Camera correction modifier

[Ras]

Hi

 

I am trying to do what the camera correction modifier does in VIZ to my camera in MAX 5... Converting a 3 point perspective view with converging vertical lines to a 2 point perspective with parallel vertical lines. I can´t find any such function in MAX - it is quite easy to find in VIZ so why can´t I find it in MAX?

 

Thanks

[plastic]

because it isn't there. there is no perspective correction in max. if you have both max and viz get the viz copy tool from discreet.com

it copies all viz only plugins and scripts to your max installation.

 

i'm not sure if discreet likes what i say now, but you can even use it if you don't have viz, since the perspective correction modifier comes with a script in the viz copy tool package.

you have to copy the script files to the right folders and add the 'camera correction' menu item to your UI after (re)starting max.

[Christopher Nichols]

First I would like to point out a few flaws in your statement... warning: this is just my rant and you can ignore them.... the answer you are really looking for is at the bottom.

 

The terms "correction" is wrong even if it more commonly known term. Really you are distorting a perspective that is correct so that it looks more like a traditional hand drawn rendering. The moment you look up at a building (assuming that your building is straight out of the ground), the top will "converge." This is only needed on 99.99% of architecture which come out of the ground straight up. You will not need perspective correction on many Gerry buildings. Basically proving my point that it is the architecture that needs correction not the camera. It reminds me of the movie "Deconstructing Harry" where Robin Williams was out of focus and the world had to wear glasses around him not to get sick... forcing the world to see things his way.

 

OK... that said, what is done in VIZ/MAX to compensate for that is a taper modifier on the camera, which distorts the camera cone, so that the viewing plane is perpendicular to the ground. You can experiment with that of you want. This is what a perspective "correction" lens does in a real camera. There is, however, a MAXscript that will do it for you.

 

Got to:

 

http://www.scriptspot.com/start.htm

 

... and look up "technical camera"

 

it works a lot like the one in VIZ...

 

Chris Nichols

 

[ October 14, 2002, 10:25 AM: Message edited by: Christopher Nichols ]

[plastic]

the VIZ perspective correction/distortion modifier is better than the technical camera script, since it does everything automatically.

 

its a very important tool IMHO.

using cameras with perspective correction lenses still is the key to great architectual photographs...the thing is, photos suck. you have to fake/distort view angle and perspective to get a feeling of how it looks when actually you stand inside a room. i think a (well faked) 2 point perspective is far closer to what the eye/brain sees, in real life. projecting 3d to 2d is always problematic, i don't think there is any 'correct' way.

 

[ October 14, 2002, 11:04 AM: Message edited by: Marc Lorenz ]

[Christopher Nichols]

The only thing that the "automatic" setting does is to automatically make the camera plane perpendicular to the ground plane. It doesn't take much work to do this manually.... maybe a few extra seconds.

 

and yeah, yeah... I hear that argument from all the old fart architect about how the eye/brain works... It is not true. They still use the terms 2 point and 3 point perspectives which are terms you use for hand constructing perspectives... my camera (and 3d studio camera) doesn't think in how many point perspective it works in.

 

When your eyes see an image of a building without perspective correction and see the convergence, you don't think that the building is smaller at the top... you think that the building is tall.. which it is. Some architecture uses that knowledge to their advantage. Some Disney architecture makes their buildings smaller at the top, so that a 10 story building looks like it is 30 stories... Basically prooving that people actually see convergence, and accept it as a queue to height.

 

Another way to prove that it doesn't make sense is to move the camera. Animate a camera (the way your head moves in real life) using a perpective corrected lense, and you will see that the world looks very distored... hense wrong...

 

Seeing perspective correction is not the way people are really used to seeing buildings up close... it is the way that architects started to draw perspectives during modernism as buildings got taller. It is easier to draw 2 point than 3 point... that become the desired method of architectural representation.... to the point that photography was forced to alter... perspective correction lenses are only mostly used for architecture.

 

For now, architecture and architects still live in a world where perspective needs to be corrected... this is all based on traditional rendering techniques... as that art will become more rare, so will the idea that perspective correction is needed.

 

It will be come a dying "art" like the T-square.

