Conceptual Architectural Visualisation

Example of a conceptual image, used in this tutorial

Introduction

A conceptual representation of a project. Project: S. Verhage

A more definitive representation of the same project (note the differences in the amount of detail). Project: S. Verhage

This tutorial serves as an introduction on how to use multiple computer programs to quickly create an conceptual image of your architectural project. In this case we will use Autodesk Maya, Adobe Photoshop (or Gimp) and Illustrator (or Inkscape), whereas each of these programs has a specific quality, that serves the quick production of an image. This tutorial will cover workflow and communication between the different programs and the specific techniques per program.

Although this tutorial is mainly focused on using Adobe Illustrator and Adobe Photoshop, the basic principles are the same for Inkscape and Gimp. Because some options are a bit different within Inkscape and Gimp, an extra part has been added at the end of each section to explain how the basic functions work within Inkscape and Gimp.

Often it is not necessary in the beginning of a design process to spend a lot of time on visualization, as the project will change regularly. Also the produced image should represent this state of "indefiniteness". To create images that articulate a certain degree of finalization makes no sense, as a teacher, group members or client might think the project is already finished.

The conceptual image has a specific place in the design process, it is less suitable for communicating the finished design, but works better in a situation where the represented unfinished project is in a stage where it needs some form of visualisation, but is not yet ready. Its name is therefore very descriptive, basically its purpose is to show the concept of a project.

Style, Detail and Communication

Visualizing a project can serve two purposes, one is to communicate the visual appearance of materials, color, space, thus the design, to a client, teacher or groupmember. The other purpose is to serve as an evaluation tool, where different design ideas can be evaluated. In this tutorial both are integrated into one project.

Therefore, based on the images' purpose, one must consider:

• The amount of details (strongly related to the temporary state of the project). Think of trees, cars people, materials, furniture. Are these made in 3D or added later on in Photoshop or Illustrator? Are the realistic or stylized/abstract. Too much detail will give the impression the project is already finished.
• The topic of the image. What are you communicating? The concept, the architectural space, the colors, the materials, infrastructure, the use? Based on your purpose of the image you can add more information to it.
• The style of the image. Also artistically, how are you representing your project, cartoonlike, grungy, sketchy, realistic, abstract, warm, sterile etc. The style of the image can help you getting your message across.

Considering these topics is relevant for the conceptual phase of the project and what you are trying to communicate with it. An abundance of information may lead to drawing away the attention to something other than you want people to focus on. For example, in an image you want to emphasize the shape of a building, but by adding photo-realistic trees to the image, the building itself will look meager.

With everything you do when creating your visualisation, always ask yourself, what am I trying to communicate?

Project description

For our project we want to quickly create, within 3 hours, a fast visualisation of our project. Because modelling rendering, setting up light, creating and mapping textures in Maya, can consume quite some time, we will not use these techniques. Besides, we are aiming at an conceptual and not a final representation of our image. Therefore we are only going to use Maya as an underlayer for our image. After which we are going to use Illustrator to add some color and people silhouettes, and finally we will use Photoshop to create materials, shadows and light.

As we are still in the conceptual phase, a lot of emphasis will be put onto the communication between those programs, as workflow is very important to achieve fast result.

The construction of our project, with the phases Maya, Illustrator and Photoshop

Our project is not finalized, but we would still like to communicate the design. Therefore we have decided on the following:

• The space is an public service desk. The image should represent this
• The image should focus on the architectural expression and usage of the space (not on the details, furniture etc.)
• It should be visible that the roof lights illuminate the room
• The image should focus on the interior of the building, not the exterior
• It is not yet clear what kind of colors and materials will be used, flexibility in colors is necessary
• Finally, it should be clear that this is a preliminary design, and not the definitive project

It is very convenient for yourself to make such a list beforehand, so you have something to guide you while you're working on the project.

The Start, Creating the Image Underlayer with a 3D software

Massmodel

Because our project finds itself in the conceptual phase, we will keep our 3D model relatively simple, at massmodel level.

You might consider the model to have a level of detail which compares to a 1:500 CAD drawing of a building. This means:

• The rough shape of the building is modeled, without making installations, rails, details etc.
• Openings in facades are present, dooropenings, windows, atriums. These openings however, do not have framing, glass and doors in them.
• Some essential simple furniture is modeled very quickly, solely to emphasize the use of the space. This should not take more than 5 minutes

We want to get a line drawing of our model. Preferably a vector drawing, because we will be able to easily modify this in Illustrator or Inkscape. To create the vector image of the massmodel we can use a number of different 3D software. In this tutorial we will discuss the following three:

• Print to pdf in Revit
• Make 2D in Rhino
• Vectorrender in Maya

Print to pdf in Revit

After we created the model, we create a 3d view (Revit Views#3D Views).

1. Make sure the shading is set to Hidden Line (see image above). If we choose a different setting here, we might loose the vector information and get a pixel image.
2. File » Print » Print
3. Choose a pdf-printer.
4. Choose a path and a file name to store the pdf file.
5. Click Save

Make 2D in Rhino

After we created the model and placed the camera in Rhino, we make a 2D view of the geometry.

