Luminous Landscape Forum
Equipment & Techniques => Medium Format / Film / Digital Backs – and Large Sensor Photography => Topic started by: AndrewDyer on January 12, 2010, 03:48:32 am
-
Well... it does have some large format cameras in it.
Here is the description that is on the page.
"A FULL-CG animated piece that tries to illustrate architecture art across a photographic point of view where main subjects
are already-built spaces. Sometimes in an abstract way. Sometimes surreal."
http://www.vimeo.com/7809605 (http://www.vimeo.com/7809605)
It is hard to believe that one guy did all this.
I hope you peoples like it... it blew my mind.
enjoy.
Andrew
-
I agree it's a totally awesome piece of work. Especially when you consider it's all done by one guy. There are also a couple of videos showing how he did it.
It isnt totally CGI though - he has composited some live footage in as well I believe - but it is still an outstanding piece of work that is also quite beautiful to look at. Creating quite a stir on a lot of CGI sites as well
-
Thank you Andrew for posting this.
Stunning
Cheers,
Frank
-
thanks for posting...
stunning work makes one think..
-
Makes me hope my clients don't start to use renderings instead of photographs.
Stunning piece. I linked it on my blog too. Thanks for sharing.
-
That was mesmerizing! Thanks for the link.
-
I liked it, but I liked his piece on Kahn's Exeter Library more. It's on the same website. What a beautiful building, beautifully captured!
Best,
Mitchell
-
Beautiful
-
reminds me somewhat of Gattaca.
music and lighting
so how many of you think that this is the possible future of photography/illustration/advertising?
-Dan
-
OMG! what is this? Can somebody tell me what is this CGI stuff?
Is it a threat to photography?!!!! lol
The lighting in this video sure makes my photography look like a monkey
-
I work in the CG industry, coding a renderer, and this kind of ability has been around for years. Mostly, there is no threat to photography from this kind of production, since it takes months to put something like this together, much more time than simply lighting and shooting a situation.
The only area where CG is making real inroads, to my knowledge, is, as mentioned above, when the CAD data are ready, but no device/car yet exists. A complete advertising campaign can be underway before the first car is built.
-
I work in the CG industry, coding a renderer, and this kind of ability has been around for years. Mostly, there is no threat to photography from this kind of production, since it takes months to put something like this together, much more time than simply lighting and shooting a situation.
The only area where CG is making real inroads, to my knowledge, is, as mentioned above, when the CAD data are ready, but no device/car yet exists. A complete advertising campaign can be underway before the first car is built.
briefly, how is it done?
-
briefly, how is it done?
How is what done? The whole thing?
1) Modeling of all the parts in the scene. Some parts can be done with particle systems, some can be generated on the fly, like leaves on trees, but well over 90% of what was in the movie had to be modeled per hand. Some of the more detailed things would have taken significant time and research, including measuring the (mostly, if not all, real) buildings, and so on.
2) Texturing and shading. Capturing the right textures by going out with a camera. Some can be generated, or re-used, like the concrete textures. For realism, most textures will need to be real captures, or perhaps painted by hand, such as the brassing on the cameras in the movie. This also includes capturing the movies of the sky, for later comping in. Given the visual richness in this short, I doubt that many off-the-shelf shaders were used. Probably almost everything was textured.
3) Lighting. We are not only talking realistic light, but also creative light. This could also have taken real time, although some of the outdoor lighting could be done with a "sun and sky" shader, which takes the time of day, location on earth, and given weather, into account, but to enhance or draw attention to certain items, much will need to be done manually, and experimentally, with many revisions until the result is perfect. Since test renders are necessary, this is also a place where the equipment can slow you down.
4) Cinematic planning. Camera paths, angles, lenses, effects, and so on. Flare was present in many shots, and this has to be planned in combination with this stage. Some can be automatic, but since it is mostly present for creative effect, much has to be done manually.
5) Rendering. A computer sits and generates the images from the data on hand. This can take months by itself. I believe that V-Ray was used, which does help with some of the tasks listed, as it is somewhat automatic compared to other systems, and efficient with rendering.
Of course, a real production is much more chaotic than this, and there will be more than a trivial amount of system administration, futzing with licenses, getting networks tuned for speed, setting up render queues, and on and on. Normally a sequence like this would be the result of tens of people. That one person did almost all of it is pretty incredible, and he must have been working on this for many months, maybe longer if it was not full-time.
Note that I did not schedule the creative time, just the technical time. Someone has to have thought through all of this... FWIW I thought that this was one of the weakest parts of the production. While the richness and stunning vistas were there, there was a much weaker sense of "why?".