[nisus]

pfff... Why does one need all these modifiers if the only thing to do is keep the camera and it's target on the same level and crop the image a bit...?

 

rgds

 

nisus

[Ernest Burden III]

Originally posted by Christopher Nichols:

Really you are distorting a perspective that is correct so that it looks more like a traditional hand drawn rendering.

There is so much to aggree with and disagree with in your statements...

 

A perspective is itself a distortion because it is a projection onto a plane rather that a curved surface (your retina) and because it is monocular, whereby most people see biocularly. Furthermore, you cannot increase or decrease your eye's field-of-view, but you can project a perspective to 180 degrees FOV.

 

I do happen to believe that when we look at buildings our brain is doing a 3 dimensional study, and so, as we look up and see convergence we still FEEL the verticality of the surface. When we take a photo or do computer images, we look at a planar image (paper, monitor) and are struck by the convergence--very different experience. So the convention of 2-point perspective serves to keep our images from distracting the viewer with convergence. Besides, they are easier to draw.

 

Another important characteristic of two-point perspective is that it maintains the ratio of distances along the entire height of the building. If a typical floor-to-floor is 12', and that measures .5" at the base of a drawing of a tower, it will still be .5" at the top. Renderings are part diagram, part art. Clients pay a lot to put their skyline-defining crowns on those towers and want to see them. Never mind that from the ground the upper floors are, in fact, farther away from you and in perspective things that are farther away should be smaller...

 

And, yes, it is much easier to construct a 2-point perspective. Drawing a 3-point is a nightmare. But in the computer it is not. Leading many who endevour to produce CG architectural renderings to bypass the historical convention of 2-point without stopping to consider why it was developed in the first place. Also of note, almost ALL renderings done BC (before computers) were done that way, meaning that the older client base (people with enough money to buy renderings) are used to that look.

 

I'm not saying we should all just march in a line just because it USED to be done a certain way, just that there are reasons beyond easier hand construction. Sometimes its in your best interest to listen to the dying.

 

Your point about animating 2-point 'cameras' is well taken. I find it often creates more problems than it solves. But I do have some 2-point sequences in an animation I am doing now, which look great. What I do is modify HOW the camera moves in order to keep the views from getting weird. For example, a straight-on tracking shot or an elevator-type shot.

 

my point that it is the architecture that needs correction not the camera.

No arguement there. I'm disgusted with most of what I am paid to render. But, if I'm accepting their money, I have a responsibility to make it look as good as I can. But lipstick on a pig...

[Christopher Nichols]

I'm going to agree with most points that Ernest is saying, with the exception of one... The statment about the clients being used to 2D perspective... If your clients are old fuddy-duddy architects... yes... but being in LA... you will be surprised as to the age and experience of our clients... they look for cool, they look for a filmic look, not a "beaux-art" or modernist look.

 

On the other hand... nisus is the one that is correct. The key "correct" way to do it is to render an image with the camera being perpendicular to the ground... use some insane wide angled lense, and and crop out the bottom... essentially that is what you are doing when you are drawing a 2-point perspective...you just don't draw the bottom half. In fact you can save render time by only doing a render region.

 

BTW... I have seen some crazy looking animations with PC lenses in it... it was done by Zaha-Hadid... looked cool and warped your world a lot...

 

BTW there were some great papers and class at Siggraph on the failures of liniar perspective... they talk about all your points.

 

Just as you are horrified by the architecture you are forced to render, I am annoyed by "traditional" methods of rendering being imposed on a new medium. Be aware of history, but never be afraid to find a new expression... an move away from it.

 

BTW... I'm forced to use PC lenses all the time... but I do fight it.

[Ras]

Thanks for your help - and the interesting philosofical discussion it led to...

 

I think that openmindednes towards different artistic approaches is important.

I am often surprised to see how appealing other peoples methods are even though they have nothing to do with reality.

 

- besides - I like tools!

 

Thanks again - I´ll get right to it...

[Ernest Burden III]

Originally posted by Christopher Nichols:

If your clients are old fuddy-duddy architects... yes... but being in LA...they look for cool, they look for a filmic look, not a "beaux-art" or modernist look.

