SU in the design process: the matter of “Workflow”.
-
Don't get me wrong. I didn't say I had given up CAD altogether.
All I have done is rationalise so that I spend less time undoing, or re-doing things already drawn in SU, and lots of things are easier drawn in SU anyway - especially using pre-prepared components
One of the main things in CAD is dimensioning so I do that in CAD, as well as assembling all the views and the detail work - very much as as Greg does in his workflow.
I have stopped using layers though, although you could easily set these up in a template.David
-
gf.greg
here is the finalized version of the piece from the gallery thread. (it might be cropped a little funny...)
i'll try and get some images of the model up.
re: photomatch - for a lot of environment design concept artists get pushed into the 'look/feel/atmosphere' department. Many times i get 3d underlays from the AD to paint over or at least a generalized block out of where they want certain elements to be. I've gotten into the habit of taking their block out and photomatching to it and building/painting on top. It makes the photorealism easier and makes sure you don't get too far away from what your AD wants. I eventually started doing this with my own compositional sketches. i somehow feel it retains the initial artistic spark and keeps the traditional media in the workflow. I always let the sketch dictate the model and not the other way around. This means (to me) that i'm not letting the software dictate the design. In a way is a bit of the opposite of your painter sketches.
I really admire the amount of detail in the planning you have to go through! Your stuff is extremely well thought out and planned and the resulting compositions are both pleasing and useful. I think this is a fundamental difference between what an AD/PD does and production illustration. You don't get the luxury of saying 'someone else will figure out the details!' - which is something you find amoung concept guys. They will love to do the really cool painting, but are never quite sure exactly what the form is doing, or what is behind those trees. I think sketchup is a tool I can use to make that gap smaller. AD's love it when you have a model, to scale, that you can rotate around and show exactly what you are thinking.
But of course they always want changes
.the trees in the above image are all photoshop. I put in proxy geometry in my rough model, then dropped in photos later on.
let me ask you, as an AD how could what the production illustrator does make your life easier down the line? are you only looking for how the set will be dressed, color, atmosphere etc or more comprehensive concepts of the environment at the outset?
-
@pbacot said:
Do your templates contain layers and and scenes that make up typical CAD drawings? So when you start a project, you already have a place to put groups, for example "second story walls"? This is what I do in CAD. I'll take a look at the link. I have read that thread (wish more installments were made).
in my last office I developed a layer system that can bring all models in the different partner offices to one standard. it has 4 different sections:
- CAD Drawing Imports
- Master Planning Scale
- Building Elements
- All Additions (furniture etc.)
of course for production design you had to add a 'Camera' layer besides the 'Lighting' layer in the fourth section or maybe even introduce a fifth section for all technical installations.
this layer system works with a clever grouping structure that gives you advanced control when modeling and presenting the design. I can write a short description of the grouping system, if you like.I don't know, if this is too technical a structure for production design. I myself tested it on several projects and it worked very well for me. I would love to hear your opinion on this system.
SU_Template.skp -
@dcauldwell said:
Greg
I only recently got fed up with preparing the design in SU only to end up re-drawing a lot of work in a CAD program (I was at least exporting the 2D stuff). I suddenly realised that with better preparation and a more disciplined approach in SU, then I could save myself a ton of work in CAD.
David
I am really glad to see that someone else got tired of the back and forth. I have just reached that plateau, so I appreciate your approach. I also appreciate the kind of cross-discipliness (newly coined phrase) that this thread entails, given that the workflow and might I say "deliverables" that one person in the process provides may be different in the set design arena versus architectural design, but there is a lot to be learned by seeing examples. Plus in the end the intention is to be able to visually express your ideas to others.
I would also like to add one other difficult situation I find. This is mostly based on the way our office functions, but we have drawings done in AutoCad, Vectorworks, and one contracted cad guy even uses Softplan, as well as the napkin sketches. To add to that the boss who uses AutoCad doesn't use it well. Cross platform exports are weak at the best of time, and the time spent "cleaning up" others drawings can be costly. -
Plot-paris
I personaly find this sort of layer structure too complex. I guess if you are sharing information and working on medium to large schemes it will be important to structure the information. I found that, when I worked in an office trying to impliment this type of formal structure, getting people to adhere to it was very difficult - so the system tended to fall apart at the seams. (Do some CAD systems force you to organise in this way?)
