3D Physical vs. Virtual
18 messages in this thread
To all the 3D builders both physical and virtual,
No doubt that 3D computer rendering and animation has brought many benefits and
has been a great innovation to design visualization and presentations. Photo
realistic renderings are now widely used where 3D physical models made of
paper, wood, and plastic were needed in the past.
The questions I pose for this new thread are:
1. Physical models do provide a type of information to the viewer that a flat
print or screen can not provide. What is this intangible benefit to using a 3d
physical model?
2. Is the 3D computer model really faster to create than its physical counter
part?
3. I would like to hear from those people that still build physical models.
Why do you choose to build physical models? Are they quicker, cheaper, easier,
more effective?
I believe that physical model building is alive and well today. Why…?
Scott Gaul
Scott:
I can't speak for other professions but I can give you a quick take on the
architectural community.
1. Please elaborate on the type of information that a physical model can
provide that a computer model cannot. Most architectural models worth keeping
usually end up under plexiglas to keep from getting damaged and dust covered so
physical texture is not a benefit to physical architectural models. A computer
model is not necessarily a flat print or a flat screen. You couldn't walk the
floor at either SIGGRAPH and AEC Systems without knocking over a virtual
reality system. Depthography in NYC and other companies have captured
stereoscopic views on flat prints which will make your eyeballs pop out. The
intangible benefits for physical models include the client's ability to touch
the finished product (therefore tangible, yes?). I only say this because some
people need to see, touch, have the product they pay for. Lots of clients
don't like buying databases as they can't touch 'em. Intangible benefits to
using physical models include the ability to have any human pitch in to create
the thing. No computer skills are required.
2. This one is tough to answer because it is so dependent on the
complexity/scale of that which you wish to model.
3. I no longer build physical models, I generate computer models exclusively
now. I don't think I've worked on a computer model that wasn't cheaper than a
physical model to create. Quicker, most times. Easier, always. More effective,
mostly but I'd definitely say more valuable to the client.
4. Physical model building is alive for product prototype/product design,
motion picture miniatures, and other applications. Architectural modellers are
struggling to survive and I'm sure the good ones will continue to provide those
eyecatching models which sit in the company lobby for visitors to peruse. How
long before they are displaced by holographic projections or virtual dispays?
Let's just say I'd hate to be an aspiring young model builder who wishes to get
rich off the construction industry.
Bob
Bob,
>>Please elaborate on the type of information that a physical model can provide
that a computer model cannot. Most architectural models worth keeping usually
end up under Plexiglas to keep from getting damaged<<
Try this… look and a still rendering of a virtual model. Then make an
animation of a small camera move of maybe 10 degrees. What you should find is
a much greater feel for the mass and volume of the animated virtual model.
Wile the photo realistic image of a customers project has a lot of initial
impact, I think it also lacks many things. The viewer is forced to fill-in all
the pieces that are off screen, and there is a certain frustration level of not
being able to handle the design.
As for the virtual 3D head gear, I like how Bruce describes them.
>This is something you can't do with digital, barring wearing silly looking
headgear that costs a bomb and makes you seasick to boot <g> <
Scott Gaul
Scott:
/__________From your message___________\\
Wile the photo realistic image of a customers project has a lot of initial
impact, I think it also lacks many things. The viewer is forced to fill-in all
the pieces that are off screen, and there is a certain frustration level of not
being able to handle the design. __________________________________________
I have never seen a physical scale model of architecture that comes close to
having the detail that can be placed in a computer model when you desire to
have both interior and exterior shown in the same model. To be useful you have
to build a full scale mockup! I have seen highly detailed interior models of a
portion of a building, I have seen highly detailed exterior models but scale
keeps one from melding the interior to the exterior in a useful *viewable*
composition.
Even if you are doing residential work when it is conceivable to physically
model both the interior and exterior it is economically inconceivable to put
the effort into the model to give it the detail which would approximate
realism. Besides, without a modelscope you can't get your head into the model
to get a reasonable vantage point unless you chose an incredibly large scale.
Bob
I'd love to see a highly detailed computer-based architectural model turned
into a high-resolution stereolithography model that would fit in your hand…
you should be able to make a plastic model that has windows, doors, desks,
telephones on the desks, etc.
