Importance of photorealism?

[Fran]

Hi Everyone,

 

I would like some thoughts/opinions/rants (maybe) about the importance of photorealism in conveying a design to clients. By photorealism, I mean that the image is truly indistinguishable from a photograph. There are situations where pass/fail depends upon photorealism, such as in the film industry. But I don't think the success of an architectural image (ie its ability to convey the designer's intent and represent it to the client) depends on the render looking like a photograph.

 

Thanks for any thoughts on this topic.

[ingo]

Well i think our work doesn't need photorealism, its sometimes even a problem. Just think of a client looking onto your rendering and can't find his building (this happens, really) because it looked too photorealistic.

 

In my opinion its all about believability, if all fits well and no 2D trees or people distract from a 3D rendering its okay and better than a pure photorealistic rendering.

 

So for my projects i'm working on getting my own style, so that a client wants me, and not one of the other guys who can do photorealistic images too. Just look at Ernest, he has a lot to do because he has his own unique style.

 

Just my two cents

[Ernest Burden III]

I would like some thoughts about the importance of photorealism in conveying a design to clients.

I think we do world a dis-service when we use the word 'photorealism' because really it's two words--photo and realism. They are not the same thing. We have come to believe that a photo is a 100% true and un-biased view of something. It is not. Just in terms of how it interprets the light that the film or sensing device is tuned to, the way the scene is translated to a 2D plane through a lens, and then how we perceive that image vs. how we would have perceived the scene with our own eyes, it's not the same. We do not process visual information as a data field--giving each part equal weight. Our brains are wired to put greater emphasis on some things than others. So assigning 'realism' to a photo misses the mark. Artwork interprets a scene too, maybe more, but neither art nor a photo are an objective record of the visual world. Often more can be said with art than with a photo, so which is more real, if you can 'feel' the subject better through artwork?

 

A photo or still rendering is a poor substitute for experiencing a space for yourself, it is missing many of the cues that help us what we are 'seeing' --moving perspectives, full-dynamic-range light, sound.

 

I do not see photorealism as an absolute goal, nor have I had much success in achieving it. Non-photo-realism suits me.

 

photorealism, such as in the film industry.

The film industry is all about portraying a lie in such a convincing way as to make you believe it over all else. From makeup to lighting to special effects, movies are contrived, contorted and concocted to distort reality to sell the story. I don't mean that as a criticism, but they play lose with the 'rules' of photorealism.

 

I'm done, I have to work all weekend, no time for this thoughtful stuff. Gotta work...

[bigcahunak]

Hi all

Fran, we had a very good discusion about this topic a few months ago. I think it started as a reply to someone's NPR WIP.

We even argued the "realness" of a professional photograph (filters and such).

I'll try to dig that link for you ( I can only remember Ernest and Dibbers were involved.)

[Iain Denby]

I agree totally with Ingo (Blimey, that's a first! ). For our kind of work, making the image beleievable is the key. I've allways hated the term Photo-realism, it's over used, over emphasised and nearly always untrue. There's a web site out their who claims 'Hyper photorealism' for goodness sake! Guess what it's exactly the opposite. It's the Software companies who have pushed this, as up 'till now this ability is what their programs are sold on. And why is everything a measure against Photo realism anyway? ie Photo realistic or Non Photorealistic?

[Fermi Bertran]

Hi Fran!

 

Well, former posts set it.

 

Not the goal.

 

I agree with Ernest that the word itself a poor choice (painitng, with "hiperrealism" made a better choice), so well, can we reduce it to "realistic"?

 

I agree with ingo that fooling people can't be the goal of arch. viz. but instead of that, help them understand what a project is actually proposing.

 

If used correctly (stressing the desired qualities of the project) a certaing degree of abstraction tends to be a plus, how much may depend on the stage of the project or the public the visualization is aimed to.

 

It seems that after some years of Cool Chrome Balls raytraced reflections and Glass refractions, culminated with the burst of G.I. and H.D.R.I. illumination , there is a certain tendency back, towards what is called N.P.R. (Non Photorealistic Rendering)... Well, I am ecleptic here, anything that serves your purpose is perfect

 

cheers

[Fran]

Thanks for the insight everyone. As with most things, architectural visualization is not a static thing. Sometimes a photomatch is called for and in that case I do believe it's important for the cgi portion to be as close a match as possible. In most other situations, there is not a photograph to compare with because the scene does not exist.