I took a lead from one of the SU workshop lectures and now use groups to the same end, keeping it simple - just, 'Ground floor', 'first floor', roof, etc.
I find that manipulating information using groups and the 'outliner' facility in SU, allows sufficient control, and is something that was not available in CAD.
However, I think the main thing is to find a system that works for you and there are many ways to organise the model information.David
-
In my experience as a Production Designer 'rigid' layer-type workflow systems don't work as a departmental framework. Film and Tv production is about budget, time and the ability to turn 180 at a moments notice. The going joke is - by the time it's finally drawn, it's already built, shot and binned. You spend days or weeks concepting a Spanish town and they decide to change it to an African village and still want the concepts tomorrow. It's the nature of the beast. I need to pull people off one thing and throw them into the African village - then it goes back to the Spanish town but the deadline is pulled up to tomorrow because the producer forgot he has to be in divorce court in two days.
You pull it all together and the presentation goes great... then the day before it all goes to camera the head of the studio's girlfriend (who is 19 and works as a personal trainer) comes to set and tells him she thinks it would look better painted pink-ish.
DE -
Plot, David:
I'll have to upgrade to SU 7 to see your file, Plot.
In CAD I use a simple layer system--usually only to hide or dim different layers in different sheets. Other than that I can keep it simple. So one sheet will have only 5 layers. A file can have 10 or more sheets, sharing some layers. Choosing a layer setup (called "sheets" in my program) gives you the output or view you need. I don't know how you can achieve this with groups or blocks in CAD. I would probably recreate these layers in SU eventually for those images or objects I would export to CAD. Not yet though.
New methods and applications arise constantly. So far in CAD, modeling, and other imaging, I think we have to keep fluid. Maybe you have to redo some things when you go to the next step. Pick up a tool and the faster it draws the better, and move on. Only highly disciplined large firms can institute elaborate systems, and they run the risk of tying themselves up while new technologies come out. You're going to have to mix it up sometimes.
Some CAD programs create certain layers and these serve in analysis for BIM etc. Not likely to be my cup of tea ever, but important to many in AEC.
Double shot: PD sounds like building design in many ways, especially in the stops and starts, twists and (arbitrary) turns. Sometimes just as tight a design / build. You may have to "document" things while they are being changed in the field. Usually it's no hurry until suddenly it's urgent.
Peter
-
@pbacot said:
Plot, David:
Double shot: PD sounds like building design in many ways, especially in the stops and starts, twists and (arbitrary) turns. Sometimes just as tight a design / build. You may have to "document" things while they are being changed in the field. Usually it's no hurry until suddenly it's urgent.
Peter
The biggest difference is I don't care if mine falls down in three months. lol
-
@unknownuser said:
I personaly find this sort of layer structure too complex. I guess if you are sharing information and working on medium to large schemes it will be important to structure the information. I found that, when I worked in an office trying to impliment this type of formal structure, getting people to adhere to it was very difficult - so the system tended to fall apart at the seams. (Do some CAD systems force you to organise in this way?)
Layers do slow you down. On most jobs I use this layering convention:
ground 2d (cad - everything component I make has a plan illustration drawn on the 0 plane with dimensions attached)
ground 3d
first floor 2d
first floor 3d
And so on. Then I export the 2d info back to the relevant cad layer.I know it's hardly rocket science. Even though it takes a long time to give every component a 2d illustration it saves so much cad time and if you use these components in other drawings it soon pays for itself. This allows me to mainly just use cad as a paperspace layout tool and spend as little time there as possible.
-
Gaganraj, thanks for posting the final version of your visual. I think it’s extremely interesting to see how you have integrated and mixed the different techniques and programs. The result is really excellent. Your use of the photomatch technique makes complete sense to me, and you prove that it is very effective. Good to mention that you care about keeping some of the traditional media in the flow. Also very important, I agree. Am still hoping you’ll be able to post a view of the SU photomatch job from different angle….By the way, do you ever use “Vue” for landscaping?
