i tried to start a renderings thread a while back and kinda got shut down by one of the 'nect admins (something along the lines of "there is a gallery for renderings"). the problem with the gallery is that, more often than not, it doesn't say who the image was created/posted by, what techniques they used to make the image, etc. so, in the hopes that some of you are as interested in rendering techniques as i am, here's another shot at starting a discussion/gallery concerning such techniques.
(i'm working on next post after having creating the topic... i can't post images in the first post for some reason... gimme a minute).
okay, so, the inspiration for trying to get this thread going again was this image, which i saw this evening in the galleries:
absolutely stunning in my opinion. i have read some brief mentions of piranesi in the gallery from time to time, but this image made me look up the program online. i had no idea what the software is for (i'm still a little blurry on it, actually), but after visiting the company's site and looking at the gallery, it appears to allow you to 'paint' a rendering, either in post-production or (somehow?!) actually import a model and set it up pre-render. these two functions seem entirely different and to require entirely different types of interfaces (2d post-production vs. 3d pre-rendering). i am, to say the least, intrigued. if anyone uses this software frequently, please share some of your experience with it.
to speak briefly to why i find the image above so stunning: the abstract composition of the background, the starkly contrasting, semi-reflective (not exactly reflecting what is above/behind it) foreground, the pastel-drawn quality of the characters all offset the pure (ack, i know, talking about composition can get messy sometimes, bear with me) and simple lines of the architecture in a remarkably cohesive and graphic way.
i work in 3ds max and then do post-production in photoshop, usually working mainly on levels, saturation, adding highlights, shade, shadow, etc. where necessary. this has worked pretty well for me and i am very happy with how my rendering abilities have developed in the past two years, but it seems like a program like piranesi offers a very exciting sort of middle-ground type of production somewhere between pre-rendered 3d and post-prod. 2d that can possibly increase the artistic possibilities (though, i have to say, most of the renderings in the company's galleries looked somebody had a bit too much fun with a bunch of photoshop filters).
so, if the person who made this image is reading, how did you create this image (more specifically than your descriptions: "Modeling exercise from magazine photo. Modeled in SketchUp, rendered in Lightwave and finished in Piranesi. Syd Mead was the inspiration for this particular image.")? if i was to do it using my 3ds, photoshop method, it would involve spending a lot of time composing the graphic background to use as an environment map in max, then simply rendering the scene and doing post-op in photoshop... where does piranesi fit in/what did it do to the image?
i will gladly post some of my renderings to add to the discussion as soon as i get photoshop back up and running on my machine... i formatted recently and haven't reinstalled all my programs yet :(
Given that the architectural model is quite simple, it would be quite easy to do most of the work in photoshop, and save a lot of time probably, and yes i dont think those people are from round here. I guess they must be the developers!
But to be honest, i'm not really sure the background really adds anything- its not a landscape, but it doesnt really suggest an anti context point of view- for me it ends up in the middle.
that horrible middle ground - suburbia..
the background is the part that I like. The rest is kind of bland without the contrast. The glass needs more work. Perspective looks whacked.
It's interesting and catches your eye, so in that regard it's successful. But I think some of the time should have been spent on the building and it's immediate setting, not just a pretty pic in the bg.
i agree, the image is not correct photographically in its perspective, but that could be corrected with 5 seconds of photoshop. i actually think the characters work quite well. the whole image is very balanced stylistically. the whole thing has a painted quality... even through the clean-lined architecture the image maintains its painterliness, and i think that's why this image sings so much: its consistent graphic style while blending a sort of maximalist background and a minimalist foreground. i think the sharp horizon line definitely helps achieve this. the characters, if anything, seem like they could live in the same world as the background, but in a more civilized part of it (like maybe the house behind them).
still waiting for some technical feedback on piranesi's role in this whole thing...
does anyone attempt to do renderings by hand anymore? I have always loved the feel of pencil on heavy paper and just wonder if it is becoming a lost art?
every semester i sort of tell myself that i'm going to make some awesome pencil renderings, but the type of design/produce schedule that the computer has created at columbia (and i'm sure a lot of other schools) makes it all but impossible. with a 3d model i can crank out all the perspectives i need for a presentation, interior and exterior, with a high level of detail in a weekend. i'd be lucky to get one fantastic hand-drawn perspective done in a weekend. the critics know this, and prefer a quantity of output that allows for a broad discussion of your project in review, and one or two pencil renderings doesn't cut the mustard anymore.
despite the fact that i really miss drawing by hand and will likely pick it up again sometime in the future, i really enjoy exploring the possibilities of creating images digitally and don't feel that it's any less of an art, as some people feel.
one of our principals can do airbrush quite well. he did one a while back. he is also very good at colored pencil/marker and he can do it fast.
A rival firm does all their renderings in really cheesy free hand perspective marker drawings.
Our interior designers, all out of one particular school, are very proficient in marker renderings. It is all technique, some skill, lots of practice. They learned in school - god forbid someone would learn something marketable like that in architecture school.
I do the majority of my firms renderings and have done several dozen by now, but don't know how to post one. I'll see if I can figure it out when I have time.
You can upload images to imageshack from your computer and it will host them for free. You can then link them in these forums.. I'm not sure how they make any money, but it works for now..
i like the background graphic too.
the one thing i would change is the people. Make them walk towards the bldg. and not away from it. The image feels lonely.. maybe put some more people in the bldg. too.