Maybe I should come work there, then. Yeah, the east coast is way too conservative. Usually when I suggest an interesting treatment for a project I hear "that would be great...but not on this project, perhaps on another one". They aren't BSing, they just never want to risk the success of their work to a new approach. The best times for that is design competitions, when they take so long to get their design together that you can do the renderings any way you want because if they don't let you they won't get anything.

 

But again, if I take their money I owe them a product they will be happy with. I am free to turn down the assignment, and I do. But I cannot make them think outside their anachonistic, unimaginative boxes simply by telling them they are not cool enough to hire me. Perhaps a good beating might be the trick.

 

I am annoyed by "traditional" methods of rendering being imposed on a new medium.

Don't take this as critical, just discussion:

 

So would you call my hybrid picture annoying or inovative? I am trying to creat a new look by combining aspects of both, but at its root, its nothing more than what you describe.

 

I would ask then why you put camera shake and a focus change in the fish animation? Those reference cameras and in truth have nothing to do with digital imaging. They did, I thought, make the clip more accessable and interesting by bridging the traditional look of film with CG.

 

Without motion blur and depth of field and GI CG tends to reveal itself as an aberation to the expected. Not that that is always bad.

 

I just watched Monsters Inc. with my son and we always look for the concept paintings and 'how we did it' features on DVDs. There are 'tours' of the locations. One is of the factory, and in it a camera passes a mirror and briefly reveals two characters pushing a camera on a dolly. Its funny because in CG you look at a mirror and don't see what you most expect to see: yourself (or a camera).

 

Thanks all for the thoughts!

[Christopher Nichols]

Well Ernest… I have to say your points are well taken. Better yet, they are right on the money and I respect them a great deal...

 

When you quoted "Interesting... But not on this project." I hear you... it is one of the main reasons I moved out of Houston. Rice is an amazing avant-garde school, but has no place in a city where the people with the checkbooks have no imagination. In LA, the industry (entertainment) is about imagination... not to say that there is no bad imagination or no corruption... there is plenty of that.

 

You are right about Fishy... completely right. I did mimic reality; but not the architectural kind. You can probably give me that. Architects are not used to seeing focus, or even a moving camera. So my point was a means of illustrating a space the way I would want to see it... I guess I would call it a little more hyper-real? Maybe not...

 

Anyway, about your hybrid, I don't see how it fits in my annoying category, as you call it... yes you embraced architectural representation, but not architectural illustration.... you accepted model as model, and even used curved lenses (not a typical 2d illustration tool)... everything you did was about traditional 3D design, not 2d illustration. In a sense you did what I did with my fishy... I embraced film, you embraced model making (at least the way I see it).

 

And you called me on that... I am proud of my Fishy piece, I had fun doing it... but it didn't do what I said, and you quoted... and that is to break the traditional from the mold... I reintroduced flow... and I had so sacrifice a lot to do it: one hour a frame!!!

 

If you really want to get into this, I can get into Citizen Kane and how I think of it with respect to architectural representation.

 

Why do I start these things? I guess it is because I'm looking for answers.

 

Great discussion BTW. I'm glad Jeff picked up on it.

 

Respectfully,

 

Christopher Nichols

[seismograph]

eh .. select camera + camera target, apply to both the taper modifier and play around.

It's a very ugly solution but if you only need exact verticaly edges of you building it will work pretty good.

 

another solution would be to render blowup with a horizontal aligned camera... this is a very clean and Lightscape compatible way.

 

If you get a "shifted" photography from a 60/90mm film then don't forget to flame your client for that .. no!.. just a joke, my impression is that it's much harder to compose an CG scene into a shifted image, because you can be sure that you will see every mistake you do in CG and also you can forget the perspective correction in max because it will be not the same!

[Ernest Burden III]

I am proud of my Fishy piece...Christopher Nichols

As you should be. To recall what I posted after first seeing it "I love it"...or words to that effect.

 

I am very interested in the counterpoint to the arguement of maintaining an aesthetic link to traditional architectural rendering. I agree that it is in need of a version upgrade, but since it is largely client driven, its a slow process.

 

The most important pieces of architectural presentation that the public has seen in decades have been in films: Blade Runner, Star Wars Ep. 1 to name just two.