Also agree with “not letting the software dictate the design”, which I think is true for any tool we use anyway, specially with digital tools. This being said, SU is quite good for that in a way, as you said yourself: “ I find using sketchup, that i draw less and less traditionally, and move into sketchup earlier in the process. (i guess that's why its called sketchup...).” That’s one of the reasons why I really enjoy working with SU, and with only a little more practice I now feel the same with Modo.
Regarding the “fundamental difference” between a PD/ AD and an illustrator that you referred to, I agree but I would add that perhaps it depends on a few things, for instance which stage of the designing we are at. In my opinion, it’s OK even necessary to free all the crazy ideas at the beginning. Then gradually it becomes crucial to focus on the constraints of reality, and at this stage I guess that it is important for an illustrator to switch into “design” mode and think well before graphically expressing an idea. I think that make for a great illustrator. Not easy. But that’s also why we work in teams isn’t it?
In a way I think that it is better to try and avoid the “someone else will figure out the details”…If possible.
Again at this point I agree when you say: “…I think sketchup is a tool I can use to make that gap smaller…” and that’s also my answer to your question: “how could what the production illustrator does make your life easier down the line?”
Then, I believe that the Designer or the AD should make it clear whether they expect a visual that will translate the overall mood and feel of a story, or if they need something quite precise for a particular set.
I’m sure that there are many points described here which are similar with the Architectural Project. -
@double espresso said:
In my experience as a Production Designer 'rigid' layer-type workflow systems don't work as a departmental framework. Film and Tv production is about budget, time and the ability to turn 180 at a moments notice. The going joke is - by the time it's finally drawn, it's already built, shot and binned. You spend days or weeks concepting a Spanish town and they decide to change it to an African village and still want the concepts tomorrow. It's the nature of the beast. I need to pull people off one thing and throw them into the African village - then it goes back to the Spanish town but the deadline is pulled up to tomorrow because the producer forgot he has to be in divorce court in two days.
You pull it all together and the presentation goes great... then the day before it all goes to camera the head of the studio's girlfriend (who is 19 and works as a personal trainer) comes to set and tells him she thinks it would look better painted pink-ish.
DEwow, that really sounds like fun. isn't that one of the most horrible things? when some people, who have no idea about design somehow end up in a position where someone listens to what they say (pink spanish/african town) and destroy the work of months...
dcauldwell, you may be right in this model structure being too complicated. but on the other hand, it gives you a great deal of control. I attached an excerpt of a manual I wrote for the last office, where the layer system is explained. and I added the SU file in version 6.
model structure.pdf
SU_Template.skp -
@plot-paris said:
wow, that really sounds like fun. isn't that one of the most horrible things? when some people, who have no idea about design somehow end up in a position where someone listens to what they say (pink spanish/african town) and destroy the work of months...
What can I say... beats working in a bank.
DE -
This model structure is most helpful.
thanks for taking the time to explain it, diagram it, and clarify it.
aloha
red -
you are most welcome, red.
but now for something completely different (slightly off topic):
in film business (probably more in post-pro than in set-design) you can choose between three big programs - 3D Studio Max, Softimage and Maya.
I heard that whereas 3D Max is mainly used for computer game design, Softimage and Maya are the ones one should have a look at when considering to work in film. it is difficult to get an objective opinion on which of the two is better. Maya has the reputation of being the Queen among modelling- and animation software; but Softimage users are convinced that this queen has seen her best days already.however, a few months ago Softimage was bought by Autodesk (who already owns Maya and 3D Studio Max). they affirm their intent to still develop Softimage further - my question is for how long though...
so does anyone have some advice on which software to learn? do you have any experience with workflow and ease of use or any information of distribution of the software within the film business?
I would be happy to hear your opinions -
Hi plot-paris
You are right about the main tools used in film business (3DS Max, Maya, XSI) but it's not that simple. Everyone of them can do everything and in the end it will probably depend in wich firm you're working (if you're working as a freelancer normally it doesn't matter because you'll probably be asked to do a specific model and deliver it in a 3D format that anyone can open). Autodesk owning the 3 just means, right now, more compability between them.
3DS Max it's not the one "just for games", Metal Gear Solid 4 for example was done in XSI (but Hideo Kojima's firm is changing the workflow for Max too...) but XSI represents less that 10% of the market, Maya is still the most used for animations, and Max will always be strong because it's the standart of the industry and because the many plugins it as.