Is that thing sitting on a sheet of ice? Is there some kind of magic tree that grows out of the ice like that? The image seems an odd mix of abstraction (the environment) and realism (materials and those hideous people out of entourage)
it reminds me of one of the images in the tunnel in "charlie in the chocolate factory" that damn wille wanka and his crazy boat and tunnel of nightmares
I agree, the acid flashback in the background does little to relate the building to the site. Its akin to the architectural animations I sometimes see in which the grass is blowing in the wind and the sky is passing overhead...and the building is for shit b/c the designer was too caught up in making a 'sexy image'
I like the image. I think it's an interesting mix between the hyper-realistic renderings and the overly abstract renderings that are out there today. Like all renderings, I think it would only work for certain clients, but I think it's a succesful image.
/>
I am a first year MArch student s be gentle, you know our ego's are easily bruised and then we run 3 pints of Hagen Daz Dulce De Leche Ice Cream.
I did this one while working on this town house project, it is a study to see how much sunlight I can sneak into the houses giving the existing southern building.
Rob
Lets try this one more time. Been awhile since I played with html, sorry
I did this one while working on this town house project, it is a study to see how much sunlight I can sneak into the houses giving the existing southern building.
Rob
not bad, I think it's quite nice in a way, but the fact that you have one room lit up so strongly, shining a light up and out, suggests the wrong thing perhaps? To me, it doesn't ooze sex appeal, which I assume is that this one should be all about.
I would lose the the window frame. lie. just make it a pane of glass. Seriously, I never considered doing this until now, but just lie, it's what all the best students/architects often do, I hate to say it, and I don't want to do it, but it just ruins the idea I think.
In fact, it's more telling 'certain truths' than telling lies, I've realised this is how you get marks/clients, though I'd love for someone to tell me I'm wrong...lying sucks!
Unless the idea is to show how to subdivide a large area of glazing
The bar (handrail?) looks wrong and out of place, plus I don't see a door onto this balcony, if that's what it is?
in layman words, it a bit too dark are you trying to render a night view??? in thish caseit might make sense to wash the light over the cielind of the first floor.
The frames look too commonplace imho - you can either lose them, or else add a lot of detail to show what the frames would be life in actual life. Some reflectivity on the glass will also help.
good call on bumping up the reflectivity; even if you're not going for an ultra-realistic rendering, reflective glass always adds a layer of visual appeal. i would also make the frame elements more responsive to light (more glossy/specular/reflective/etc.).
also, to expand s.o.d.'s comments on the darkness of the image: for an exterior night shot such as this, you have to either really amp up the lights inside, making the interior the obvious focus of the shot, or add some local lights outside to brighten up the exterior in a manner appropriate to outdoor lighting.
as a general comment, having your walls angle sharply inwards is an amateur photographic mistake that most people make without either noticing or, if they notice it, without knowing how to correct it. in the case of digital images such as these, it's as easy as doing a perspective-correcting crop in photoshop. for analog photo-taking, they make crazy expensive cameras that can compensate for lens distortion. you'll find that, way more often than not, architectural images appear much more professional, realistic and convincing when presented in such a manner (unless you are trying to evoke the sense of vertigo that one will experience while standing nearby the project).
I rally like the image- i think what is really great about it is the flatness of the areas of monotone colours, which is quite refreshing i think-
usually i feel that making everything glossy improves the image and makes it look very seductive and sexy, but not in this case, partly because if you add reflectivity to the glass, it will emphasise the barrier between the inside and the outside more, and will not let you see the architecture so clearly. So although it may be technically the right thing to do- i dont think you should do it. I think another thing with renders is that increasingly renderings look more and more the same and they are often not always about accurately showing the building, but in my experience they are used quite often to sell the building. So therefore if you come up with a render that looks a little different- that can be a really positive thing. Having said that- if the architecture has to rely on a night shot to sell itself there is usually a problem with the architecture....
How about in this post someone posts a model, and we all render it quite quickly and then have a discussion about that- then at least we all start from the same point- and i could be really useful.
i agree w/ mao... if anybody has a relatively simple model they could host, i think we should all render out a couple of images and post them, learn from each other's techniques. i suppose i could get around to making a simple model, if everybody's cool working with .max files. maybe something along the lines of the building in the first rendering posted on this thread? it seems the maker of that image isn't reading this, but if he was, he could just host his model...
anyhow, here are three images from a project i'm currently working on in studio. the project is (an international competition for) a pedestrian border crossing station/bridge between mexico and the US. the first image is an initial sketch (part of my process is mocking up sketch models in max and doing a fairly high quality rendering to get a sort of immediate, zoomed-in look at an idea. these usually take a couple of hours to model and render), the second image is one i used in the competition entry about a month ago (design fairly well-honed by that point), and the third image is another sketch, post-competition, when we have begun to add supplementary programs (theater/forum, shopping).
in this first sketch, the thing i spent the most time on was mapping the corrugated metal. first i took a relatively hi-res photograph of the material from an internet source. then i did a perspective-correcting crop to make sure all of the corrugations were perfectly horizontal. at that point it was ready to bring into 3ds max, where i uvw mapped it to the geometry itself (which was created in max by extruding a drawn corrugated section), being very careful to line up the highlights in the map photograph to the edges of the corrugations, where they would naturally lie. again, this was a sketched image and, all told, i spent maybe 6 hours on modelling, materials and rendering.
