BIM! WHAT MORE?

[Anonymous]

This is meant to be the twin thread of "BIM? WHY?" for those who are in favour of it and want to share their view on how to improve it. Those who are not in favour of BIM may still read it, and mouthwater as they change opinion...

Then the question is, what more would I want in it? This is my answer so far.

- include a section on topography that would serve the obvious purpose in design: when drawing a slab, AC also draws the foundation by following the actual land, on the fly. This is not easy, as one may want to have two or three connected slabs, as they follow a hill.

- AC is able to draw all sorts of roofs on the fly. Let this feature be smarter. Given the geographical north and the slab (see above), draw the roof so that it is oriented towards the maximum exposure to sun light, for the purpose of solar panel installation. This would solve a pesky problem with all new constructions, at the click of a button.

- let interior design meet its next generation cad by adding constraints on object placement, constraints that apply locally to any designer in the studio. Let call the collection of these constraints, a style model. For example,

- I want all beds to have the head towards north.
- I want no less than 1.2m around any master-bed.
- I do not want to see the dining table in front of the main entrance door.
- I do not want to see the toilet when I open the bath-room door. For example, if the door opens to the left, I want to see the washbasin on the right and the toiled on the left.
- and so forth

This is possible by adding a relevant option to the relevant object, where I can specify the parameter. However, one may forget about it, and thus one would rather set them all on a separate page. The collection of all these constraints defines a style model that applies to the design. Style models are not meant to be unique. For example, I would dream to have Feng Shui or other similar "philosophy" reduced to a set of constraints into a style model, and have AC enforce them on the fly.

Now it's time for coffee...

Robert Hunter

[Anonymous]

>hat's "Mr." Graphisoft to you.
>The writer assumes he is reaching Graphisoft. haha.

No Dwight, no! I simply changed subject!

Ahh, I see, you are from Vancouver... (joke)

>"How will this lead to increased sales?" Not an easy answer considering how few Archicad users ever make a rendering.

Is this due to a constraint of some sort, or just an observation without notice of cause? Perhaps the cause is the overall cost of making such pictures, and I include the difficulty in the costs. If "proper" renderings were a standard part of AC10, via Artlantis for example, people would use them more often. It is just a guess, but the poor quality of Lightworks might be the cause of underuse. It goes without saying, that underuse of the rendering engine causes, as a side effect, the poor understanding that AC10 has of object file formats other than its own... Good rendering comes with good objects, and AC10 has rather bad objects too. Or, perhaps, it is the way around; AC10 is so bad with those objects that it uses a castrated version of Lightworks on purpose. In both cases, the delivery is poor, and I am myself looking around for a better rendering engine, which leads me to the problem with 3d object files. However, this all ends in smoke if Graphisoft is not interested in moving forward. Reading your message, I understand that this is indeed the case, which means that Revit+3dsmax+vray is going to take over the market.

Bob

[Anonymous]

>is your vision of the future that limited? jerk.

Let see if Djordje takes this one...

[Dwight]

jdk wrote:
It is just a guess, but the poor quality of Lightworks might be the cause of underuse.
Bob

You are most certainly just guessing. But it is a good guess, laden as it is with sinister suggestion of incompetence on the part of Graphisoft.

I like that.

Notwithstanding the truth of how bad and difficult LightWorks is when used in Archicad, most users never make renderings because they don't understand or use them, no matter how easy they could be made.

My experience from writing two books about this topic and traveling the English-speaking world to give full-day seminars on computer illustration, most user don't make renderings because of the meticulous nature of the work plus the time it takes to assemble a meaningful entourage. The pure mechanics of making a rendering, no matter how simple, eludes the patience of most users.

Then there's aptitude. Our users predominantly come from a technology background, not an artistic background. They might have been sold the bill-of-goods that rendering was automatic. Wrong. Born at night. Last night. The artistic guys are so far ahead of the curve that they have gone on to more sophisticated imaging tools.

This forum has examples made in LightWorks that are hideous and examples that are fabulous, both by beginner lightworks users. Some of those guys even bought my book.

Even the old internal engine makes useful imagery. Most of our users, sadly, remain reliant on elevations and plans to communicate their design work. They may use BIM, but they are still flatland thinkers.

[Anonymous]

>Notwithstanding the truth of how bad and difficult LightWorks is when used in Archicad, most users never make renderings because they don't understand or use them, no matter how easy they could be made.