But you even have to keep in mind other players like Zbrush; that it's the King when it comes to adding detail, character and organic creation (with firms workflow like create the model in Zbrush export it to other and animate, or importing a mesh and addind details in it, or to make a hi-resolution model and use it to do the normal maps), Houdini 3D, for animations and SFX and then exporting to other renders; Modo, mostly because the modeling tools, and even Sketchup, for doing the (very) basic model concept before importing it other 3d apps or Photoshop for mate paintings.
So as i said in the beggining it's everything but a simple choice but if we're talking about animation Maya is probably still the best choice (the animation tools are stronger) and most in the business work or have worked with it.
P.S.: I'm no expert in 3D, i'm just a guy that "plays" a litle with 3D and likes to keep up with things in the 3D business, beeing fortuned for having good discussions about this with close friends that work in this area (TV, Advertising, 3D, Web, Programing...)
David
-
thanks, DacaD, for the detailed information. so there is no really waterproof solution of which is 'the' program to muster.
actually, I had a bit of a play with some of them (Maya and 3D Max). and while they are both very impressive when doing more complicated stuff, I was utterly disappointed by the awkwardness of their work flows in terms of very basic modelling.
as an example: you have two cubes and want to place one on top of the other with all the edges aligning.
the workflow in Maya (at least as far as I know):
- first I select one cube with the selection tool, then hit the 'Ins' button to change the centre point
- then with the move tool I move the centre point to one corner of the cube. then I press 'Ins' again to exit the 'centre point changing mode'
- I now move this cube to one of the top corners of the other cube (by pressing down 'V' it even snaps to the corner)
- now I right click it and change into 'face' mode
- I change to the select tool again and select one face
- now I switch to the move tool and move the face to align it to the cube below
- I repeat the last two steps twice to align the remaining two faces...
and the same with SketchUp:
- with the move tool I grab the first cube at the desired corner and place it on top of the other cube (it snaps to the corner automatically)
- I then select the push/pull tool, click one of the upper cube's side faces and click again onto the equivalent face of the lower cube to align it.
- I repeat the last step twice to align the remaining two faces...
see what I mean? it is sooo fast and easy with SketchUp!!! why cant the other applications adopt this simplicity of work flow? or why cant SketchUp become a proper professional, mighty Software?
-
plot-paris
There's not just one "correct" workflow when working with this aplications, anyone can do the same things using diferent methods, and because the main tools are about the same if you master one of the apps you can probabaly do just about the same things in the other two (as soon as you know where every tool is). And that's not the only way to put a cube in top of other. You can move the entire cube and align it using the axes putting the units, snaping to other primitive (i supose the cube is one of the primitives), and so on. Moving each face of the cube can't be the only way (imagine you're moving the mesh of dragon with 30.000 faces...you won't move face by face). But you can see what i mean.
The main diference between sketchup and the major 3D apps it's sketchup just working with faces and the others working with solids, nurbs and also faces. Other bigger diference, that for me is one of the things that is killing sketchup right now as a tool for 3D creatives, is beeing at the same time a presentation and modeling aplication. There's no render in sketchup, there's a optimized output, you present the model in the same space you use for modeling, "what you see is what you get" so the engine can't be fully optimized to just modeling or just rendering/output. In the other apps modeling and rendering it's completly diferent parts, what you see in the "model space" isn't what you're gonna get with the render.
So sketchup and Maya/3DS Max/XSI can't really be compared, and don't worry about beeing faster doing in sketchup simple things because that's an illusion if you're comparing a master in sketchup with a master in 3DS Max (or any of the others), and a master in 3DS probably means that he also masters photoshop (because he also makes is own textures, bumps, spectacular maps and so on...) so his final presentation will always be better than the one from sketchup...These software can also be very specifics, there's firms that have people specialized just in modeling, other people just for texturing, animation, or rendering. There's even a animation with maya class in England (for example) of 3 years with a one class just for texturing that takes 6 months...
The thing you have to ask is "Are you willing to have the patient, time, and dedication this kind of software needs to master it?"
David
Advertisement