 

I know people who cannot understand the concept that an image can be successful (sometimes more so) without it being indistinguishable from a photograph. I know people who cannot stand to look at my work because it isn't "photo real". In film work, the point is that the cgi HAS to match a particular thing that already exists - cheesecloth filters on Liz Taylor close-ups and all. The fact that photography is often manipulated is beside the point.

 

When I see visualization that impresses me, it is not necessarily the level of realism, but how well the viz artist has conveyed the design. I don't judge bad cgi based on lack of realism. I pass that verdict on cgi that has poor texturing and modeling, bad lighting and dumb camera angles.

[Iain Denby]

I pass that verdict on cgi that has poor texturing and modeling, bad lighting and dumb camera angles.

 

I wish more clients would too!

[nisus]

Sometimes a photomatch is called for and in that case I do believe it's important for the cgi portion to be as close a match as possible.

Photorealism... A lot has been said about it already... (To Ernests comments I would like to add that the choice of 'framing' is NOT a realistic but artistic choice...)

 

I'll just stick to my daytime job in commenting about this, with a few anecdotes from clients.

 

Yes, we had several clients that could not find their building in a photomatch (and since we do scanline max3.1 this is highly surprising...) Many times we have to keep the rendering 'clean' instead of simulating photo-grain to make the building 'pop-up'. Awful, since this is not a lack of skills, but a true desire from clients! The worse the photomatch, the better they like it!

 

Anyway... Many times we've made 'photorealistic' images, by which I mean that the building is 'undistinguishable' from it's environment, that clients didn't find 'photorealistic' enough because they want oversaturated colors, happy blue sky,... You know: Thàt's realistic they say... it is NOT! So I would like to opper the question: WHAT PHOTOREALISM? The real or the fake realism? Reality or enhanced reality???

 

Basicly, (in belgium) any image with a blue sky is NON photorealistic... I could even take an photograph of a building, show it to an architect who will find it NOT photorealistic at all...

 

rgds

 

nisus

[Sawyer]

I think that alot of what happens with renderers is that we get really into the technical perfectness of on image but we forget the impact. I used to work for realtors & I would watch them do an open house. They would spend days getting it cleaned and then bring in high end furniture, take out some of the old original furniture and put in new rugs and couches. They would edit the books on the shelves (nothing offensive no Mappelthorpe picture books) put down new mags and even ask the owner to take out the dog for the day of the open house. Not reality but creates an image the image that someone wants to see (I even overhears one realtor telling her client to bake cookies that day for the smell).

 

Most people wont notice 1/100 of the effort we put into our work. They won't say "gee the light source looks a bit off from the way the light is hitting the clouds on the background plate". Sure architects will but most people will notice the image see if it looks pleasant to look at and then judge if it is a good picture. So many times I see interiors that look perfect but sooo boring. Not places that I would want to go to or buy or think about putting down some money to help build. I think more time should be spent on making images into presentations and yes marketable.

[Iain Denby]

I totally agree with Sawyer and Nisus' last comments. It's all about marketing at the end of the day. Thant's what our job is! Just look at all those glossy car advertisments on TV...They're nearly always shiny cars with often black tinted glass(so not to distract from the exterior) and the only one on the road! You are selling a dream and a lifestyle.

[garethace]

The 'buzz' for architects, is to meet people, get involved in things, organise things and 'sell new ideas'.

 

Do you honestly think they worry about visualisation - it is not like they get up tomorrow morning and go, "I want to improve the specular reflection algorithm on my new revised widget rendering".

 

http://www.nikclark.com/

 

That guy seems to be able to render almost anything, and if you study carefully, there ARE some quite strange looking objects on his web page. Done in beautiful rendered image finish.

 

At the moment, I really do find it difficult to 'separate' the progress of computer rendering and modelling sides of software from:

 

1) Manufacturing new consumer products. Virtual prototyping so to speak and patent diagrams, so that noone can copy your great idea. So you need to record it in realistic detail - like the music industry where you sue if someone copies your song.