If someone wants to know more about all of this, the creator of the movie probably hangs out on some forum, and is answering questions in this very moment, perhaps cgsociety.
-
The worst part, I see wedding photographers possibly being part of the immune crowd with where technology is going.
Interesting way to insult about half the members on this board
Care to elaborate on why that's "the worst part"?
-
Technically impressive, but am I the only one who finds repeated and totally gratuitous focus-pulls and shallow DOW to be gimmicky and tiresome after the first 10 times?
-
Nothing about it was captured. All done in the virtual sense. (Except the skies I think were from video and used as a background plate)
This video shows more about what was and wasn't CGI - looks like the photographer is captured - as well as certain items like the flying white pigeons.
Compositing Breakdown (T&S) (http://vimeo.com/8200251)
Like I said earlier it is a brilliant piece but it looks like the guy is excellent in many fields - compositing being one of them
-
Hi Carstenw.
Thanks for explaining to the forum some of the processes in creating this kind of video.
I also work with CG in 3D and 2D, so I can agree with what you have said.
People need not fear that this kind of work will take away photographers jobs... it would cost a ridiculous amount of money
in man-hours - I am sure it would be much cheaper to get a building photographed than modelled, textured and rendered like this.
The talented person who made this video I believe took a year off to make!... which I can truly believe possible.
Where CG comes in handy and would be preferable to photography is in animation and impossible camera moves/positions.
So keep your cameras guys!
-
In 10-20 years though the story might be quite different. Not so sure if it means much to many of the grizzled and seasoned shooters here but for the up and coming shooters or even 30 year olds like myself, we'll certainly see it become mainstream, quick and easy within our working lifetimes. Something to consider...
-
So watch this: http://www.vimeo.com/8217700 (http://www.vimeo.com/8217700)
15 minutes of video from sped up motion screen grab to show how he created the library in another piece. Didn't seem to take a year or a month, maybe a day.
What some people don't consider is that he can create views and get angles and do this in motion with pulled focus and a whole host of other real world techniques that require huge ass motion controlled cranes, large crews and big budgets. Some scenes I would say actually are not possible or feasible in real life.
Right, there are techniques which can help to quickly model existing buildings. For this particular shot, he went to Google Maps, grabbed a SketchUp model, imported it into 3D Studio MAX, refined some details, copied and pasted a bit to get the empty bookshelves, grabbed some existing textures, layered them in Photoshop, added some ambient occlusion, and refined the look of the whole thing. The movie is sped up, however, and keep in mind that he is only showing the successful part of the process, not all the planning, thinking, failed searches, re-thinking, and so on. The process which he shows in 15 minutes could easily have taken a week, depending on what level of fidelity he eventually wanted, and how many failed ideas he had on the way. It is like showing the 5 minutes of a day-long photo shoot which netted the best shot, without showing everything else that built up to that point. Also keep in mind that if he had been able to do the whole movie at the pace shown in the video, it would have taken him less than a week, but in fact it took a year... Finally, the movie shows only the setting up of the building. There is no knowing how long he took to do the camera motion, falling paper, and so on, which were needed to complete the shot.
And that is only for a building which happens to already exist as a relatively detailed model. That is going to be the case for more famous buildings in certain countries, but not for the new subdivision someone might have to photograph. In other words, this relatively easy path is only available for a few select people. The rest of us have to do much more work to get the same result in a different place.
Another aspect to keep in mind is that he is not going for absolute realism, but rather for a convincing alternate reality with some neat tweaks to make it pop more. Some photo jobs would call for absolute realism, and he would have his work cut out to match that.
By the way, the second half of the movie or so show him setting up a bunch of lighting tweaks for a single shot! He has to repeat much of this work for every shot, so although the modeling work is amortized over big chunks of the movie, a lot of the work is not.
The other thing is cost. To do this he had to take a year off and work on this full-time, probably the kinds of hours that studio-based pros in big cities are used to, i.e. 60-80 hours weeks, if not more (and he sat on his ass in front of a computer the whole time!). The machine(s) necessary and the software involved are not cheap. For a full setup (and I doubt he had everything registered he is using...), this would not be much cheaper than a full MF digital setup including studio lights and so on. This means that going the CG route isn't necessarily going to cost the customer any less...
His shots also show only static objects, and leaves, as well as a few falling papers. Making a realistic CG shot of a human can take weeks or months, where with a camera, you could do it in 1 hour.
