Planning before Executing a Project

[Dave Buckley]

I'm interested to see the different approaches to planning for an arch viz project.

 

More specifically the modelling part of things.

 

Let's say you have a building that requires each floor level, room, interior and exterior to be modelled textured and rendered.

 

How do you attack it?

 

At some point you must sit down before opening your chosen software and decide how best to approach it.

 

For example, do you break the building down into areas, create lists for what needs going in that area and then model in any specific order from that list?

 

What do you do once that area is finished? Do you move onto your next area, stop again, create a list fro what goes into this area and model again?

 

Or do you texture the first area before modeling the next and then shut it off?

 

I think you get the idea. How do you keep things organised and not waste time?

 

Interior modeling first then exterior modeling? Objects that require high resources (foliage etc) last. etc etc?

 

Interested to hear your thoughts.

[frank1331]

Sounds like a major project. Can't say I've ever had a project where I had to model and texture everything. The first question would be, how far along is the building design? If the form of the building is established, I would propably start with modeling of the exterior. I think that is the least likely to have major changes. Textures and colors may change, but the modeling should stay the same for the most part. I would leave all of the texturing for the end. Those are the things that are most likely to change during the design. I think you absolutely must have some sort of checklist inorder to keep this thing organized.

 

Good lUck with this one.

[Dave Buckley]

Ok, so it's not a project I'm working on. I'm just interested to hear peoples approaches to planning for a project in order to keep things organized.

 

I often look at the end results that people post online and then try and figure out how I think they could have got to that result in the most efficient way.

 

Especially on detailed renders, I think to myself - "Where did they even start with this". There are so many aspects to think about what came first. Even small renders.

 

I'm just looking into how much planning people do for projects large or small. I feel I waste a lot time by not being organised.

 

Lets take a garden shot looking back into a house for example. Usually you can see the interior fully textured and modelled as well as the exterior.

 

What order did they approach the modeling in in order to keep it efficient and organised.

 

It's just a general thread that I'm hoping becomes useful for a few more people.

 

it could be a personal piece with no set deliverable

 

but how do you plan how best to create it (you may not plan, and you may just make it up as you go along)

Edited October 11, 2010 by Dave Buckley

[carlotristan3d]

I've observed that most of those detailed renders posted online (where every blade of grass, etc are 3d) are personal projects. Hence, the artist are free from deadlines and where able to continously improve the renders to their satisfaction. I've experimented with one, with the intent of doing an animation later and what I've learned is to get everything on their proper layers. Modelled the exterior first and then populate the environment with proxies, modelled the interior in a separate file and bring in as proxy.then I can play the lighting and test renders by turning layers on and off.

[Dave Buckley]

Ok so you got your ordering straight, exterior, environment, interior.

 

How do you plan what trees to use, what furniture is going to be inside etc.

 

You must plan what to model?

 

Collate reference material of ideas, or create lists of specifics etc

[carlotristan3d]

furniture is not a problem, if you're doing it for an architect you can't get wrong with modern classics, and these are plentiful on the net. and these would be the last thing i would even model. Trees? I'm more concerned which trees that are nearest to the camera, because you have to show detail. In this instance, Evermotion is your friend or If you have time to texture- Onyxtrees. I don't mean to plug Evermotion, but if you can find the trees you want in their catalogue, buy it. It'll save hours plus you can reuse it over and over again.

[Dave Buckley]

nice to see your thoughts Tristan, but I'm not necessarily looking for an answer or solution to a specific project.

 

I'm just looking at the many different approaches to visualisation.

 

I'm not looking for an answer to 'What should I do?'

 

it's more a case of 'What do you do?'

 

And again i'm talking specifically about planning before even executing the project.

[Tommy L]

The way I approach projects depends entirely on deadline and fee....

Modeling obviously comes first but I don't think it matters where you start modeling, it all needs to be done. Internals and externals (on the same building) I usually render from separate files. If I have the exterior modeled it can occasionally be a starting point for modeling the interior but not always.