this image is a rendering of a model of my entire project, which took several days to model. at this point the project had shifted to being made up entirely of expanded metal fence and, as such, i spent a long time honing the mesh material for this image. again taking an internet image, killing the perspective on it, working on levels/contrast, and, to tile the material in max, cropping it so that it would mirror perfectly in all directions. the image map itself is doing most of the work, with a simple opacity map and a carefully-made bump map (which, i think, has succeeded in giving the fence a little depth... though it's so subtle i think i might be kidding myself). the white ground is a plane with a shadow/matte material applied, which, when the render is outputted as a file type that preserves alpha channel information (i use .png), the shadows basically render as objects with their proper opacities which can then be layered on top of other images (i used a subtle ground texture as a background for my competition boards, and i wanted the images to have almost a spray-painted-on-the-wall quality to them).
and this last image, as i mentioned, was another sketch. i did this one for the first desk crit after we sent the competition boards in, when we added the new programs. i was running a bit late that day, and threw this together in about two hours before class. basically it's an imported acad plan with most of the shapes extruded to form the earthwork. the slanty concrete thingies were made in max by creating a simple plane and dragging vertices around, then applying a shell modifier to give it all a thickness. i rendered this using mental ray, with low samples (which gives the earth material, which is actually just a plain yellow color, some visual depth) and high energy (to get some more lighting action in this relatively shadowy scene). i forgot to mention until now: for all my outdoor renderings i use only a simple skylight for fill light and a direct light for my shadows. pretty basic. since i was in a rush, i did most of the fine tuning of this image in photoshop (a lot of levels work and some saturation correction). the hazy glass reflections were done in photoshop with a low-opacity white layer for glass and a slightly-higher-opacity layer with some white, rough cloud shapes created by a single, really large stamp of a messy photoshop brush (again, as with the fence bump map, may be too subtle to really notice, but if you saw the original you'd definitely see the difference).
i'd be happy to answer any questions about these images, and equally happy for some constructive criticism.
gonna post again for convenience. sorry about that. the 'preview' button sure is helpful...
i agree w/ mao... if anybody has a relatively simple model they could host, i think we should all render out a couple of images and post them, learn from each other's techniques. i suppose i could get around to making a simple model, if everybody's cool working with .max files. maybe something along the lines of the building in the first rendering posted on this thread? it seems the maker of that image isn't reading this, but if he was, he could just host his model...
anyhow, here are three images from a project i'm currently working on in studio. the project is (an international competition for) a pedestrian border crossing station/bridge between mexico and the US. the first image is an initial sketch (part of my process is mocking up sketch models in max and doing a fairly high quality rendering to get a sort of immediate, zoomed-in look at an idea. these usually take a couple of hours to model and render), the second image is one i used in the competition entry about a month ago (design fairly well-honed by that point), and the third image is another sketch, post-competition, when we have begun to add supplementary programs (theater/forum, shopping).
in this first sketch, the thing i spent the most time on was mapping the corrugated metal. first i took a relatively hi-res photograph of the material from an internet source. then i did a perspective-correcting crop to make sure all of the corrugations were perfectly horizontal. at that point it was ready to bring into 3ds max, where i uvw mapped it to the geometry itself (which was created in max by extruding a drawn corrugated section), being very careful to line up the highlights in the map photograph to the edges of the corrugations, where they would naturally lie. again, this was a sketched image and, all told, i spent maybe 6 hours on modelling, materials and rendering.
this image is a rendering of a model of my entire project, which took several days to model. at this point the project had shifted to being made up entirely of expanded metal fence and, as such, i spent a long time honing the mesh material for this image. again taking an internet image, killing the perspective on it, working on levels/contrast, and, to tile the material in max, cropping it so that it would mirror perfectly in all directions. the image map itself is doing most of the work, with a simple opacity map and a carefully-made bump map (which, i think, has succeeded in giving the fence a little depth... though it's so subtle i think i might be kidding myself). the white ground is a plane with a shadow/matte material applied, which, when the render is outputted as a file type that preserves alpha channel information (i use .png), the shadows basically render as objects with their proper opacities which can then be layered on top of other images (i used a subtle ground texture as a background for my competition boards, and i wanted the images to have almost a spray-painted-on-the-wall quality to them).
and this last image, as i mentioned, was another sketch. i did this one for the first desk crit after we sent the competition boards in, when we added the new programs. i was running a bit late that day, and threw this together in about two hours before class. basically it's an imported acad plan with most of the shapes extruded to form the earthwork. the slanty concrete thingies were made in max by creating a simple plane and dragging vertices around, then applying a shell modifier to give it all a thickness. i rendered this using mental ray, with low samples (which gives the earth material, which is actually just a plain yellow color, some visual depth) and high energy (to get some more lighting action in this relatively shadowy scene). i forgot to mention until now: for all my outdoor renderings i use only a simple skylight for fill light and a direct light for my shadows. pretty basic. since i was in a rush, i did most of the fine tuning of this image in photoshop (a lot of levels work and some saturation correction). the hazy glass reflections were done in photoshop with a low-opacity white layer for glass and a slightly-higher-opacity layer with some white, rough cloud shapes created by a single, really large stamp of a messy photoshop brush (again, as with the fence bump map, may be too subtle to really notice, but if you saw the original you'd definitely see the difference).