On the other hand, he can do things which cannot be done with physical equipment in the real world. He can move the camera places which would cost thousands to do with a real camera, at the click of a mouse. He can change the light however he wants without regards to realism. He can speed up and slow down shots without needing special cameras. And so on. So there are benefits for CG too.
---
In the end, some of the things he had to do could be sped up with better software and faster machines, as well as access to large texture and model libraries, but a lot of what he had to do cannot be sped up, and just like in the photography segment, a good director and cinematographer cannot be replaced.
There will always be a need for both. The balance might shift slightly, and companies like RED will have some effect here, but there are many things which are just done better/cheaper/faster/more easily via photography than with CG, just as the opposite is also the case. Take care not to be a specialist in a segment which can easily be replaced with CG though
As for this particular guy, Alex Roman, nothing I have written should denigrate his work in any way: he is obviously extremely talented, hard-working and his vision is top-notch. I am guessing that this movie has guaranteed him a career at the top of his field for the rest of his life. This kind of talent is as hard to find in CG as it is in photography.
-
Thanks again Carstenw...
I was about to to try and reply to John-S a bit, but I think you have done it much better than I could.
Yes the 'making of' video only shows a fraction of the work involved... and the talent needed that very few have.
-
One thing that I saw with a potential furniture client, is they shot a room (photography) then rendered new furniture into it (saving the cost of building multiple configurations and materials options). I know they have been doing this in Automotive for a while, but this was the first time I'd seen it in an architectural setting.
I think one thing we'll start to see, is more of this kind of combination of technologies. Rather than the death of photography proper, I think it will instead morph in an integral part of the process of combined image making techniques.
-
One thing that I saw with a potential furniture client, is they shot a room (photography) then rendered new furniture into it (saving the cost of building multiple configurations and materials options). I know they have been doing this in Automotive for a while, but this was the first time I'd seen it in an architectural setting.
I think one thing we'll start to see, is more of this kind of combination of technologies. Rather than the death of photography proper, I think it will instead morph in an integral part of the process of combined image making techniques.
Yes, that matches my thoughts. Product photography, especially for new and possibly not-yet-built products, will go to CG. Not food photography though. There will be fields where CG works, and fields where it doesn't.
[!--quoteo(post=0:date=:name=John-S)--][div class=\'quotetop\']QUOTE (John-S)[div class=\'quotemain\'][!--quotec--]Every time I turn around, I see the same old stuff, boring landscapes hashed out a million fold before, people thinking that emulating a few styles is their path in photography, Dave Hill look, HDR, whatever, same old ideas of what the fine art market is.[/quote]
John, if you think that bad photography is going away, I am sad to disappoint you, but I am certain it won't
-
Well, that movie was impressive!! Having consulted on CG projects as a lighting director (car rendering) I know that the time involved in producing that piece was considerable. But who cares, it's the end result that matters and Alex has proven how great his talent and capabilities are. Now back to the conversation of it's application in advertising (commercial photography) and I know that it's forever changed the business of product photography. Ten years ago there were automotive photographers everywhere, some alot more talented than others but there was so much business that the even the less capable shooters had enough business to make a living. Five years ago the business had already changed, the basic images that every manufacturer needed now could be accomplished with the same data that the engineers created for their CAD data. No more vehicle transportation, vehicle prep, preproduction vehicles for marketing (at least, alot less of them) and no more photographers invoices. This knocked a few guys out of the business. Now as it turns out the cost savings were minimal because the cost of creating the CG and the additional retouching the images required (due to usually either math issues or operator issues) was considerable. But if the vehicle only existed as a clay model (at best) it worked very well and it still does. In fact, the quality of CG is equal to what I can do. Did I really write that? Well, maybe it's almost as good. Anyways, it still takes alot of time and the day in and day out renderings don't look that good althought the consumer would probably not know the difference. Luckily, my clients do and in addition to quality, a big advantage automotive photographers have now is speed. If I have a retoucher working with the client (and myself) as we shoot, I can produce up to 3 vehicle majors and a few details in a long day in the studio. The next day the retouching is wrapped up and the project goes to the printer or online. This scenario isn't ideal but happens more often than I'd like, it also doesn't apply to car launches where everyone from the vehicle designer to advertising manager to creative directors need to be involved (that slows things down a bit). But still most of the day in/day out images of vehicles are now rendered and that's taken a number of the remaining photographers out of cars and into another area of expertise. Anyways, there's no denying the impact that CGI has had on my business but I've had worse years before it even existed (2008 was one of my business's best, out of 29 years). So I just work harder, try to make my client's life a little easier and diversify at the same time, this, hopefully, will allow me to work the rest of my life!! Jim
-
adobe knows why they are putting the biggest advances in photoshop in 3D tools.....