1. Dimension » Make 2-D Drawing
2. select the objects you want to see in the 2D drawing.
3. The following window will open:

1. Leave the settings as in the above image and click OK.
2. A new layer called "Make2D visible lines" contains the lines of the 3D-view.
3. Select these lines and click File » Export Selected...
4. Set Save as type to Adobe Illustrator (*.ai).
5. Choose a path and a file name to store the Illustrator file.
6. Click Save

Vector render within Maya

After we have created the model and correctly placed our camera in Maya (for more information see the Cameras page), we can make a vector render. The vector render render engine will produce simple black lines, which we can then use as our underlay for the rest of our visualization.

Since we want to make full use of Illustrator's vector editing capabilities, we want to render directly to an .AI file, which is a vector based native Illustrator file. If we were to render to JPG, TIF, TGA or any other pixel format, we would not have the flexibility in colors adjustments, line thicknesses etc. that vector files have. See also: information on pixels versus vectors.

When rendering with vector render, check the following:

• Your model uses only Maya base shaders: Lambert for diffuse and Blinn or Phong for reflective materials. Replace any other shader, such as MIA materials with Lamberts.
• That the render engine is set to Vector render in the render settings
• In the Common tab, for Image format make sure you check Adobe Illustrator (ai)
• In the Maya Vector tab make sure that Fill Objects is disabled (as we only want lines)
• Make sure Include Edges is on
• Make sure you have set your project (as the .ai file will be placed inside your project)

If you've selected the output to be Adobe Illustrator (.ai), but Maya doesn't write and AI file, but instead gives a warning and produces a IFF file, you're probably using a Student version of Maya, which doesn't allow you to save vector formats. Use the educational version available to students of the faculty of Architecture instead.

For more information, see Maya Vector Render to Illustrator.

Illustrator, adding Color and Silhouettes

Once we open vector image in Illustrator we can start changing the appearance of the image. Some things that you could do in Illustrator are:

• Add/change color
• Add scale figures like trees, cars and people. You can even use your favorite CAD blocks or trace pixel images
• Change the appearance of lines, thickness, dashed, color, transparency
• Add text (which is always sharp)
• Draw new lines, which then can be altered easily

Adding color with the Live Paint Tool

Adding color with the live paint tool

Now we want to use Illustrator to assign colors. Since Illustrator CS2 Illustrator has a Paintbucket-like color fill tool, called the Live Paint Tool which works quite handy. The way to go about is:

1. Select the entire drawing
2. Select the Live Paintbucket tool from the toolbox
3. Click once on the selected drawing
4. Deselect your image
5. Select a fill color of your preference
6. With the live paint tool you can now fill the area's with the fill color you have selected

If you have difficulties creating a Live Paint Group, refer to Live Paint troubleshooting

The difference in depth between mono-color and color differentiation

When you fill area's with color you can already take into account which area's are lit more than others, because of the daylight, artificial light, or materials. Also by using different color tones you can create depth in the image. As is shown in the example.

Another handy feature of the live paint tool in a context of easily changing colors, it the triple-clicking. When you click three times in a speedy manner, all the area's that have that similar color will change to your new fill color. With this you can switch the colors really fast inside your image, evaluating different options (as opposed to individually fill every single area with a new color).

Adding scale figures

Adding scale figures gives a sense of scale to your scene. Also they tell something about the use of your design and its surroundings. Therefore it is recommended to not randomly paste stuff into your image, but consider what you want to emphasize with them. In our case we have a public service area, with a waiting area and with a service desk, thus using a lot of sitting people seems to make the most sense, as opposed to people doing irrelevant activities like running or skateboarding.

There are several ways for adding people figures as scale elements to your image:

• Drawing them yourself (with a pencil or with Illustrator/Photoshop)
• Actual photo's (best composed into your image in Photoshop)
• Use pre-made vector drawings (from DWG blocks etc.)
• Trace photo's with Illustrator

Tracing pixel images in Illustrator

Illustrator, which is a vector editing program, has numerous advantages if you want to be more flexible in changing the appearance of the artwork. Therefore it would be handy to have your scale figures for example in vector format. The process of converting a pixelimage (photo, drawing etc.) to a vector image is called tracing. Illustrator has such an option, but there are certain aspects you need to take into account:

• The subject that you want to trace has to be distinguishable from its background. A white background is thus preferable
• A lot of color differentiation will lead to more detail in the vector image. If you merely want a silhouette or an outline, make sure the pixel image is simplified, or completely mono-colored
• Black and white or greyscale images produce the best "abstracted" vector result
• There are several tracing options, it would go too far to discuss them all. But the "Threshold" option is very influential on how the pixel image is read. Play around with the settings and see what the result is

Using the live trace option to convert pixel images to vectors

The LiveTrace process in Illustrator

• Open a prepared (preferable black and white, clear background) pixel image in Illustrator simply by using File > Open
• Select the image
• Object > Live Trace > Tracing Options
• Check the preview option in the options menu
• Adjust the settings so the result complies with your wishes. Hit the Trace button
• You have just created a live trace group, which we will convert to editable artwork, by going to Object > Live Trace > Expand
• Now rightclick the artwork and Ungroup it
• You can also release the Compound path (also by right clicking and selecting Release compound path). This will explode all the art work in separate small pieces
• Now you can change the appearance (fill and stroke color, transparency etc.) of your traced artwork

You can now put the scale figures in your Conceptual Architectural Visualisation image. Taking into account the laws of perspective. It is recommended to put your people scale figures in one layer.

Scale figures and their appearance

Using Inkscape instead of Illustrator

Using Inkscape instead of Illustrator

Inkscape is a free 'Vector Graphics' software available from inkscape.org. