor structure HAS driven me nuts... not initial setup but adjusting heights afterwards, especially forgetting defaults that may affect changes I am attempting..

[BryceEngstrom]

I think I've actually learned how to use Chief in what I would describe as a "sketchy" way but I agree the model can become a mess.  For more complex designs, I often get through most of the schematic design phase, and then actually reconstruct the model from scratch based on the schematic design.  As long as your walls are plumb, I think Chief is a great way to go.

 

I learned to draw and sketch pretty well in college but I honestly find it a huge waste of my time, and my clients money, in today's world.  Just my two cents.  I know a LOT of architects are still very passionate about the centrality of sketching by hand.  I think the most important thing about learning to draw is learning how to SEE.  But once you've learned how to see, you don't necessarily go back to being blind because you don't sketch any more.

[yusuf-333]

Method and habits may become confused. I learned from Dan Baumann and I ascribe to his philosophy from the standpoint of planning out your design before you actually start to create it. By that I mean sketch out the floor levels using 2CAD and dimensions so I already have a good understanding of floor levels, floor platform thicknesses etc. Then I set ALL applicable defaults (it saves time and clicks later on). Then you start actually creating the 3D model. Often times as the design takes shape, certain fixed ideas I or the client had fall away and evolve. There is no system that would fit all users and clients. The key is taking the time to fully understand what you are doing and how to best use your tools and then enjoy the process. Let the nit-pickers waste their time worrying. There are some constructs that are more difficult than others (requiring thinking out of the box), a great example of this is the work of Yusuf in Ethiopia. He has figured out ways to create effects that others missed using Chief. DJP

David, thank you very much!! Many men are really grateful to your kindly treatments. You are so kind and encourager to all of us here. Thanks again for addressing my works as an example David.

[Rashid_Garuba]

Here's to you Yusuf!... There is no doubt one's mindset of thinking of problems brings more problems and a resistance to learning new ways. Thinking of possibilities (open mind) is the core of progress.. I think we get more rigid as we get older...

 

[Lighthouse]

I think some of this could be addressed by the programmers at Chief simply stating what the organizing principles of the software are, and how they can be worked around.  Obviously that could get complicated fast, but I would appreciate a list of basic principles baked into the programming.  What I'm suggesting is a little different than reading the manual, which of course is an expression of the programmed principals.  But for example, the statement could say something like (these examples are not meant to be true, but just examples):

 

1) the primary function is drawing walls, which create rooms

2) rooms are understood to have walls, floors, and ceilings

3) ceiling heights determine floor levels based on platform thicknesses, which default to X unless changed

etc etc

 

Then the more interesting part would be:

1) a ceiling will always overide a wall change

2) walls will follow a roof up or down in these particular instances (give examples)

3) a manually created roof will not form a proper valley unless x or y

 

Clearly the program has a particular logic based on choices that the programmers made, but the logic is not visible except in practice.  The logic is discovered by users, who often post both the limitations and useful work-arounds.  However, this is a somewhat ad-hoc, hit or miss way to inform all users.   For example, Scott seems to suggest that a complete understanding of the program would enable me to work in the way that I would like.  I would like that to be true, but there must be a more efficient way to get to this stage of advanced knowledge than simply spending thousands of hours making mistakes and then learning how to fix them.  Does anyone else beleive that a concise, organized statement of the programming language and assumptions, apart from the user's manual, would be useful for them in order to develop a working method that enabled them to use Chief to it's maximum advantage?

 

To my mind, this is similar to any design program I encounter- my first question is: what do I want to acheive and what are the parameters and limitations that would inform my choices?  If I knew for certain (and not just because someone posted it on Chieftalk) that I should always establish floor heights higher than I might want, because it's easier to move them down than up, I would do that.  But I tend to see things offered as "best practices", which imply that better outcomes are more likely with this method but other methods might still work.  This is not as specific as actually describing the logic assumed in the programming.   Is there a reason to not clearly state the programming logic?  Is it too complicated?  Trade secret?  Not good from a marketing standpoint?  Time consuming and not interesting to most people?  To be clear, I'm asking Chief Arch to do this, not the user community

[VisualDandD]

I do this all the time.   That is how I design many of my projects.   I use chief to "sketch" and I dont worry about all the 1/16"s etc.   I am just sculpting.   Sometimes, I even alter room DBX's to create offsets in roof planes very quickly..etc.