people and animals...hair in general is the hardest to do...almost unlimited surfaces to render....so this will take a while...
it makes so much sense for cars....shiny surfaces that see everything...i remember when i got an email for a location! service...specialising in CG generated locations....
anyway, i think the movie is great, yes a little heavy in the techyness of it all, but that is what it is supposed to be.... a showreel, nobody will hire him to direct a feature...
i noticed one shot: the camera very slowly comes up the stairs and the shot is a time lapse shot....a very nice, subtle effect...i don't think this is doable with a camera? you would have to move the camera up the stairs a single foot and take several hours to go that distance....
for impossible shots done without CG...."soi cuba" opening shot...one shot....also: "andre rublov" (tarkovsky) i think it is the 3rd part, opening scene, the bell maker.....one shot....i don't know how these guys did that....breathtaking....both great movies btw.....
-
Sweet movie, very visually stunning.
I did notice that the windmills are spinning backwards at 7:45 though... a little bit of an "oops" when you've put a full year into a project.
-
It is certainly a discussion worth having. I come across CGI more and more on my visits.
To give a you a few examples (where photography is still important) ;-)...
A 'Survey' company here in the UK does a lot of laser scanning of buildings. To give you one example of many...
They had to survey a church in disrepair in Europe. So they first laser scan the whole building, then photograph sections of it with an H3D39. The laser scan and images are combined to provide a environment project workers can visit.. ie a Virtual Church with massive resolution.
So if you are the stone mason, you can zoom into any part of the structure to asses it. Bearing in mind the stone walls are made up of many 39MP captures.
So many people can calaborate on a project like this from all sides of the world without visiting it. Sometimes they use 35mm but on this kind of work where they need the upmost detail they use MF.
The same company also visits movie sets to capture textures. I didn't know (Carsten most likely does) but a lot of movie sets are now only one storey high - if they need a taller building, then the texture of Storey 1 can be copied up and up and up! So I saw a whole movie set at one level, and then the final rendering of the scene. Again this was captured by an H3D39.
As a final example some Ikea 'photographs' (One of our largest clients) are part CGI, part photography - sometimes using textures captured by camera.
I think photographers should have a serious look at...
http://www.bunkspeed.com/index.html (http://www.bunkspeed.com/index.html)
I downloaded a trial version of Hypershot and rendered a pretty impressive looking diamond in about 30 minutes. This was an older version of the software with a truly appalling user interface so it took me half a day to figure out how to use it! I think it has come a long way since then.
I don't think it will replace photography but it is a way to combine your skill of capturing high resolution images but doing more with them in applications like HyperShot.
Going back to the Diamond I would not have been skilled enough to shoot it as a photographer. On the bunkspeed forum there are some quite impressive jewellery examples which people have posted. There are some duff ones too, but the good ones are interesting to look at.
Here are some I was looking at a while back...
http://hypershot.hyperboards.com/index.php...mp;topic_id=763 (http://hypershot.hyperboards.com/index.php?action=view_topic&topic_id=763)
David
EDIT - I see there are some issues with Bunkspeed and the licensing of a core technology from the developer who designed the ray tracing algorithms, so it looks like you can't download anything right now. Check their blog for updates.
-
One thing we certainly agreed upon that many people just don't grasp that PHOTOGRAPHY is being redefined at the moment and honestly nobody knows exactly what it is becoming, how different it will be used, who will use it and pay for it differently from before, but the fact remains WE do not have control of where photography is going. Consumers, advertisers, content users, businesses all will play the role of morphing photography as we know it. We are along for the ride to give them what they want.
It is changing, it will change from what most are doing right now w.r.t. to selling imagery. Alex's video represents one avenue of real potential change. Instead of blinders, might as well put on some 3D glasses to help enlighten our perception of the future.
John, I don't just believe what you are saying from your first posts. I am experiencing it.
There are a number of industries that have dropped 70 to90% now done in CGI. the model is already made, and can be tweaked. No need to remake it. Shoot a house....different house, just add a room, change the roof..Done. Change color...zap (with reflections of color).
Its a good thing having something to fall back on to fill some gaps. 3, 7 maybe 10 years, and you too will feel it creeping in. Sure there are always weddings and highschools one can turn to (no jabs at you)....then again, as John said....
Very nice video. I guess I have seen this and know the abilities for a few years now. .. not so much on a tech achievement level. (The Movie Godzilla years back was a lot of 3D, using SoftImage, a MS software) Though impressive enough to read through and appreciat the work. Beautifully done.