i'd be happy to answer any questions about these images, and equally happy for some constructive criticism.
very nice joed, I don't do 3d anymore, don't have the energy, but I like what you're done, though it's hard to guage the project from just these of course, not that this is the point here!
can you expand a bit on the alpha channel concept? i.e., what is it, how do you use it, what other file formats support it other than .png, how well does it translate across multiple programs (max to photoshop or formz to photoshop etc....), why the necessity of a "shadow/matte" plane(and what exactly do you mean by "shadow/matte", (i think i realize the matte part but do you mean the plane accepts shadows?)
i like the idea of posting a model, no time this morning...maybe i can later this evening
Targa supports alphas, when you save as a .TGA it gives you a choice of 24 bit (no alpha) or 32 bit (with alpha) .
In PS you ca find your channels between the Layer pallette and the Path pallette, normally you dont have an alpha channel, if you select the are you want to be transparent, and go to the bottom of the selection menu, located acroos the top between Layers and Filters, you can "save selection" . When using the default settings this creates an alpha channel. You can edit an alpha channel by clickking on it and making sure the RGB or CMYK channels are turned off, basically an alpha channel designates absolute black as 100% transparent and absolute white as 100% opaque, and gradients in between are treated accordingly.
3DMax will create alphas from what ever is background no mater what color you designate the background to be, if you choose a file format that supports alphas.
I should learn more about .PNGs it sounds like a powerful file format but I am used to working with .TGAs ( old school, video).
Basically you import your 3D model into Piranesi and actually paint on the surfaces of the model. You can also treat the model as a 2D drawing and apply textures from a certain view/perspective and it will use the 3D information to do things like highlight edges in front of other edges, blur receeding edges, etc. Also you can place scale figures throughout the model and it will automatically scale/rotate them for you.
This was a piss-poor explaination of the way it works, but no one else was piping up, so I gave it a shot. I think it sometime gets a bad rep because most people use it to "watercolor" over their models and it ends up looking like amateurish crap. Here's a rare decent example, and this one's not bad either.
Like any other software, the output is as good as the person at the keyboard.
to be honest i am not too sure about the difference between .png format and 32 bit (alpha embedded) .tga.
to clarify a bit: any transparent/translucent conditions in a model have alpha (transparency) properties. when you save a rendered image (in 3ds max, the only modelling/rendering software i know), you can choose a file type that will embed this information.
say, for example, you have a glass office tower, and you are rendering a stree-level perspective, looking up at the tower (no environment/sky map, just white). then, you save the rendering as a .png or 32 bit .tga (or 32 bit .tif also, if i remember correctly). when you open the image in photoshop and put another image layer, say a sky, behind the tower, the transparency properties of the glass material will be retained; you will see the sky through the glass as much as the glass should let you see the sky.
shadow/matte is a type of shader in 3ds max.
open the material editor in max and follow my rad red arrow to the box that says "standard." click it and a window will open with a list of different material shader types. double click on 'matte/shadow' and you're ready to go.
what matte/shadow does is, when applied to an object (say a ground plane), it will make the entire object render invisible, but it will 'capture' any shadows that would land on that surface. if you save as an appropriate type, the alpha info (transparency info) will be retained as well. this is really handy if you want to do a lot of photoshopping. in this example, you might want to create your own ground plane with a few different images of grass in photoshop. with a matte/shadow shader, you can then place (as another layer) your rendered model with its proper shadows over your grass layers in photoshop.
if you had rendered the image with a 'standard' material applied to the ground plane, you would need to work carefully around the shadows in order to not draw over them if you wanted to add to/change the ground plane in photoshop. with matte/shadow, it basically makes the shadows into objects and gets rid of any non-shadow ground plane so you have more freedom to place your object and its shadows into another image.
here's an image to give you an idea. the ball was sitting on a plane that had a matte/shadow shader applied to it. thus, the shadow has been rendered as an object with transparency, or alpha, properties intact, and the rest of the ground plane doesn't exist.
a masterpiece, i know [*bows*]...
hope this helps. i know it sounds a little complicated, but it's really not bad at all.
Mar 21, 05 3:55 pm ·
·
Block this user
Are you sure you want to block this user and hide all related comments throughout the site?
Archinect
This is your first comment on Archinect. Your comment will be visible once approved.
renderings... talk amongst yourselves
i tried to start a renderings thread a while back and kinda got shut down by one of the 'nect admins (something along the lines of "there is a gallery for renderings"). the problem with the gallery is that, more often than not, it doesn't say who the image was created/posted by, what techniques they used to make the image, etc. so, in the hopes that some of you are as interested in rendering techniques as i am, here's another shot at starting a discussion/gallery concerning such techniques.