John:
A high-resolution stereolithography model that would fit in your hand would
snatch the cash from your other hand. <g> Today's CADD driven laser cutting
tools do a fantastic job of getting detail into scale models but the time of
finishing and composing the pieces puts a limit on what is practical to build.
I have no idea where stereolithography is now in terms of accuracy. The last I
saw was far too crude to be used for miniatures but o.k. for full size study.
Bob
He, he… I was asked this same question 15 years ago while in school
(architecture). It turned into a very heated debate divided between the "old
school" (every single person in that campus) and the "geek" (moi).
I think of it in two ways. First, though you can synthetically reproduce a lot
of different elements you couldn't using conventional model materials, the
[synthetic] depth of tridimensionality is still far from reality. That can only
be accomplished with a true physical model. We have two eyes and the screen is
a single flat surface. The counter argument I suggest for this is that the
benefits of being able to transform reality using synthetic tools far outweighs
this lack of tridimensionality.
The second thought I have is that you could just as well relate this whole
question to two PBS shows. One is called Wood Right Workshop and the other is
New Yankee Workshop. The end result is the same. It's just the tools. I guess
it's something somewhere between art and old habits. As art, the method used
should not be evaluated (Were Ana Loka here she would jump all over this). As
old habits, well, there still are people looking for Elvis.
Gus,
I see the need for both. The virtual rendering can give a sense for the
lighting, colors, and materials. But the physical model gives a feel of the
whole mass and volume, this is important also. The viewer may be excited to
see the renderings at first, but he may leave feeling he dose not fully
conceive the design.
Scott Gaul
>>To all the 3D builders both physical and virtual,<<
As someone who does both (although it's been heavily weighted on the digital
side for the last year or two), I'll try to give you my take on the subject.
>>1. Physical models do provide a type of information to the viewer that a
flat print or screen can not provide. What is this intangible benefit to using
a 3d physical model?<<
The most obvious advantage to a physical model is that if you want to see the
back of it, you just turn it around. You don't have to wait 3 hrs for a new
rendering. This is something you can't do with digital, barring wearing silly
looking headgear that costs a bomb and makes you seasick to boot <g>. The
intangible factor would have to be the "wow – neat!" element of a physical
model. I have to be careful now, as the following are my observations *not*
neccessarily my opinions. There is still a power in solid objects that speaks
of craftsmanship and skill more directly than computer models. I find a lot of
people *expect* computer models to be perfect ("you just push a button,
right?") whereas these same people are delighted with a good physical model.
I don't think that either method will disappear. I would envision the
techniques overlapping more, however. I could easily see more desktop CAD/CAM
work leading to a marriage of the two arts. After all, we're only talking about
*output* here. One is flat (dye-sub or video screen) that looks 3d but is easy
to mail or carry about. The other *is* 3D.
Bruce,
>>I could easily see more desktop CAD/CAM work leading to a marriage of the two
arts. After all, we're only talking about *output* here. One is flat (dye-sub
or video screen) that looks 3d but is easy to mail or carry about. The other
*is* 3D.<<
Good point. If you had two out-put devices, one out-put a flat photo
realistic image, the other out-put a simple but accurate 3D physical model,
which one would you use? Assume both devices took the same time and resources.
Scott Gaul
Scott:
>>If you had two out-put devices, one out-put a flat photo realistic image, the
other out-put a simple but accurate 3D physical model, which one would you
use?<<
It would depend entirely on what I was trying to do with the model. If I was
trying to make a specific point (eg "what does the building look like when I
drive past it on the highway?"), I'd go for 2D computer model renders. If I
need to give a general description of the project, from any angle the client
wanted to view it , then I'd prob. use a physical model. In reality, we use
both alot.
The proposal itself often dictates what's needed. They often ask specifically
for a physical model. In practical terms, it's easier to ship flat renderings
than $30,000 physical models, especially when you have jobs in Jakarta,
Shanghai, Abu Dhabi. etc. like we do.
Again, they are both different in effects and they are both needed. I wouldn't
want to put myself in the position where I have to choose one or the other.
Bruce
Scott:
The main problem I have with physical models is scale. Some people may be able
to acurately judge a space when they are looking at it through King Kong's
eyes, but most, I believe, get a better understanding of and feeling for a
space when they can see it from the perspective of a normal person.