 

When I start to get close, and am happy with my shape, I turn off all the layers that are not important, and do a cad detail from view.   I now start a new plan (from one of my save as templates that best fits it) and I redraw the house.

The CAD overlay makes this VERY quick.   This time, I know where all my important points are (eg important centers, and things that need to align or offset).   I now bring the model into dimensional accuracy and eliminate all the odd dimensions.

All told, I can redraw a model in less than an hour in this method.   I open the original and copy over areas I spent a lot of time on (cabinet layouts in kitchens..etc).

 

I cant tell you how useful this is to me.   I am working on a new personal house and the versions just keep piling up.   When I am finally happy, I will just redraw.

 

EDIT....I attached some pics that show the evolution of a project.  These are all basically the same plan.   It is still in flux with me doing the "dirty" drawing right now.

[Lighthouse]

the audi commercial reminds me of one of my favorite cars, my audi "super 90" sport wagon.  there were only a couple hundred imported to the US back in the early 80's, and only to the west coast, so I never saw another one in the Northeast where I live.  It had a center brake rotor for both front wheels, which befuddled many a mechanic when I would go to have the brakes done.  They would come back with a serious expression and tell me that the reason my brakes didn't work well was that I had no rotors!   I think audi gave up that particular innovation fairly soon

[yusuf-333]

I'm interested in the insights of advanced CA users re their workflow in highly custom designs.   Part of my work is designing very modern houses that are somewhat sculptural in shape, with may different floor levels and unusual roof lines.  The interiors often feature clerestories and lofts.  Sometimes I'm starting with an existing house which is incorporated into a new building, sometimes it's from scratch.  The common demoninator is that I have no idea what the house is going to look like when I start designing.  My process is to iterate very quickly, essentially sketching with Chief like I used to do on trace years ago.  I want to pull and push on the walls and roofs while constantly viewing the model from various angles until the shape I'm looking for starts to emerge.  Typically I'm doing "save as" when a plan has a few elements I like but I want to try another direction on some part of the house, while retaining parts I like.  On average I usually go through 10-20 major iterations and maybe another 20 minor ones.

 

The problem with my method is that all this pushing and pulling on the model (manual roofs, manual wall edits) leaves me with a hopeless kluged-up model.  A couple of Chief wise men (I will let them identify themselves if they so choose) have told me that my method is fundamentally incompatible with Cheif, because I am not establishing defaults at the outset, and all the manual editing of roofs and walls creates screw-ups that can't be undone.  Accepting the wise mens' word as gospel (which I do), what is to be done?  I can only think of three options:

 

1) use my method to arrive at the design, then start from scratch and recreate the entire model with defaults and auto-roofs derived from the finished plan info

2) do my initial modeling in sketch up or another program, if it has to be recreated from scratch anyway

3) continue my method and accept the fact the my finished model is going to be full of patches and polyline solids masking the kluged up parts

 

Anyone have any suggestions or insights as to how to address this problem- in other words, how do you use Chief if the house is going to be uncoventional and you want to sculpt the shape as you go?

 

Thanks for your thoughts!

Surprisingly my method and yours have 0% deviation and that is how I use it exactly. You left nothing for me to add even the number of personal archived files and I put them in the order of stage 1,2;stage 2 1.... And some times I have special folder for some unintended outcomes during work, that obvieously help for an other job. I call that files "ideas folder from PR x stage y....I said the same thing to yours in an earlier thread, I can't remember. This method some times would lead you get frustrated and put me in an endless ocean. But I suggest it, because it isn't only important to learn the programme but also let's you develope a natural understanding to architecture and design by default time after time.

[yusuf-333]

Here's to you Yusuf!... There is no doubt one's mindset of thinking of problems brings more problems and a resistance to learning new ways. Thinking of possibilities (open mind) is the core of progress.. I think we get more rigid as we get older...

 

Rashid thanks for this inspiring pick.

I like itit really.

[johnny]

Again, since this thread it talking about the subject, I thought I would show you a quick reference on 2d shape to 3d shape to 3d model to 3d structure video in Vectorworks.

 

Please forward this below video to 13:04 and see how this process works in Vectorworks.  btw...even though VW is much faster at concept work, detailing in 3D later in the project flips to Chief Architect being much MUCH better for residential projects.  This is why I am trying to force myself to learn CA.