Dwight, after two books and related teaching experience, I expected you would state something along this line: <<I think I have contributed to people's better understanding of rendering in CAD.>> As this is not the case, then either people have missed your books and lectures, or they have hated them. Teaching is hard. If a method fails, perhaps it was not the right one. Or, parhaps, and this reinforces my former uninformed guess, the rendering engine was less then perfectly integrated with the CAD. Computers have been invented as clerks, to do the grunt work swiftly. If CADs impose grunt work themselves, in terms of "how the hell am I going to get there from here?", then the reason for underuse is plain clear. I guess, again, that we are saying the same thing here. I have about 20 years of experience with computers to know that present difficulties will be smoothed. The integration between CAD and rendering engines is only at the beginning, and its limitations and induced grunts are also due to underuse. Things are changing, and BIM had its positive role in this. Both CAD and rendering have a common interest in the business. Small companies like Graphisoft may not have the resources to make the big step happen. GS had the idea of BIM, but AD is catching up, and seems committed to go the extra mile. True, there are many CAD users who have no clue about rendering. Most of them are AutoCAD users. But the scenario is changing. Let people play with Revit a bit more, and sniff close to a 3dmax+vray engine, and the trend will kick off. At that point, as in all steps of progress, there are people who will adjust, deliver a better service and take away the clients... Architects tend to forget that they are delivering a service, and that the clients are demanding. How quickly the mass of clients will start demanding it, it is anyone's guess, but those who demand it today, are also those who pay best. In our firm we are tired of 2d/fake3d oldies. We have rather wealthy clients, and the step to photorealistic rendering is just about to happen.

[Anonymous]

jdk wrote:
...Graphisoft, I happen to have a PhD in computing, I guess I know what computers can do, and everything I have mentioned is doable. The technology is already there. You just need to integrate it. Do you need money?...

Bob,
Please tell us a little more about your background...Do you have an Architectural education as well? It seems like your desires are more technologically focused than architectural.

I don't think the issues are a matter of being "doable" as much as they are practical. I've been involved in several (architectural) technology integration initiatives...writing sytem specs, creating process flows,etc....In my experience the more rules/intelligence you add to a system, the more constraining and harder to use (as a design tool) it actually becomes.

As far as your LightWorks integration goes...I think there are more important BIM things to address. ArchiCAD's Internal and Skectch Rendering Engines are more than adequate to get a rendering job done....If an architectural firm spends time producing ray-traced photorealistic images of everthing they do, they're losing time and money...In my experience, the majority of clients don't appretiate the subtle differences of a ray-traced image anyway.

Dan K

[Anonymous]

Dan,

>>I happen to have a PhD in computing
>Do you have an Architectural education as well?

I am close to my third degree.

>I don't think the issues are a matter of being "doable" as much as they are practical.

I think they are doable. On whether they are practical, I can see the frustration of people, including my own, when using present CADs, and have a vision about what research on CAD could do. Present difficulties do not need to sit here forever. It is not written in the holy bible that architects must suffer because of poorly integrated technology. Graphisoft does not need to reinvent the wheel with rendering. LightWorks is not a good engine, in my humble opinion. There are better ones (from best to worse: 3dsmax+vray, cinema4d, artlantis, the last one being far better than Lightworks). The idea to integrate the rendering inside ArchiCAD is a winning one, as people can do a little bit of work in a 3d environment. I tried working directly in the rendering window, but the 2d drawings are still the best. To allow people to truly work in a 3d environment, GS should go at length with virtual reality. Given the cost of it, the best GS can do is to maximize the present use of rendering in AC10, namely allow for state of the art presentations. For this to be possible, AC10 should have a better rendering engine in the first place. LW's renderings are not the best one can get. We all know that such applications are expensive, and that 3dstudio, which only runs on windows, will never deliver a mac version built into ArchiCAD. The difficulty with system integration is this one. Even Autodesk, who owns both Revit and 3dstudio, did not merge them into the same application, because 3dstudio has more applications than just architecture. The way to go with ArchiCAD, in my opinion, is to replace the internal rendering engine with a well-written interface to external engines, and let people free to purchase whatever rendering engine they can afford. In this way, the quality of the rendering depends on your budget, rather than Graphisoft's decision. One may object that it is still possible to use an external engine, but to anyone who as tried this approach, it is evident that a proper integration is missing. But there is a good news on this front, at last: Cinema4d has just released a plugin that does just this. I am indeed in the process of trying it.

[Anonymous]

jdk wrote:
The way to go with ArchiCAD, in my opinion, is to replace the internal rendering engine with a well-written interface to external engines ... One may object that it is still possible to use an external engine, but to anyone who as tried this approach, it is evident that a proper integration is missing.

One objects.
I have worked with Art.Lantis for 6 years, and never found any major problem with AC exporter. I get in Art.lantis what I take from AC, instantaneously. No bugs. No loss of data.
I really don't mind paying the extra (how much anyway?) for having LW in AC, which I don't use, providing AC is worth it's price (which it has been so far, most certainly for version 10).
It's like having an panorama glass roof on my Mercedes s600 http://www.mbusa.com/models/main.do?modelCode=S600V
I don´t use it, but it does not bother me, and it is not worth worrying about the extra I payed.

[Djordje]

jdk wrote:
>is your vision of the future that limited? jerk.

Let see if Djordje takes this one...

Having the time and the inclination to write extensive prose does not excuse you from the Etiquette. Read it again.

Next swearword and you are banned for a week.

[Dwight]

Only a week?
How about a year?

[Djordje]

Dwight wrote:
Only a week?
How about a year?

He will have nine degrees (and know how to spell degree) by then ...