And don't bring up model scopes, 'cause they suck.
Also, changes in lighting (heck, lighting in general), changes in material, and
physical changes are all generally more easily made to a "virtual" model than a
physical one.
The main advantage to a physical model (and if it isn't obvious by now, I'm
talking from an architectural perspective) is that you can give the client
something physical that they can stick out in a lobby, or take to their board
meeting without having to orchestrate VCR/monitor setup. Also, I think people
have an emotional attachment to scale models — they're neat — just like that
HO train set you used to have as a kid (And heck, they can be fun to make!).
Are 3D (virtual) models FASTER to make than physical? I'd say this depends on
the situation. If you've something that contains a lot of repetition
(fenestration on a tall building, say) then virtual will definitely be faster.
If the item to be modeled contains lots of unique little pieces then maybe
(MAYBE) physical would be faster.
BUT, regardless of which is faster, virtual is more FLEXIBLE.
Dave
David,
We have been talking about comparing finished models to computer renderings.
But what about the use of study models and massing models. Even if these are
not ever shown to the client, arent these used to help prove designs, and help
the architect study his/her design?
Scott Gaul
Physical vs Virtual
As a dedicated computer artist of over 15 years, I typically only look at
problems that are interesting to solve <or at least explore> on the computer.
Even if physical models do currently have an advantage or two to sustain the
practise of building them, I feel that as the SOP shifts from physical to
virtual modelling, physical modelling will be overlooked even when appropriate.
People will forget that physical modelling is an option and it will fade away.
My architectural clients that still make models say that its biggest advantage
is in being able to plop the thing down amoung ten people and let each of them
simultaneously experience it in a fully, independently interactive stereo mode.
They can all talk about the model at the same time while each looking at what
they want! What technology!<G>
-Paul
>>But what about the use of study models and massing models. Even if
>>these are not ever shown to the client, arent these used to help prove
>>designs, and help the architect study his/her design?
In this case, I'd pretty much come down on the side of virtual — especially
with either the new video cards that allow real-time manipulation of a shaded
model, or w/ the new quick preview features of r4. Again, even though
architects, et al are SUPPOSED to be able to visualize 3d from 2d, there are
often times where 2 or more elements are coming together in a way that's not
only hard to visualize, it's hard to draw to convey info to others (ie where
certain items may only be apparent from certain angles). We had a situation
where an exposed steel truss came in to a gypbd wall and pilaster that had a
little ledge for lighting where this exact situation occured. At the time, we
didn't have 3DS — just ACAD and Alias' Upfront (which they have apparently
abandoned, at least for us PC/Windows users — they came out w/ 2.0 for the Mac
recently). It was just almost impossible to see where certain conditions were
occuring in 2d settings. In such a situation, who wants to clear off their
desk (not a possiblity in my case, I'm afraid <g>) and get out some foam core,
balsa and chip-board to spend an hour cutting and gluing little pieces?
Especially with relatively simple (primative) forms, the thing could be put
together (in ACAD or 3DS, etc) in a few minutes, and freely examined. Then
there's shadows. At school, we had a room called the Heliodon, a circular
space with an adjustable table in the center and a domed ceiling set w/ flood
lights at various positions. Using a control panel and the table, you could set
up lighting conditions for different locations around the world for various
seasons and times of year. How many offices have this? (How about those
little sun-angle calculator wheels, eh?) But, with a computer, it's easy! And
you don't have to set there and draw little lines on your model with different
color pens so you can remember where the shadows fell.
Physical models are cool. I don't dispute this, and I agree there's something
humans connect to with a model — but from a standpoint of how versitile is the
result for the time spent (ie greatest bang/buck), I'd say it's virtual all the
way. After all, when the client has the phys model, what do YOU have?
Dave
>>…and Alias' Upfront (which they have apparently abandoned, at least for us
PC/Windows users <<
Upfront has been sold back to it's original developers:
SketchTech Inc.
43 Main St., Suite 410
Minneapolis, MN 55414
Tel: 612-379-1435
Fax: 612-331-4962
I don't know how serious they are, but when I talked to them on the phone, they
said they were working on ver 2.0 for Windows. They found out that more than
half of the registered users of Upfront are on that platform.
We'll see.
Bruce:
Thanks for the info! I think I'll "remind" them that we're out here.
Dave