 

[KevinWaldron]

You should probably check out FormZ 8......... fairly easy to draw as you are inspired and few of the limitations that Chief forced upon you..

 

Blessings,

 

Kevin

[johnny]

You should probably check out FormZ 8......... fairly easy to draw as you are inspired and few of the limitations that Chief forced upon you..

 

Blessings,

 

Kevin

 

Wow.... i've heard of FormZ but never seen a video.  The few I just saw where very impressive - looks more intuitive than SU.  Here is a 5 min. creation of a Wright house.  I'm going to check that out thanks.

 

[Kbird1]

Here's to you Yusuf!... There is no doubt one's mindset of thinking of problems brings more problems and a resistance to learning new ways. Thinking of possibilities (open mind) is the core of progress.. I think we get more rigid as we get older...

 

 

 

 

I don't like Ads much ...but I like that one....  nice reminder of the WRC of old too...

[Lighthouse]

Wow.... i've heard of FormZ but never seen a video.  The few I just saw where very impressive - looks more intuitive than SU.  Here is a 5 min. creation of a Wright house.  I'm going to check that out thanks.

 

I remember formZ from many years ago- was popular with furniture designers.  I'm glad to see they are still at it!  I think you are right- a pure modeler is probably the way to go, and then recreate in CA once design forms are decided on.  Certainly it can be tweaked in CA for finish details and client changes.  So painful for an old dog like me to learn a new program.  I'm curious if anyone has used both SU and FormZ- I'm interested in your comment that it seems more intuitive- they seem very similar to me, although SU seems to have more of a line and plane basis, and formz is maybe more form oriented?   I do love the powerful "follow me" tool in SU, I wonder if formz has something similar.

[yusuf-333]

I do this all the time. That is how I design many of my projects. I use chief to "sketch" and I dont worry about all the 1/16"s etc. I am just sculpting. Sometimes, I even alter room DBX's to create offsets in roof planes very quickly..etc.

When I start to get close, and am happy with my shape, I turn off all the layers that are not important, and do a cad detail from view. I now start a new plan (from one of my save as templates that best fits it) and I redraw the house.

The CAD overlay makes this VERY quick. This time, I know where all my important points are (eg important centers, and things that need to align or offset). I now bring the model into dimensional accuracy and eliminate all the odd dimensions.

All told, I can redraw a model in less than an hour in this method. I open the original and copy over areas I spent a lot of time on (cabinet layouts in kitchens..etc).

I cant tell you how useful this is to me. I am working on a new personal house and the versions just keep piling up. When I am finally happy, I will just redraw.

EDIT....I attached some pics that show the evolution of a project. These are all basically the same plan. It is still in flux with me doing the "dirty" drawing right now.

Very nice of you. Thanks That is how chief make easy to develope ideas one after the other, until you decide the stage that best for what you wanted from your inside.

[Lighthouse]

Wow.... i've heard of FormZ but never seen a video.  The few I just saw where very impressive - looks more intuitive than SU.  Here is a 5 min. creation of a Wright house.  I'm going to check that out thanks.

 

Also, re SU- I saw a demo years ago where they took a house photo and pinned it to a primitive object and the object "snapped" to the geometry of the house in the photo.  It seemed like an incredibly powerful tool for working with a client on preliminary design where you could photo their house and instantly begin to show design possibilities-  has anyone used that?

[Kbird1]

Formz , does look interesting , I also watch the "30 minute House"  video ,bit of a cheat , (there's some fast forwarding) and  the guy is obviously a Formz expert but definitely worth a watch , Formz Jnr is about the same Price as SU Pro , the Raytracing is a $395 addon for Both Jnr and Pro.  

 

https://www.youtube.com/watch?t=35&v=0n0FGm3kw2I

[Rashid_Garuba]

This really illustrates the power of marketing and positioning.. A lot of these apps are as powerful (or more) as Sketchup AND have been around longer BUT Sketchup resonated much more with Architects... "Sketchup" sounds a lot better than "Formz"... Also didn't hurt that Google made Sketchup free.

 

Utlimately I see a lot of Modern architecture that is a lot less ornate with all straight lines.... The devil is in other types of houses (High end US) AND construction docs.. For massing studies they sure are good though..