About the Photorealism Graal...

Who can explain its dominance and when are we going to evolve beyond it ?


I must say that I don’t care much or at all for it.
I can understand that it is needed in special effects or advertisement but I see little if any inherent artistic value to it, unless there’s a clear intent as in surrealism like in the work of Duane Hanson. Yet again, photorealism remains the main criteria for appreciation of CG it seems. There is hardly any image on the CG Choice Gallery which is not photorealistic, even if the subject is fantastic. Sometimes the result is impressive when the essence of the character finds its full expression but more often than not the art is simply absent.


OTOH, it can be so fruitfull to go beyond reality. For example this image (just for tests) from Hannibar reeks of the smell of CG but I can assure you that many photographers I know would sell their soul to achieve such an effect.


Jean

In order to understand WHY one as to look at IT.

Why is photorealism so dominant you ask,
well here’s my opinion on that:

We live in the real (at least I like to belive so),
what we perceive as “REAL” is everything around
us because we’re in it…and therefor it must be “real”
to us.

Remember the movie “Liar Liar” with Jim Carrey?
He was under a spell from his son so he couldn’t
tell a lie at all after that, and the movie of course…
was hillarious… I particularily rember a scene when
Jim tried to convince himself that the blue pen was red
but he just couldn’t. He knew what he belived was the
truth (the real) that the pen was in fact Blue so he
couldn’t just call it red.

My point with this example if not obvious is
that what seems real to you and I reflects what
we know. We know a wheel to be a wheel, we know
rubber to be rubber and a potatoe to be a potatoe.

If it’s drawn by hand it can still illustrate a potatoe
and we’d know this since we have the abillity to
associate. Even if it just “looked” like a potatoe…
we’d recognize it and say…hey …that’s a potatoe!

If it get’s blurry and not very well drawn at all…
we might start disagreeing about that and while
I call it a potatoe my friend might call it a stone
and to HIM it is in fact a stone…while in my mind
it’s a potatoe.

I guess abstract art would take us there :slight_smile:

Personally I prefer “photo-real”. Why do I prefer
photo-real? I prefer it because if I have something
on my mind AND I want to show it to YOU…I’d
like to convey my message (image) as clearly
as possible.

Example:

I’ve invented a new “super-cute” character.
First on my drawing board…I draw this cute
character…and it looks…well…cute.

But…then I create it in 3d…and it still looks
kinda cute and “cartoony”

Then…I work REAL hard on it and give it materials
and an enviroment (with lights) that resemble
our REAL world.

Now to me…something magical is happening, this
creature that doesn’t really exist…is becoming
PHOTOREALISTIC. What makes my creature
photorealistic? Everything WE KNOW makes it
photoreal. Little details such as dust and dirt in
the nails and corners, wrinkles with sweat and oil
on it, hair that doesn’t go in the same direction,
lights that tells us that there are plants/furniture
and other creatures standing in its way…
Volume lights telling us that the air is filled with
dust and dustmites alike…

…that’s Photoreal…

at least to me

/JoOngle

Photorealism is an imitation of a photograph, and also a realistic rendering of an instant in time when you are looking at a particular object in a scene. The photorealistic effect prevents you from focusing on other objects in the scene (which is unrealistic.) While that’s an understood and accepted limitation of the camera, it’s not a limitation of CG. You naturally try to focus on those background objects. In reality, when you look from one object to another your eyes adjust focus automatically and instantly. While you are looking at that object, however, other objects are out of focus. This unfortunately makes an image without DOF unrealistic. But at least you have the ability to focus on any object in the scene (which is realistic.) You sacrifice realism either way. The most realistic rendering I can think of is a stereoscopic rendering to be viewed with 3D glasses. You can focus on any object, and when you do, other objects become out of focus.
Of course, it may be the intention of the artist to force the viewer to focus on a particular object, and that can be a poweful tool. In animation, the viewer can be led around the scene, the camera becoming almost a narrator.
Having said that, I have to agree with IamInoccent, that there is rather an obsession prevalent with that type of realism, and there are other types.

I agree with JoOngle, there is magic in creating something so well that it seems real. Photorealism helps people suspend their dis-belief of the fantastic things you show them. Its a part story telling, just like what authors do when they give their character depth. The realism of the characters helps distract from the unreality of the situation…

I think maybe the (amature?) CG community has gotten too caught up in PR. Due to Pixars movies, and some early console 3d games Computer graphics have formed an assocciation with cartoons. And while it may not be the case everywhere here in the US cartoons are not taken seriously. They are seen not as art, but as childish and for the simple minded. So it is a natural reaction for artist who want to be taken seriously to want to disassociate themselves from that look/feel.

I don’t really think of NPR as the opposite of realism, just as stylised realism. In essence everything we (CG artists) make has some level of unreality. Whether something is considered ‘Photoreal’ just depends upon what aspects of reality are focused on and which are disregarded…

This is a good topic to think about! Many interesting angles from which to look at it. It is an issue more CG artist should concider; what to make realistic and what to make stylized, and how it relates to the story your trying to tell. Well I could ramble on for a while more but I’m getting excited to try some of the things I’ve thought of while writing this!

photorealism is good but way over rated.

I think that its because a lot of people coming into CG automatically look at photoreal CG and aspire to create such ‘impressive’ work.

Personally, I prefer the sylized looks of basse and andy. They contain a lot of fantasy and rarely photoreal appeal.

I dont think ive ever set oout to make a photoreal peice.

the idea that we want photoreal in CG is utterly ludicris to me. It seems to me that the majority of ‘great art’ (im talking Mona Lisa gauge here) isnt photoreal at all. Much of it is stylized or surreal.

Ok just adding my two cents worth

I have recently posted about this in another thread asking for help on this very issue. I got a very mild response… possibly because my post was Way! to long or maybe my request wasnt made clear enough.

I feel that the importance of photorealistic CGI is far from overated and its dominance is totaly understanble. Obviously it is not universally suitable, if the artist in question is not trying to convince the viewer that what they are seeing is real then they wont use it and they can style there render however they want. But IMO the majority of work being produced is attempting to convince viewers that what they are seeing is real! not for the purpose of replicating real life but in order to “create” somthing new that does not exit and convince viewers that it is real.

Additionaly the creation of photorealistic renders seems to be the primary method of judging the quality and value of 3d packages at present. given this, it surprises me that there is not more prominence given to this topic in the documentation and tutorials for blender. As i have said previously i feel that blenders sucess depends on convincing not only current, rigid, 3d professionals but also, impressionable, novice 3d artists that blender can create whatever visions they have roming their heads.

The only way i can see this happening is through current blender artists creating copious quantitys of quality imagry. dispite this i can find very little documentation, descriptions and tutorials discussing methods for achieving photorealism with blender, while there are significant amounts written for the two other programs i have looked at, cinema4d and lightwave.

Dispite the fact that many of you may think photorealism is overated if any of you have links to or information on the creation of photorealistic renders with blender or yafray i would really apreciate the help.

IMO photorealism is ‘the’ most important feature of 3d. it allows us to take our ideas and make them apprear real!

Dave

I would also agree with JoOngle, I share his views. To me the “magic” in doing CG is bringing something - an idea in my mind - to life. And how else can it be best brought to life, if not by looking real? I also think that by learning to make something real, we also inherently learn more about our world, how we perceive things, which I believe is very useful. I’m sure I am more avare of the environment I’m in since taking up CG.

Another thing that comes to mind - and I do not want to hurt anyone here - is that the lack of realism in most cases is not because the person decides to make something non-real, but simply shows a lack of talent or experience. Not only in CG, but in several other arts as well (eg. poetry) I have seen too many examples of this. Sometimes anyone who wants can become an “artist”, without any talent whatsoever, can do so freely. It reminds me of a few khmm… cr*ppy shows in our national television, where would be stars show their complete lack of talent (singing, doing ultra-low quality talk shows, etc), and then become millionaires and famous and well-known.

Umm, skipping this little sidetrack, I also believe that once someone has mastered realism, he can continue beyond it, overruling it - when needed - in favour of artistic quality, and the message. This is what I see in the works of such people as @ndy, or the picture of Hannibar that you mentioned.