(i'm working on next post after having creating the topic... i can't post images in the first post for some reason... gimme a minute).
okay, so, the inspiration for trying to get this thread going again was this image, which i saw this evening in the galleries:
absolutely stunning in my opinion. i have read some brief mentions of piranesi in the gallery from time to time, but this image made me look up the program online. i had no idea what the software is for (i'm still a little blurry on it, actually), but after visiting the company's site and looking at the gallery, it appears to allow you to 'paint' a rendering, either in post-production or (somehow?!) actually import a model and set it up pre-render. these two functions seem entirely different and to require entirely different types of interfaces (2d post-production vs. 3d pre-rendering). i am, to say the least, intrigued. if anyone uses this software frequently, please share some of your experience with it.
to speak briefly to why i find the image above so stunning: the abstract composition of the background, the starkly contrasting, semi-reflective (not exactly reflecting what is above/behind it) foreground, the pastel-drawn quality of the characters all offset the pure (ack, i know, talking about composition can get messy sometimes, bear with me) and simple lines of the architecture in a remarkably cohesive and graphic way.
i work in 3ds max and then do post-production in photoshop, usually working mainly on levels, saturation, adding highlights, shade, shadow, etc. where necessary. this has worked pretty well for me and i am very happy with how my rendering abilities have developed in the past two years, but it seems like a program like piranesi offers a very exciting sort of middle-ground type of production somewhere between pre-rendered 3d and post-prod. 2d that can possibly increase the artistic possibilities (though, i have to say, most of the renderings in the company's galleries looked somebody had a bit too much fun with a bunch of photoshop filters).
so, if the person who made this image is reading, how did you create this image (more specifically than your descriptions: "Modeling exercise from magazine photo. Modeled in SketchUp, rendered in Lightwave and finished in Piranesi. Syd Mead was the inspiration for this particular image.")? if i was to do it using my 3ds, photoshop method, it would involve spending a lot of time composing the graphic background to use as an environment map in max, then simply rendering the scene and doing post-op in photoshop... where does piranesi fit in/what did it do to the image?
i will gladly post some of my renderings to add to the discussion as soon as i get photoshop back up and running on my machine... i formatted recently and haven't reinstalled all my programs yet :(
the image got a bit distorted and cropped for some reason. let's try it again:
ouch. it was a lot better without the people. they look like they just stepped out of an ad for health insurance.
Given that the architectural model is quite simple, it would be quite easy to do most of the work in photoshop, and save a lot of time probably, and yes i dont think those people are from round here. I guess they must be the developers!
But to be honest, i'm not really sure the background really adds anything- its not a landscape, but it doesnt really suggest an anti context point of view- for me it ends up in the middle.
that horrible middle ground - suburbia..
the background is the part that I like. The rest is kind of bland without the contrast. The glass needs more work. Perspective looks whacked.
It's interesting and catches your eye, so in that regard it's successful. But I think some of the time should have been spent on the building and it's immediate setting, not just a pretty pic in the bg.
i agree, the image is not correct photographically in its perspective, but that could be corrected with 5 seconds of photoshop. i actually think the characters work quite well. the whole image is very balanced stylistically. the whole thing has a painted quality... even through the clean-lined architecture the image maintains its painterliness, and i think that's why this image sings so much: its consistent graphic style while blending a sort of maximalist background and a minimalist foreground. i think the sharp horizon line definitely helps achieve this. the characters, if anything, seem like they could live in the same world as the background, but in a more civilized part of it (like maybe the house behind them).
still waiting for some technical feedback on piranesi's role in this whole thing...
does anyone attempt to do renderings by hand anymore? I have always loved the feel of pencil on heavy paper and just wonder if it is becoming a lost art?
every semester i sort of tell myself that i'm going to make some awesome pencil renderings, but the type of design/produce schedule that the computer has created at columbia (and i'm sure a lot of other schools) makes it all but impossible. with a 3d model i can crank out all the perspectives i need for a presentation, interior and exterior, with a high level of detail in a weekend. i'd be lucky to get one fantastic hand-drawn perspective done in a weekend. the critics know this, and prefer a quantity of output that allows for a broad discussion of your project in review, and one or two pencil renderings doesn't cut the mustard anymore.
despite the fact that i really miss drawing by hand and will likely pick it up again sometime in the future, i really enjoy exploring the possibilities of creating images digitally and don't feel that it's any less of an art, as some people feel.
one of our principals can do airbrush quite well. he did one a while back. he is also very good at colored pencil/marker and he can do it fast.
A rival firm does all their renderings in really cheesy free hand perspective marker drawings.
Our interior designers, all out of one particular school, are very proficient in marker renderings. It is all technique, some skill, lots of practice. They learned in school - god forbid someone would learn something marketable like that in architecture school.
I do the majority of my firms renderings and have done several dozen by now, but don't know how to post one. I'll see if I can figure it out when I have time.
You can upload images to imageshack from your computer and it will host them for free. You can then link them in these forums.. I'm not sure how they make any money, but it works for now..
i like the background graphic too.
the one thing i would change is the people. Make them walk towards the bldg. and not away from it. The image feels lonely.. maybe put some more people in the bldg. too.
Is that thing sitting on a sheet of ice? Is there some kind of magic tree that grows out of the ice like that? The image seems an odd mix of abstraction (the environment) and realism (materials and those hideous people out of entourage)
the architecture itself leaves something to be desired.
the background is cool, i don't know that cool=good though
it reminds me of one of the images in the tunnel in "charlie in the chocolate factory" that damn wille wanka and his crazy boat and tunnel of nightmares
I agree, the acid flashback in the background does little to relate the building to the site. Its akin to the architectural animations I sometimes see in which the grass is blowing in the wind and the sky is passing overhead...and the building is for shit b/c the designer was too caught up in making a 'sexy image'
I like the image. I think it's an interesting mix between the hyper-realistic renderings and the overly abstract renderings that are out there today. Like all renderings, I think it would only work for certain clients, but I think it's a succesful image.
/>
I am a first year MArch student s be gentle, you know our ego's are easily bruised and then we run 3 pints of Hagen Daz Dulce De Leche Ice Cream.
I did this one while working on this town house project, it is a study to see how much sunlight I can sneak into the houses giving the existing southern building.
Rob
Lets try this one more time. Been awhile since I played with html, sorry
I did this one while working on this town house project, it is a study to see how much sunlight I can sneak into the houses giving the existing southern building.
Rob
interesting image but,
a lesson to all:
There is no "Design" button on any rendering program.
Shit rendered in the fanciest modeler will still be shit.
thx for the lesson, wise one.
i'm just trying to get at rendering techniques here... i don't see how criticism of the architecture is going to help anything.
so, does anybody have experience using piranesi? what is it useful for?
What do you people think of this image?:
its a night vue to the living room / kitchen of a house i designed during my second semester (that i later rendered in 3d for another class).
i did everything with formZ (no photoshop etc). please let me know how i could improve the image. thanks
erm, it's gone off the page!
not bad, I think it's quite nice in a way, but the fact that you have one room lit up so strongly, shining a light up and out, suggests the wrong thing perhaps? To me, it doesn't ooze sex appeal, which I assume is that this one should be all about.
I would lose the the window frame. lie. just make it a pane of glass. Seriously, I never considered doing this until now, but just lie, it's what all the best students/architects often do, I hate to say it, and I don't want to do it, but it just ruins the idea I think.
In fact, it's more telling 'certain truths' than telling lies, I've realised this is how you get marks/clients, though I'd love for someone to tell me I'm wrong...lying sucks!
Unless the idea is to show how to subdivide a large area of glazing
The bar (handrail?) looks wrong and out of place, plus I don't see a door onto this balcony, if that's what it is?
ummmm.... damn i guess its not clear: the balcony (it is a balcony) has 2 sliding doors.. i probably should have given them a thicker frame (oops).
in layman words, it a bit too dark are you trying to render a night view??? in thish caseit might make sense to wash the light over the cielind of the first floor.
The frames look too commonplace imho - you can either lose them, or else add a lot of detail to show what the frames would be life in actual life. Some reflectivity on the glass will also help.
good call on bumping up the reflectivity; even if you're not going for an ultra-realistic rendering, reflective glass always adds a layer of visual appeal. i would also make the frame elements more responsive to light (more glossy/specular/reflective/etc.).
also, to expand s.o.d.'s comments on the darkness of the image: for an exterior night shot such as this, you have to either really amp up the lights inside, making the interior the obvious focus of the shot, or add some local lights outside to brighten up the exterior in a manner appropriate to outdoor lighting.
as a general comment, having your walls angle sharply inwards is an amateur photographic mistake that most people make without either noticing or, if they notice it, without knowing how to correct it. in the case of digital images such as these, it's as easy as doing a perspective-correcting crop in photoshop. for analog photo-taking, they make crazy expensive cameras that can compensate for lens distortion. you'll find that, way more often than not, architectural images appear much more professional, realistic and convincing when presented in such a manner (unless you are trying to evoke the sense of vertigo that one will experience while standing nearby the project).
I rally like the image- i think what is really great about it is the flatness of the areas of monotone colours, which is quite refreshing i think-
usually i feel that making everything glossy improves the image and makes it look very seductive and sexy, but not in this case, partly because if you add reflectivity to the glass, it will emphasise the barrier between the inside and the outside more, and will not let you see the architecture so clearly. So although it may be technically the right thing to do- i dont think you should do it. I think another thing with renders is that increasingly renderings look more and more the same and they are often not always about accurately showing the building, but in my experience they are used quite often to sell the building. So therefore if you come up with a render that looks a little different- that can be a really positive thing. Having said that- if the architecture has to rely on a night shot to sell itself there is usually a problem with the architecture....
How about in this post someone posts a model, and we all render it quite quickly and then have a discussion about that- then at least we all start from the same point- and i could be really useful.
i agree w/ mao... if anybody has a relatively simple model they could host, i think we should all render out a couple of images and post them, learn from each other's techniques. i suppose i could get around to making a simple model, if everybody's cool working with .max files. maybe something along the lines of the building in the first rendering posted on this thread? it seems the maker of that image isn't reading this, but if he was, he could just host his model...
anyhow, here are three images from a project i'm currently working on in studio. the project is (an international competition for) a pedestrian border crossing station/bridge between mexico and the US. the first image is an initial sketch (part of my process is mocking up sketch models in max and doing a fairly high quality rendering to get a sort of immediate, zoomed-in look at an idea. these usually take a couple of hours to model and render), the second image is one i used in the competition entry about a month ago (design fairly well-honed by that point), and the third image is another sketch, post-competition, when we have begun to add supplementary programs (theater/forum, shopping).