[Dave Buckley]

I agree but before and during your modelling stage do you have a system in place to keep track of everything that needs to be modeled, what has been modeled, what remains to be modelled?

 

I've started sketching over draft renders of the space once I've established my viewpoints to help streamline and stop me sitting there pondering what to model next.

 

I used to be a 'list' man

[Tommy L]

Im very much a list man.

[Crazy Homeless Guy]

EDIT: Sorry, I outlined the rough process I take for a project, and not the planning before hand. I reduced the post to the start of the project..

 

_____________

 

I would say more than writing anything down I quickly form an outline in my head of what I want the final project to be, and what I want it to look like. Photo impression vs. NPR, 3d trees vs. 2d trees, what type of people, daytime vs. nighttime, dreamy vs crisp, etc.. This initial mental picture will provide the road map for how the project is broken down, and what working methods are needed to get there.

 

At this point I don’t create a list, rather I start files that represent views, lighting, and model pieces. Those might break down as follows... They more or less provide a road map that chunks the project into what needs to be done. Similar to a list I guess.

 

- Lighting and Views.max

 

- Architecture.max

- Site.max

- Lobbby.max

- Café.max

- Entourage.max

- Transportation.max

etc...

 

I usually start with the lighting and the views, and let those be a large part of the guidelines for the rest for the project, with the model pieces filling in the rest.

 

Overall any adjustments or surprises are usually found by this point in the project, so if needed I go back and adjust the plans for the workflow.

 

I am a person who likes things to be hands on, so I learn and decide by going ahead and diving in and let experience fill in the details of what needs to be accomplished. I do have lists, but they come later in the project, keeping track of things that still need to be done.

Edited October 12, 2010 by Crazy Homeless Guy

[Dave Buckley]

Nice work Travis

 

I saw the full post in my email notification which was quite an insight but thanks for noticing the focus of the question.

 

Do you have outlines for your working methods that get you to your various end results?

 

For example, if you or the client decide it's a concept work with NPR style, do you work from anything different than you would for a photoreal output?

 

"These are the steps I need to go through for a NPR image, this is where the Photoreal work differs . . . " kind of thing, obviously throwing in the artistic license and employing fresh techniques whereever you feel fit.

 

I guess I mean loose guidelines for different outputs.

[notamondayfan]

Where exactly do you think you are spending too much time?

[Dave Buckley]

Where exactly do you think you are spending too much time?

 

Personally for me it was 'thinking' about what to do next. Spend too much time thinking and not doing.

[BVI]

It also depends how many people are going to work on the project. If you have a team, then the process can be production lined, with multiple activities happening at once.

[notamondayfan]

Personally for me it was 'thinking' about what to do next. Spend too much time thinking and not doing.

 

Maybe you need to break down your process into bite size stages. Are you thinking about props, light, camera, materials, effects, etc all in one go?

 

What are your clients generally like, are they decisive or really loose and leave the decisions to you?

[Dave Buckley]

Maybe you need to break down your process into bite size stages. Are you thinking about props, light, camera, materials, effects, etc all in one go?

 

This is what I do and also used to do (1st and 2nd sentence respectively). I'm just interested in seeing how other people approach it.

[notamondayfan]

I generally dont worry about the finer details in a project until the initial stages have been signed off. This means I only model, light, texture, etc exactly what I see. Unless the brief is massive with loads of images or animation, you will probably only see 10% of the building.

[Dave Buckley]

I generally dont worry about the finer details in a project until the initial stages have been signed off. This means I only model, light, texture, etc exactly what I see. Unless the brief is massive with loads of images or animation, you will probably only see 10% of the building.

 

Do you have a process of what you model first similar to Travis? for example, he goes large to small?

 

Let's say for example you are doing a single shot of a kitchen. I would imagine something along the lines of, walls, windows & doors (if seen in shot), fixtures/fittings/furniture (cupboards, benches, seating), large appliances, small appliances.