-
Carsten,
Thank you very much for explaining the process. Basically this guy built the world he is using for each scene in 3d studio for example and then created textures for everything and finally added movement. Much like video games are made, except this has realistic textures, correct?
Thanks
-
Thank you very much for explaining the process. Basically this guy built the world he is using for each scene in 3d studio for example and then created textures for everything and finally added movement. Much like video games are made, except this has realistic textures, correct?
You are welcome, it was a nice change of pace to write about something I know about instead of asking questions about MFDBs
I see several companies mentioned in this thread which are customers of the company I work for. Interesting. For example, David, IKEA is also one of our customers, but for the CG part
Video games are getting better and better, but generally to texture whole levels or worlds you cannot use the same high resolution that you would for offline rendering, like the video of this thread. The same goes for model polygon counts, and so on. If you need to render 30-60 frames per second, your approach is very different than for realism. There are also areas which exist in between, with a few frames per second but higher fidelity, like the SketchUp models used in Google Earth for example. Bunkspeed is neat software, but aimed very narrowly. Other software is not so fast, but handles more general work.
Anyway, yes, CG will certainly take over some areas traditionally done with photography, at least partly, mostly focused around product shots and perhaps architecture. As David mentioned, this will also sometimes open up new areas of photography, like capturing the textures for the church, and so on.
-
Amazing video!
-
One thing we certainly agreed upon that many people just don't grasp that PHOTOGRAPHY is being redefined at the moment and honestly nobody knows exactly what it is becoming, how different it will be used, who will use it and pay for it differently from before, but the fact remains WE do not have control of where photography is going. Consumers, advertisers, content users, businesses all will play the role of morphing photography as
Once I became fully immersed in shooting digital I noticed a change in the way we worked.
About 7 years ago I was shooting an ad campaign on a movie lot in LA and I walked past the stage next door where they were shooting a Steve Martin scene and noticed how similar our still set was to the movie set.
Obviously they were working in a larger scope, with more crew, larger sets and the ability to move the camera in fluid motion, but the differences were small. We had gaffers, many multiple lights, tech stations for viewing and I knew then and there the still world had changed and it wouldn't be too long until the mediums finally crossed paths, or started bumping into each other, at least on some projects.
Sure, there are still shoots where we go out with a small crew, though small has gone from photographer, stylist and assistant and moved to photographer, producer, 3 assistants, gaffer, two techs, one video operator, 3 wardrobe and prop stylists, hair, makeup (usually multiple teams), with multiple cameras and images are shot knowing that we will change it, manipulate it, add to it, in post.
In fact it's not unusual for the post production budget to get close to the shoot budget.
What that means, I'm not sure, other than the cost of production has gone up dramatically and everything is exponentially larger.
I do think the role of photographer has changed to that of part director, part special effects supervisor, part content provider and with the advent of the 5d2, motion image director and/or camera operator.
In fact our complete project briefs have changed. Years ago (maybe 3) when you were hired you knew exactly what the image was for. A national print run, and/or outdoor, but you knew where everything was going, what size, format, usually a detailed layout.
Now, we shoot, usually the project brief resembles a storyboard more than a print layout, the use is somewhat non pre determined. It might be a package or in-store shoot, it may go to print, it may be interactive, it definitely will have some form of web play.
And unless your a complete hands off type of guy, your going to probably have 3 days of computer direction for every shoot day.
It is just not unusual to have a complete photograph that is a composite of 5 or 6 images, so even though they may be shot on a 24x36 camera the real format is what . . . let me add this up . . . 96mm x 144mm give or take a few mm.
Maybe that's why I don't understand the talk of formats. I do understand that some cameras are better suited for some tasks, but in reality the format is one of the last things I think about.
The only thing that limits us now is fast universal bandwidth. In one of our studios we have fiber optics and I can move 400mb through the web almost as fast as I can transfer it to a hard drive.
In another we have two standard dsl lines and going back to them is like going back to phone line bandwidth.
The last large project we just completed we moved over 2 1/2 terabytes of data between our team and the client. All deliveries of 125 composited, layered, finished out images were delivered through the web.
So once we have some form of universal wireless fast bandwidth that will play from handset to desktop computer in almost real time, the expectations and our roles will change even further.
I do know that the term cutting edge just doesn't exist. There is no edge anymore. Today you just keep up so you don't get passed by and I'm not referring to camera formats, megapixels or limited software solutions.
So to segway into another topic of the Hasselblad clock, it better be ticking on something spectacular, not just a 60mpx still camera.
IMO
BC
-
The last large project we just completed we moved over 2 1/2 terabytes of data between our team and the client. All deliveries of 125 composited, layered, finished out images were delivered through the web.
Client saved a lot on tax, according the laws in the states
-
Well... it does have some large format cameras in it.
Very nice indeed.
This might not be a thread to all areas of photography, for the very reason that the link between photography and reality is important for many of these, but there are clearly some domains where there will be major impacts.
I hate to say it, but fine art images are likely to be among the first to be hit. I might invest in a 3dsmax license instead of a new tripod...
Cheers,
Bernard
-
I think all that said, after it saturates, consumers will understand, appreciate a photo....and look for true photo images to depict what a brand is and what they are selling.
...and perhaps ads will disclose real and fantasy.. Something below the ad like.... (Image of product actual photograph, not a CGI rendering).
OK ...so Art Directors, Agencies!....start recommending that to clients :-)
I too will be considering a 3D software very soon. ...seriously.
I started 3D Max years back...even ATT had a 3D software I played around with. They maya came out. I have made some models and such looong time ago, also using character studio, but this is ages ago, and would love to get reaquanted.
I had avoided it, but it is not something you really can...and it is more apparant to me now.
-
alex roman (http://motionographer.com/2009/08/16/alex-roman-thethirdtheseventh/)
-
Unfortunately consumers are very poor judges to appreciate photos. Endless crap photos splashed across countless tabloid style mags that consumers just eat up, can't get enough of. And when it comes to buying photography, consumers play it safe, quite evident in a lot of formulaic work in the fields of portrait and wedding work.
I strongly agree with Bernard, I have seen a lot of very nice looking fine art images, which because of the surreal nature of the content (nature, landscapes, abstracts) someone could just as easily conceive and produce in CG. As an example, the Antarctica trip many took a few years back with Michael, created a plethora of similar imagery from its participants. many of those shots could be created in CG. Would consumers care, can't predict that entirely. But if it's being used to sell a product, they won't care. If it's presented as fine art with no further information, who is to know. Is the buyer paying for a large print of a surreal locale that a person actually trekked to and froze to bring back such images or is the image itself strong enough in quality, look, moment that the buyer just wants the piece for the feeling the artwork invokes in him/herself, that can be enjoyed on a wall for years. There was a time when photography didn't even exist. There was a time when CG didn't even exist. Photography is accepted to great extents. CG is being accepted to great extents as well, in motion pictures, video games. Who is to say that acceptance can't bridge to other uses.
I have witnessed comments coming from painters who feel strongly about photography in general not being a fine art at all. Their reasoning was complete B.S. but they convinced themselves of such and nothing was moving them. So it all boils down to what the mob feels, that's what moves most things.
I don't see why there would need to be any disclaimers as to whether it's a "real" photo or CG. Beauty photography couldn't be more fake, with retouching nowadays people only seemed shocked for a nanosecond that something was "shopped" and then they continue to buy the same products and magazines that depict the false reality. And David's comments about Ikea and their use of photography and CG says plenty. I hear no one complaining that Ikea uses nearly perfect but not real images (at least not shot with a camera) of their products.
Look at the last Van Cleef and Arpel ad....I don't know who or what agency did that ad....but it is so lacking of anything they made in the past of quality. That ad (Birdcage with grass, is such a DOWNER...I used to consider VCArpel as a high end product....They have not only cut corner or budget, they Chopped ....their credibility as a luxury brand at stake?
I would think at some point the render vs the quality of production could be in question?
-
Something I have wanted to know how to do and perhaps Carsten you may know. I want to add people (not objects) to a scene and control the lighting completely in software. The models will be lit completely shadowless as much as possible. So we have:
1. Shoot the background, room, client set, whatever.
2. Shoot model in the studio in a completely shadowless way.
3. Add model to background image.
4. Using a software program add light & shadow to the model.
Is this possible? I know how to do everything but step 4.
-
Something I have wanted to know how to do and perhaps Carsten you may know. I want to add people (not objects) to a scene and control the lighting completely in software. The models will be lit completely shadowless as much as possible. So we have:
1. Shoot the background, room, client set, whatever.
2. Shoot model in the studio in a completely shadowless way.
3. Add model to background image.
4. Using a software program add light & shadow to the model.
Is this possible? I know how to do everything but step 4.
....make the floor a bit trans, and add a light. Think of whatever you build in 3d as you would actually create in real life.
-
Is this possible?
unless I did get you wrong...: no. If you shoot everything you are working with 2D objects. So you can not add light as on a 3D object and you can not change the perspective.
You need the elements as 3D data and then add light, texture (texture can be and often is 2D) and whatever...
-
Tho_mas is echoing what I was thinking. Until you created a 3D model, you cannot accurately map light and shadow. You can use your image as the texture map on the 3D model, but without the 3D data, you have nothing to work with.
So you need to created 3D objects of your people, even if using shot information, in order to add them into the scene. Depending on the scale though, with a good texture map, you could probably get away with a relatively simple object.
-
.
-
I just realized all he listede was 2D....
Its been a few years, but if your model is free of shadow and you place in a 3D made room..It has to be a 3D environment.
with a 2d background wall....I think it is do-able.
All images are treated as a flat plane, so anything with shape to it, consider the image will wrap around it. If its a wall..it will look flat(besides perspective). Think of isometric.
-
Argh, I am with Phil, I also thought you meant 2D people in a 3D environment, but on re-reading it, realized that you wanted 2D people on a 2D background. Photoshop. There is not enough information for automated work here.
========
Here is my previous post:
Well... depending on what you are hoping for, you could use a trick to get in the right direction.
Fundamentally, the others are right: if you have just 2D data, there is no easy way to add 3D-dependent information to it or the scene.
However, where there is a will, there is a way. First of all, the people need to be carrying their own shadows (on themselves), or you will need add it in Photoshop.
The shadows of the people on the environment, however, can be faked. It would be possible to model a really crude person, make that person invisible in the scene, but able to cast shadows. That way when it renders, the shadow will come from the proxy, and the look of the person from the photo/cutout. The feet would need to match closely, but we are surprisingly bad at telling exactly what shadows (or reflecions, or ...) should look like. Basically anything beyond directly visible objects can be fudged quite a lot without us being able to tell.
As far as shadows from the environment onto the person goes, that could be a little trickier. Since the visual appearance is given by a 2D cutout, any shadows falling on it will be flat. You might have to resort to Photoshop here again too, to fix up the details.
Architectural visualisation uses this type of workflow a fair amount, to my knowledge. Aren't there helpful CG sites for this audience? Failing that, more information could be found in the forums on sites like cgsociety.
-
Something I have wanted to know how to do and perhaps Carsten you may know. I want to add people (not objects) to a scene and control the lighting completely in software. The models will be lit completely shadowless as much as possible. So we have:
1. Shoot the background, room, client set, whatever.
2. Shoot model in the studio in a completely shadowless way.
3. Add model to background image.
4. Using a software program add light & shadow to the model.
Is this possible? I know how to do everything but step 4.
The 2D way...
The only way I can think of to do this would be to cut the 'neutrally lit' model out of the studio shot in Photoshop... place it in the room / set... then to simulate directional light
apply 2 adjustment layers (Curves or Levels - what ever you prefer to use) to the model. One curve for the highlights and the other for shading.
Then Airbrush where you want these contrasts to be... may take some drawng / illustrative skills.
And then, either airbrush a shadow layer underneath the model, or there may be some plugin software that simulates cast shadows... I am sure there used to be a filter/plugin for that.
The 3D way...
You can cut the model out of your studio shot and place the image onto a flat 3D card / plane object (that the shape of the model masks out) in your 3D package and it will render with the scene.
It will cast shadows and reflections, and the shadows will follow the contour of the objects it falls on, but will not be as "thick" if the light is on a side angle to the model.
Neither will the model have correct shading, so this will have to be done in photoshop like the '2D' way.
Example attached.
Probably just as easy to go the photoshop way...
Hope that gives you something to think about.
cheers
-
Oooops... sorrry I thought you meant a 3D room as well....
Looks like it all photoshop shading then.
-
Thank you gentlemen. Yes I was referring to a 2D background and 2D model. Some of the points raised could perhaps be done by software one day. Just select out a human form (the 2D of the model you shot in the studio) and the software will create a 3D model with a 25inch waist! Then when I drop the model onto my 2D background I grab a "light tool" and place it where I want and the angle I want. Seems like something close enough could be programmed. That would be very slick...
But light from the 2D background hitting the model... yeah that sounds tough...
-
Thank you gentlemen. Yes I was referring to a 2D background and 2D model. Some of the points raised could perhaps be done by software one day. Just select out a human form (the 2D of the model you shot in the studio) and the software will create a 3D model with a 25inch waist! Then when I drop the model onto my 2D background I grab a "light tool" and place it where I want and the angle I want. Seems like something close enough could be programmed. That would be very slick...
But light from the 2D background hitting the model... yeah that sounds tough...
There is such a thing as a scriber. trace your object and vwaalllaa!
-
The take-home point from this is that specialists are much more common than those who excel in everything.
I've said it before: that photographers make beautiful pictures, but make for poor cinematographers and shitty story-tellers. I ranted about this when Reverie came out. I understand it was mostly a tech demo, but many hailed it as a beginning of the end, just like this (visually stunning) CGI clip. Rather than the end, it was the beginning of a flood of often photographically good, but poorly conceived and dulldulldull clips from people who think a 5DII or 7D turns them into an oliver stone. And don't get me started on the orgy of ultra-shallow DOF - it must be good because it can be done.
I've seen similar CGI clips before, but I didn't believe it was CGI until I read numerous comments and saw making-ofs, so there's definitely something unique here. Just like Reverie, this is a great proof of concept, this time that someone can make believable CGI alone. That's it. There's no story, no substance, and gratuitous camera trickery which would make a freshman cinematography student cringe. All this in no way takes away from the clip being a freakishly impressive accomplishment; it's just very one-dimensional.
Further, what most commentators seem to miss is that apparently Mr Roman took a year off to produce this. Certainly a prudent investment for him, as I'm sure this will open many doors to jobs, and deservedly so. But making a 12-minute music video with fifty thousand or so euros worth of work hours isn't trivial. It will take years and years until this can be done with 3D images and video without a massive investment in time and money - and even then we still need artists to create the story, the substance.
It is extremely rare to find someone who can do it all, and even rarer who's good at more than one aspect of creativity. There's a reason why even the cheapest movies have a crew, rather than a single guy making it. Even the rare coppolas and manns who can produce, write, shoot, and/or direct movies rely on an army of creative experts.
-
Help me understand here it seems from many comments here that CGI is only economically viable versus photography when there are 3d models available for the other work to be built upon it. That would explain why cars and architecture can easily migrate to CGI, because part of the product design involves making 3d models. And that also explains why CGI would be useless for things like food and people which cannot be easily 3d molded and animated. Hope I got that right.
-
Help me understand here it seems from many comments here that CGI is only economically viable versus photography when there are 3d models available for the other work to be built upon it. That would explain why cars and architecture can easily migrate to CGI, because part of the product design involves making 3d models. And that also explains why CGI would be useless for things like food and people which cannot be easily 3d molded and animated. Hope I got that right.
It isn't quite that black and white... There is work done on CG humans, but it is incredibly labour intensive, and I have not yet seen work where you couldn't tell that it was fake at some point. Still, there are many valid reasons for using CG. With humans, one reason would be that the person is not available any more, like the digital Marlene project. Another might be that they cannot participate for other reasons, and so on. I have also seen 100% convincing food CG, but it would have been so much easier with photography. Sometimes something is done just to prove that it can be done, which is a valid reason, albeit not so extensible to everyday situations.
In general, the automotive and architectural fields are natural for CG, but there are other fields where it could also be used. One reason for using CG which already came up in this thread is to do something which cannot be done with photography, such as impossible camera moves. Another is to get a bunch of elements together which in real life might not come together in the time allotted for the shot. One example could be if a sick tree has been cut down in front of a building, but you want to have it in the shot, and so on.
Ultimately, the CG side of things will continue to advance at a rapid pace, but certain things will always remain very labour-intensive and better suited to photography.
For the artists, there is also the consideration of which is more satisfying, and photography has a huge advantage there. Sitting in front of a computer for hours, days, months, perfecting a model or animation, or a look, is not nearly as satisfying for most people as being on-site.
-
It isn't quite that black and white... There is work done on CG humans, but it is incredibly labour intensive, and I have not yet seen work where you couldn't tell that it was fake at some point.
I was quite convinced at this tech demo after the actress/spokesperson first explains the process, here (http://www.youtube.com/watch?v=bLiX5d3rC6o).
-
I was quite convinced at this tech demo after the actress/spokesperson first explains the process, here (http://www.youtube.com/watch?v=bLiX5d3rC6o).
I have seen that demo before, and it is impressive, but keep in mind that contrary to the claims that her whole face is being simulated by technology (whatever that is supposed to mean), in fact they have motion-captured a specific performance, and replaced the face of the actress with a model of... herself. These are not very trying circumstances. I am willing to bet that the time invested was non-trivial, and there is much which cannot be automated. Now try to extend that to a digital character in a movie, or some other, larger scenario. Sure, technology advances inexorably, but people still need to do much of the work.
As for the rendering in that movie, something is not right about the lighting on her teeth; it is too bright. This is a common problem.