in this first sketch, the thing i spent the most time on was mapping the corrugated metal. first i took a relatively hi-res photograph of the material from an internet source. then i did a perspective-correcting crop to make sure all of the corrugations were perfectly horizontal. at that point it was ready to bring into 3ds max, where i uvw mapped it to the geometry itself (which was created in max by extruding a drawn corrugated section), being very careful to line up the highlights in the map photograph to the edges of the corrugations, where they would naturally lie. again, this was a sketched image and, all told, i spent maybe 6 hours on modelling, materials and rendering.
this image is a rendering of a model of my entire project, which took several days to model. at this point the project had shifted to being made up entirely of expanded metal fence and, as such, i spent a long time honing the mesh material for this image. again taking an internet image, killing the perspective on it, working on levels/contrast, and, to tile the material in max, cropping it so that it would mirror perfectly in all directions. the image map itself is doing most of the work, with a simple opacity map and a carefully-made bump map (which, i think, has succeeded in giving the fence a little depth... though it's so subtle i think i might be kidding myself). the white ground is a plane with a shadow/matte material applied, which, when the render is outputted as a file type that preserves alpha channel information (i use .png), the shadows basically render as objects with their proper opacities which can then be layered on top of other images (i used a subtle ground texture as a background for my competition boards, and i wanted the images to have almost a spray-painted-on-the-wall quality to them).
and this last image, as i mentioned, was another sketch. i did this one for the first desk crit after we sent the competition boards in, when we added the new programs. i was running a bit late that day, and threw this together in about two hours before class. basically it's an imported acad plan with most of the shapes extruded to form the earthwork. the slanty concrete thingies were made in max by creating a simple plane and dragging vertices around, then applying a shell modifier to give it all a thickness. i rendered this using mental ray, with low samples (which gives the earth material, which is actually just a plain yellow color, some visual depth) and high energy (to get some more lighting action in this relatively shadowy scene). i forgot to mention until now: for all my outdoor renderings i use only a simple skylight for fill light and a direct light for my shadows. pretty basic. since i was in a rush, i did most of the fine tuning of this image in photoshop (a lot of levels work and some saturation correction). the hazy glass reflections were done in photoshop with a low-opacity white layer for glass and a slightly-higher-opacity layer with some white, rough cloud shapes created by a single, really large stamp of a messy photoshop brush (again, as with the fence bump map, may be too subtle to really notice, but if you saw the original you'd definitely see the difference).
i'd be happy to answer any questions about these images, and equally happy for some constructive criticism.
doh!... guess i don't really know how to work the imageshack images properly. gimme a minute... testing...
gonna post again for convenience. sorry about that. the 'preview' button sure is helpful...
i agree w/ mao... if anybody has a relatively simple model they could host, i think we should all render out a couple of images and post them, learn from each other's techniques. i suppose i could get around to making a simple model, if everybody's cool working with .max files. maybe something along the lines of the building in the first rendering posted on this thread? it seems the maker of that image isn't reading this, but if he was, he could just host his model...
anyhow, here are three images from a project i'm currently working on in studio. the project is (an international competition for) a pedestrian border crossing station/bridge between mexico and the US. the first image is an initial sketch (part of my process is mocking up sketch models in max and doing a fairly high quality rendering to get a sort of immediate, zoomed-in look at an idea. these usually take a couple of hours to model and render), the second image is one i used in the competition entry about a month ago (design fairly well-honed by that point), and the third image is another sketch, post-competition, when we have begun to add supplementary programs (theater/forum, shopping).
in this first sketch, the thing i spent the most time on was mapping the corrugated metal. first i took a relatively hi-res photograph of the material from an internet source. then i did a perspective-correcting crop to make sure all of the corrugations were perfectly horizontal. at that point it was ready to bring into 3ds max, where i uvw mapped it to the geometry itself (which was created in max by extruding a drawn corrugated section), being very careful to line up the highlights in the map photograph to the edges of the corrugations, where they would naturally lie. again, this was a sketched image and, all told, i spent maybe 6 hours on modelling, materials and rendering.
this image is a rendering of a model of my entire project, which took several days to model. at this point the project had shifted to being made up entirely of expanded metal fence and, as such, i spent a long time honing the mesh material for this image. again taking an internet image, killing the perspective on it, working on levels/contrast, and, to tile the material in max, cropping it so that it would mirror perfectly in all directions. the image map itself is doing most of the work, with a simple opacity map and a carefully-made bump map (which, i think, has succeeded in giving the fence a little depth... though it's so subtle i think i might be kidding myself). the white ground is a plane with a shadow/matte material applied, which, when the render is outputted as a file type that preserves alpha channel information (i use .png), the shadows basically render as objects with their proper opacities which can then be layered on top of other images (i used a subtle ground texture as a background for my competition boards, and i wanted the images to have almost a spray-painted-on-the-wall quality to them).