Another angle is that different people are different minded, I happen to be precise, and technical minded. It simply hurts my eyes to make a piece of CG that doesn’t come to life, that’s missing the little details that make it look real. That’s a personal opinion of course.

Zsolt

Hm… Everybody seems to be using the term “photorealism” when they mean “realism.”
In that context I agree with everyone.
I’d like to create an unreal world and make it realistic too.

noel: yes, I’ve noticed that too, that’s why my post never mentions photorealism, but simply realism. For example, I also don’t like the very heavy DOF images, that isn’t realism! Your eye never blurs images that much, and besides, the eye moves around a lot, focusing on different objects at different distances, so the whole scene is basically in focus all the time in your brain.
The area of PHOTOrealism is I think the one that is overdone, where people get too carried away. DOF, caustics, GI, HDRI and all the other acronyms…
For example, one of the most realistic renderings that I’ve seen, and probably my all-time favourite piece of CG is the “Wet bird” of Gilles Tran.
http://www.oyonale.com/ldc/english/wetbird.htm
Now that is quality! And it doesn’t use any HDRI, radiosity, GI, caustics, nothing, and its years old!
That is the kind of thing that can be a true “graal” as IamInnocent put it.

Zsolt

are you sure it doesnt use gi

I think photorealism or realism is prominent for the same reson people do life drawing. Its practise nothing more nothing less.

I am in the prosses of doing 3 “realistic” (boring) pictures for 1 reason to better my artwork. I dont really consider them art because they dont express anything but i will say when I do my alien on theoretical landscapes my abstracts and my characters I want them took look as they do in my head belivable. GI HDRI they are all just tools to help me potray my imagination.

I welcome any new technique to blender.

PHOTO-realism is not realism.
Just have a look at the famous paintings for exmaple in the ‘zwinger’ in Dresden. They have been painted 100s of years before the first Camera was invented.
The Painter (i think it was one of two Canalettos) used a perspective Projection (Exactly the one from the human eye, he didn’t probably know any other) and nothing else. No DOF, just a close look on what he saw.
OK, he ‘beautified’ it a bit. He did the following tricks:

  • He filled 2/3rd to 5/8rd (the golden cut) of the Canvas with Sky :wink:
  • He took the measure of the Architecture ‘on location’, and then filled the People/debris in his Atelier
  • He even painted Buildings/towers that didn’t yet exist (But if ‘Count XY of YZ desired it’)

So, conclusion:
Have a look at the old masters from ages before Photography

  • Take your time to make a
    a) neat image
    b) realistic image

(a is easier, IMHO)
And realize the change you made through while searching the ‘perfect’ Setup.
Did you also recognize how much more detailed and greater Nature
appears the closer you look?

happy blending

I for one don’t like DOF. It is a “camera” lenses effect and not something that you see with your eye. I think that 3d animators and artist need to drop the old ways of trying to duplicate “film” camera effects in a digital medium. I had a discussion over at CGtalk and I stated that such “intense” usage in or digital 3d medium of trying to duplicate a bygone era of analog film media driven imagery is totally unnecessary now. With digital we can make things look as clear as they do to the naked eye. We don’t have to filter anything out with some old fashioned film effect. Can you tell that I not too nostalgic? Hehehe.

I like old film effects like film noir, etc. But we can add a new angle to these effects that create a look that is impossible for a film camera photographer to produce. We can do things like X-ray vision on any scene in 3d that we choose or create lighting arrays that illuminate a scene on Mars based on the data that we get from scientific data on lighting on Mars. We can recreate a lost city and rebuild it detail by archeological record detail. We can make anything that our minds come up with, which is something that no film camera operator can do even with all of their fancy props. It would be way too expensive and waste too much time anyway.

So all we really have to do is just think up the art, research some of the details and get to work.

I think that some of the other 3d apps assume that we all want to make our work look like a photograph. Why can’t we just try out our software’s features and see what we can come up with on or own without clinging on to photographic theory?