 

What do you class as 'fine details'

[pand0r]

Establishing the camera angles is one of the first things I try to do, since this has a large impact on what can be seen (and has to be modelled).

 

I will be following this thread closely, since this is something we should be thinking about at our company too, instead of modelling without a decent planning.

Edited October 12, 2010 by pand0r

[Dave Buckley]

Establishing the camera angles is one of the first things I try to do, since this has a large impact on what can be seen (and has to be modelled).

 

I will be following this thread closely, since this is something we should be thinking about at our company too, instead of modelling without a decent planning.

 

exactly my point, thats what I was hoping the thread would become, something where people start to evaluate their workflows and hopefully benefit from the thread in some way.

 

what promted it?

 

watching the Julius Shulman film and hearing him say "I don't pick up the camera until I'm certain of what I want to capture - you must plan ahead"

[pand0r]

what promted it?

 

watching the Julius Shulman film and hearing him say "I don't pick up the camera until I'm certain of what I want to capture - you must plan ahead"

 

Hehe no, seeing the hours accumulate and the profits melt away was a less enjoyable way of finding out we needed a decent plan before starting.

[Crazy Homeless Guy]

"These are the steps I need to go through for a NPR image, this is where the Photoreal work differs . . . " kind of thing, obviously throwing in the artistic license and employing fresh techniques whereever you feel fit.

 

 

My initial stages are the same for the most part regardless of what type of piece it will be. However, if it is going to be an NPR I am more likely to use 3d people, or 3d trees. Not always, but more likely to. It is at the start when I am developing a picture of the final product in my head that I decide whether or not I want 3d entourage or 2d entourage for the final product.

 

The other thing about NPR work for me is that I become a great deal less concerned about details, and accuracy of materials. They can be fairly loose an accurate and still produce a nice NPR piece. However, some of the time savings for those pieces gets added back into th post work on an NPR piece. Not all of it, but some of it.

[Crazy Homeless Guy]

e

what promted it?

 

watching the Julius Shulman film and hearing him say "I don't pick up the camera until I'm certain of what I want to capture - you must plan ahead"

 

Camera views are one of the first things I try to get off the plate. I find it very irritating when camera views change late into the game. To me, changing the camera view is a great deal worse than changing design geometry at the last second. I feel it takes away from what you have been working towards, and leads to a less successful project.

 

If I am working with a designer who is flaunt in 3d I often ask them to go ahead and set cameras for me. I may refine them a bit, but I find it saves a great deal of time if they just set them, and then give them to me rather than trying to explain what it is we need to capture. The old "A picture is worth a thousand words." I can see the idea better if you show me through the lens than if you try and explain it to me and hope that the picture in my head is similar to the picture in your head.

 

This is also successful because the designer has often been working with the model for several weeks, or in some cases the design for several months, sometimes even over a year. They really know the design inside and out, and the understand the what the nuances of the design are. Things that I may not pick up on in just 1 or two hours of working with a model.

 

I think illustrators want to believe they can pick a 3d view better than the designer because that is part of the job, but the designer holds very valuable information and a keen eye. Being able to have the designer assist with this portion can knock off 25% of the time it takes to complete the project. If they pick the initial view, there is a great deal of a less chance the initial view will be changed, and it saves a lot for time going back and forth, and shooting test shots.

[Justin Hunt]

I used to get really tied up in locking down the camera view very early. However I dont get so pedantic any more. Mainly because more often than not we have to start modeling long before a design is complete (and in some cases even thought about). One view developes into an other and then another so before you know it there are ten views and we are basically seeing everything at one point or another.

 

So now we start massing out the main/ larger elements and go through a process of refining and adding detail as needed. Like Travis some projects run for years (I think the longest is going on four years now). The process is far more organic and fluid. As such our planning has to be more fluid.

 

One thing that I feel strongly about is organising you modelling naming conventions and project filing systems in a logical way. Keeping track of revisions, options and variations can be a nightmare and in the long run cost you alot of time and pain if you let it get out of control. Especially if your working in a team environment.

 

jhv