 

I am working on a business to offer visionary architectural artwork--as well as traditional rendering and photography--to the public. I think architectural artists can develope imagery for public consumption without needing 'validation' by an architect. Architects want to control the dialog with the public, but they need us to draw for them. I can draw for myself.

[Christopher Nichols]

Originally posted by Ernest Burden:

I think architectural artists can develope imagery for public consumption without needing 'validation' by an architect. Architects want to control the dialog with the public, but they need us to draw for them. I can draw for myself.

... as did Lebbeus Woods. He managed to escape the role the illustrator and turned into artist/designer.

 

[ October 15, 2002, 09:33 AM: Message edited by: Christopher Nichols ]

[kid]

...Architects want to control the dialog with the public, but they need us to draw for them. I can draw for myself...

what do you mean by that? I think I know what you mean...

 

At architecture school the architect developes a sort of architectural visual language of abstraction, representation, diagram, collage, etc, which within the architectural compound is usually understood with a certain clarity, but when that language is used to speak with the public and client it is usually not understood, therefore it usually fails to move the client, "what the hell am I looking at", they may ask. The public understand traditional (methodology as opposed to medium) artistic repesentation much clearer - a language not commonly adopted by the architectural student during their education. To clarify this babble I would liken this to the lay person viewing the art of the De Stijl, they may say "Mondrian, that's a nice pattern, it would make a great dress print or table cloth", without reading the manifesto a full understanding of the artistic intent is not possible. Likewise, without the architectural education the visual language of the architect is not understood, the lay person cannot read the intent. In contrast to the De Stijl, we would have the Classicist artistic representation seen in Renaissance,Baroque,etc art. Where the De Stijl relied almost solely on their intent to give meaning to their art, the Classicists did not. Even without a full understanding of the artistic intent of the Classicist artist, the apparent realism of their art was a sufficient vehicle for the art to move the lay person - and this is where the 'traditional' artist/illustrator is of great help to architect...

 

babble over.

 

Oh yes, 2-point perspective, um looks good on manual drawings but I think it looks wrong on 3d renders, just my opinion.

[Ernest Burden III]

Originally posted by Christopher Nichols:

as did Lebbeus Woods. He managed to escape the role the illustrator and turned into artist/designer.

I've been following Leb Woods around for my whole career. He started out in NY doing perspective layouts for [one of the best ever] renderer Brian Burr ( www.brianburr.com ), for whom I worked as assistant for three or four years, later occasionally doing perspectives in CADD. Then LW started doing his own renderings, as did I, he worked a lot for KPF, where I met him once when we were splitting a job.

 

He had struck out on his own again to do his own imagery and books. He has established himself as a force in architectural visualization. He was a professor at Cooper Union in NY.

 

I hope to follow him into most of those, we'll see how well it works. Before I imply too much I should say that there is little similarity beyond that.

 

LW has produced a prolific amount of work. Leb is a licenced architect, an FAIA in fact, and has a talent that I find intimidating to say the least. The weird thing is that he has vanished. He no longer lives at his long-time loft in NYC, isn't on the Cooper faculty, has no website or regular publisher, etc. I heard he had married a woman from Sarievo, so maybe that's where he is. If anyone has a contact for Lebbeus Woods please pass it on to me.

 

The only other artist working today that I think as highly of as LW for advancing true architectural creativity is Roger Dean (of YES album-cover fame). I was able to meet him recently at a gallery, I bought one of his sketches which hangs right over there.

 

OK, then there is Syd Mead, whom I hope to meet if I can ever get my sorry self to LA.

 

I think I'm guilty of topic-drift here, sorry.

[CHE]

Is he real? (Brian Burr) I wish I could shadow him and learn more about traditional art.

 

[ October 15, 2002, 01:19 PM: Message edited by: CHE ]

[Ernest Burden III]

Originally posted by kid:

quote:

---

...Architects want to control the dialog with the public, but they need us to draw for them. I can draw for myself...

what do you mean by that?Very articulate 'babble', kid. I enjoyed it.