and this last image, as i mentioned, was another sketch. i did this one for the first desk crit after we sent the competition boards in, when we added the new programs. i was running a bit late that day, and threw this together in about two hours before class. basically it's an imported acad plan with most of the shapes extruded to form the earthwork. the slanty concrete thingies were made in max by creating a simple plane and dragging vertices around, then applying a shell modifier to give it all a thickness. i rendered this using mental ray, with low samples (which gives the earth material, which is actually just a plain yellow color, some visual depth) and high energy (to get some more lighting action in this relatively shadowy scene). i forgot to mention until now: for all my outdoor renderings i use only a simple skylight for fill light and a direct light for my shadows. pretty basic. since i was in a rush, i did most of the fine tuning of this image in photoshop (a lot of levels work and some saturation correction). the hazy glass reflections were done in photoshop with a low-opacity white layer for glass and a slightly-higher-opacity layer with some white, rough cloud shapes created by a single, really large stamp of a messy photoshop brush (again, as with the fence bump map, may be too subtle to really notice, but if you saw the original you'd definitely see the difference).
i'd be happy to answer any questions about these images, and equally happy for some constructive criticism.
here's a closer look at the expanded metal material.
thanks for everyone's input. i'll rework the image and maybe post a second version later.
very nice joed, I don't do 3d anymore, don't have the energy, but I like what you're done, though it's hard to guage the project from just these of course, not that this is the point here!
nice work utilising 3 different programs / down and dirty, nuts and bolts renderings.
so... anybody up to posting a model for everybody to render and post/discuss?
also, is there nobody out there that uses piranesi?? i'm still waiting for an explanation/review of the program.
joed-
can you expand a bit on the alpha channel concept? i.e., what is it, how do you use it, what other file formats support it other than .png, how well does it translate across multiple programs (max to photoshop or formz to photoshop etc....), why the necessity of a "shadow/matte" plane(and what exactly do you mean by "shadow/matte", (i think i realize the matte part but do you mean the plane accepts shadows?)
i like the idea of posting a model, no time this morning...maybe i can later this evening
thanks
Targa supports alphas, when you save as a .TGA it gives you a choice of 24 bit (no alpha) or 32 bit (with alpha) .
In PS you ca find your channels between the Layer pallette and the Path pallette, normally you dont have an alpha channel, if you select the are you want to be transparent, and go to the bottom of the selection menu, located acroos the top between Layers and Filters, you can "save selection" . When using the default settings this creates an alpha channel. You can edit an alpha channel by clickking on it and making sure the RGB or CMYK channels are turned off, basically an alpha channel designates absolute black as 100% transparent and absolute white as 100% opaque, and gradients in between are treated accordingly.
3DMax will create alphas from what ever is background no mater what color you designate the background to be, if you choose a file format that supports alphas.
I should learn more about .PNGs it sounds like a powerful file format but I am used to working with .TGAs ( old school, video).
Basically you import your 3D model into Piranesi and actually paint on the surfaces of the model. You can also treat the model as a 2D drawing and apply textures from a certain view/perspective and it will use the 3D information to do things like highlight edges in front of other edges, blur receeding edges, etc. Also you can place scale figures throughout the model and it will automatically scale/rotate them for you.
This was a piss-poor explaination of the way it works, but no one else was piping up, so I gave it a shot. I think it sometime gets a bad rep because most people use it to "watercolor" over their models and it ends up looking like amateurish crap. Here's a rare decent example, and this one's not bad either.
Like any other software, the output is as good as the person at the keyboard.
to be honest i am not too sure about the difference between .png format and 32 bit (alpha embedded) .tga.
to clarify a bit: any transparent/translucent conditions in a model have alpha (transparency) properties. when you save a rendered image (in 3ds max, the only modelling/rendering software i know), you can choose a file type that will embed this information.
say, for example, you have a glass office tower, and you are rendering a stree-level perspective, looking up at the tower (no environment/sky map, just white). then, you save the rendering as a .png or 32 bit .tga (or 32 bit .tif also, if i remember correctly). when you open the image in photoshop and put another image layer, say a sky, behind the tower, the transparency properties of the glass material will be retained; you will see the sky through the glass as much as the glass should let you see the sky.
shadow/matte is a type of shader in 3ds max.
open the material editor in max and follow my rad red arrow to the box that says "standard." click it and a window will open with a list of different material shader types. double click on 'matte/shadow' and you're ready to go.
what matte/shadow does is, when applied to an object (say a ground plane), it will make the entire object render invisible, but it will 'capture' any shadows that would land on that surface. if you save as an appropriate type, the alpha info (transparency info) will be retained as well. this is really handy if you want to do a lot of photoshopping. in this example, you might want to create your own ground plane with a few different images of grass in photoshop. with a matte/shadow shader, you can then place (as another layer) your rendered model with its proper shadows over your grass layers in photoshop.
if you had rendered the image with a 'standard' material applied to the ground plane, you would need to work carefully around the shadows in order to not draw over them if you wanted to add to/change the ground plane in photoshop. with matte/shadow, it basically makes the shadows into objects and gets rid of any non-shadow ground plane so you have more freedom to place your object and its shadows into another image.
here's an image to give you an idea. the ball was sitting on a plane that had a matte/shadow shader applied to it. thus, the shadow has been rendered as an object with transparency, or alpha, properties intact, and the rest of the ground plane doesn't exist.
a masterpiece, i know [*bows*]...
hope this helps. i know it sounds a little complicated, but it's really not bad at all.
Block this user
Are you sure you want to block this user and hide all related comments throughout the site?
Archinect
This is your first comment on Archinect. Your comment will be visible once approved.