What rule of 3d says that we “have” to texture our 3d models with “real” photos"? Do we have to texture with photos or paint complex uv maps in a 2d app for that matter and apply them to our models? We can do whatever we want to do as 3d artist and still get noticed. There is no rule book to 3d art or any art for that matter.

We can make 3d images that are as real as we want them to be.

Blend on!

As an artist, I see the need for all interpretations of what we see; realistic, photorealistic, stylized, and whatever else. I personally prefer a more stylized look. I do, however, agree that in order to move beyond realism, an artist should probably explore it as fully as possible (crawl before you walk). Look at the early work of Picasso or Monet… stark contrasts to their later work.

That said, as a technician and engineer, the question of whether or not DOF is a “camera only” effect intrigues me. I will grant that DOF is attributable to the use of lenses. However, it has to be noted that our own eyes would not work without lenses. Take a moment right now and, while reading this text, try to focus your vision on something in your peripheral vision. You can’t do it. The best you can get is a blurry impression of those things around what you’re focusing on. That sounds pretty similar to depth of field to me, and tends to indicate to me that having an image with some degree of depth of field (maybe not as exaggerated as some pieces I’ve seen, but still there) is a pretty faithful rendition of what we see with our own lens-encumbered eyes.

I’ve seen that before, a true work of art that is indeed, and anyway, I don’t see any reason you have to use HDRI in photorealism and things like caustics and GI can be faked real well using a clever lighting setup. I think he selected to do something like that so he can avoid putting in that fancy stuff like caustics.

To me art is all about style too. This leaves room for a lot of imagination.

photorealism or not i think you have to rate it as another way to create 3d work. the power behind phootorealism is that somebody who can archive this masters the technique very well. everybody can take a pen scratch some lines and say wow thats art.
but to draw a perfect picture with photorealism it needs a lot more than only artistic understanding of the technique but also an thats the point he knows how is enviroment is.

most people think only gifted people can draw and thats wrong its even proofen everybody can draw but not everybody can analyse his enviroment. understanding what you draw is the key to successfull art work in my opinion. only when you know what you do you can play with it and dont tend to use special fx tricks to give the image a content.

in animation like newtron junior it works perfect with the cartoonish style and in final phantasy the hair and skin was amazing because the Cg made the artifishal created movie a bit more real.

so i would not say photorealism is to overrated maybe other styles are to underrated in my opinion!

somebody who say i cannot draw so i only draw abstrakt and thinks this way he is able to make art is wrong because to make abstract art is as difficult to draw a perfect face.

claas

I’m just an artist and a web designer but my main focus on 3d is getting a buzz from an audience. I have learned that photorealism is not enough to get a reaction. You have to enlist style that creates a look that causes people to notice your work. I use photo-realism in place of taking a photograph and compositing that photograph in with my other 3d scene renders. If I have a topic that needs to be supported by a particular image of an object I can model that up and render it out in no time flat. I even can create whatever I see from a database of real world photos of objects. Then I can use 3d to the max of what it is best used for arranging and rendering that object in as many ways as I see fit or that someone ask me to render.

I wrap all of these separate objects into a main image presentation that is more stylized than photo real. Photo-real renders can live right beside stylized work. It works for me. Since I am the one serving up the images for whatever I am promoting I control what people see coming from my end of the 3d renders. I’d say that 3d film producers don’t have as much freedom as to what they offer a client who is already coming from the angle of making extensive use of live footage etc. As an everyday 3d designer I can do whatever I want to do as far is offering up my stylized “look”. I am not totally committed to 3d ultra realism.

I think that photorealism is rather “easy” to come by with todays 3d software. But an understanding of creating a style or a look that sets one apart from the rest of the crowd is not an easy thing to do.

the pleasure of art is to go beyond reality,

to enhance it, to clean it, to “subjectivise” it, to embellish it, to dirty it and tear it to pieces…

But if your reality satisfies you, you’re still free to look at it, admire it, try to copy it or even match it… and then there surely is a problem somewhere :wink:

I shall now sleep

Ciao
Dani

Well you have to get the right lighting, texture and render setup for photorealism to come out, even my most realistic pics don’t turn out photoreal, though it’s realistic.