 

But what I meant by the statement is that architects, at least in the US, have sought to be the one and only voice of architecture in society. Sort of like you need an electricial to change a light bulb. You don't. Architects want to be the voice and spirit of architecture, the only ones qualified to present ideas and plans for our shared environments. Yet they often fail to make their idea accesable to the general public, keeping it in an elitist, acedemic air. Many of them either can't or won't draw or do renderings, leaving that vital means of communication to freelancers like me. But clients rarely want to know what I think about the projects, which is fine. But if I have ideas to express I have the tools to deliver those ideas to the masses, same for most on this board. If we are doig our own art presenting our own ideas, then where does that leave the architects? I mentioned Syd Mead and BladeRunner. Think of what a powerful architectural expression that film was, and that was the work of a renderer/illustrator/designer. And then there is Doug Chiang's work on the new Star Wars films--fabulous. Those ideas were seen by probably a billion people.

 

I think its time for me to shut up.

[Iain Denby]

I dissagree with your last comment Kid.

Converging virticals (especially when it's not neccessary ie the building isn't tall) is the biggest sign of an amatuerish illustration.

 

In my opinion, if you're dealing with tallish buildings, is to set camera and target the same hieght, and, if neccessary move away whilst increasing the lense size and reducing the FOV.

[Ernest Burden III]

Originally posted by CHE:

Is he real? (Brian Burr) I wish I could shadow him and learn more about traditional art.

Quite real, an Aussie. A wonderful man and illustrator from whom I learned so much. He is trained as an architect, and I think licenced in Australia, though he usually lives in New York and usually sticks to rendering. He has an eye for architectural detail in a drawing that I will probably never match. That made him hard to work for--he could see my mistakes easily. Otherwise, a great boss.

 

I am particularly impressed with his treatment of landscaping, something that is frustratingly hard to get THAT good in CG.

 

Look at his work, and other traditional media renderers and figure out why you like aspects of the pictures and find ways to use CG to achieve the same feel.

[Christopher Nichols]

Originally posted by Dibbers:

I dissagree with your last comment Kid.

Converging virticals (especially when it's not neccessary ie the building isn't tall) is the biggest sign of an amatuerish illustration.

Times are changing... there is no argument I can find, even with my really great conversation with Ernest, that makes that statement true. The only reason you would make your verticals straight is because it is easier to draw or because of tradition... the tradition exists because it is easier to draw. Don't give the argument that your eyes see it that way. I've already gone through that argument and find it wrong as well... go back about 15 posts.

 

Don't straighten my world... my world points up with the lines coming together...

 

Quote from Steven Wright:

 

"I had a dream that parallel lines actually met... but they were discreet."

 

[ October 15, 2002, 11:24 PM: Message edited by: Christopher Nichols ]

[Iain Denby]

Times are changing... there is noI argument I can find, even with my really great conversation with Ernest, that makes that statement true

I'm definiteley showing my colours, because I am one of those traditional perspective artists that have been talked about here. 15 years before digital. I agree that virticals are virtical because it's easier to draw (beleive me this is true!)

 

yes, times are changing, but this is because of the many illustrators who use the software as there artist rather that there tool. In general, traditional artists have a clear vision of the finished artwork before they start, and use the software to achieve this vision. The finished result is an illustration. others only see there artwork once it's rendered, and seem to have an obsession with tricking the viewer into thinking they are looking at a photo!

 

Don't get me wrong here, I'm not trying to be snotty, but digital work produced by traditional trained illustrators is obvious.

[Christopher Nichols]

OK... not that I want to really push this some more... but look at these images by Payne Rowlett... Obviously used a PC lense action on them:

 

 

Especially this one:

 

 

They don't look top heavy to you? They look like the top is bigger than the bottom... with the "laws" of linear perspective they are bigger at the top. I know some non-architects that have commented the same.

 

See this one has some depth and movement to it:

 

http://www.visuraimaging.com/images/fifth5_05.jpg

 

My point is, the PC lense is a marrage of elevations and perspectives... it doens't make sense in CG... why?

 

Because you are correct, CG is moving to be more and more photoreal. It is specifically because of that that images like the ones above look wrong to me. I get paid to give people renderings of what their building will look like... I need to be photoreal... it is my job. I also accept 3d as 3d and try not to pretend that I did something on paper. And you are right you can tell images done by tradition illustrators... there is no life or movement in the image. Ideally, I would want all of my images to look like stills from a movie, not images done with a 4x5 on a tripod and a PC lense.