 

So to avoid, huge legal costs etc - a good visualisation - if the manufacturer comes back with something that isn't your orginal design - you can say, there is my photographic rendering. Also, if Johnny ideas man couldn't find someone to build his widget idea - then if he patents the design, he owns the right to his idea.

 

Hence, it is worth a few quid on a guy with an Intel Xeon box and a copy of 3DS MAX.

 

2) Mechanical Engineering design involving huge numbers of parts, sourced from places all over the world - the last Boeing airplane designed on computer being a typical example - where the 'model' serves as a kind of 3-dimensional database interface front-end for sourcing information and specs on very detailed parts and components.

 

I.e. You click on the part in the model and it connects you automatically the the manufacturers web site.

 

Now, at the moment, Architecture doesn't suffer from people copying your patented design, or needing to design/create/order new components from a 3D database interface.

 

And so falls far short of the requirement necessary to really bring new modelling computer technology into true positive usefulness.

[Chris_A]

Garethace brings up an interesting point. Something to think about...

 

Design engineering models (i.e. Pro/E models) are far more 'complete' that any architectural model I've seen even in the realm of BIM, but that's not who's imaging. Engineers want to prove that it works.

 

For the purposes of patents you actually want to be a little vague. Less specific drawings allow for more liberal intrepetations of what may or may not be infringed upon.

 

It's arguably much easier to control context and lighting for a product rendering than that of a building. It therefore would seem to be a much more straight forward process to create photo-real images. Accurate materials, texture, and color are critical and still more difficult to achieve so it's uncommon that really nice images are created. Designers would probably rather go to a physical model. They're still using sketches and marker rendering (Alias is marketing 'Sketchbook' to mimic natural media digitally) to develop ideas.

 

A good question might be how so many Architects have come to rely so heavily on the 'realisim' thing to communicate and validate their ideas. Has the ability of the average Architect to understand and communicate their work diminished?

[STRAT]

there is not right or wrong or set fast rules. i suppose it depends on where you work and what the clientelle in that particular area is used to.

 

i'm getting more into NPR these days, but my bosses and clients MUCH prefer my imagery to be as realistic as possible.

[Christian Miranda]

Photorealism!?!?

 

Well, as I love this forum because (but not only) It has been created by an artist and is so nicely kept greatly alive by several great artists for artists from around the world... I will mention my opinion:

 

I trully don´t think "that word" or is that a mix of words? is a requirement for a successful architectural illustration...

 

In the very end what really pleases the human eye is an interesting composition, rich colors (I don´t mean changing the proposed colors but just give them life) and contrast (I hate washed out 3d renderings). And then come the detais...

 

1. With today´s modeling tools one can achieve very fine details such as a naturally laid cushions, a nice bedspread, curtains, etc. So many of us can also be called digital sculptors.

2. Natural looking textures. There are many of us who can produce incredibly natural looking or fantasy textures.

3. Lighting. I believe that monte-based GI, mesh-based radiosity, traditional spot/omni lighting are only tools available to the artist. GI is "fashion" but as we know several artists can produce extremely beautiful images using scanline with omnis and spots.

 

Instead of "photorealism" I believe a well balanced combination of the above mentioned points (which also could be called talent) is a key to a successful life as an architectural illustrator.

 

OMG!!! I hope one day we all will get there!

 

[alexg]

I really think that there's a place and time for photoreal and non-photoreal images.

 

For most of us, we are hired to produce a means of communication. One client might want a 'stand-out' building you can easily tell apart from the surrounding, the other might just want to convince council that their new building doesn't affect the neighbourhood grain. Guess we just have to respond to the needs.

 

Once we had a project that includes a heritage building and we have to make it as realistic as possible so the council just can't spot new additions. Then the developer just show them the image and said 'see? nothing funny there, if we didn't pinpoint the additions you won't find it anyway, so why don't u just grant us the project'

 

Other client need image to show future buyers their new project, so it needs to stand out, creating contrast for the 'untrained' eyes.

 

Spotting which one the client needs is an 'art' to master, but well worth 'art' to study.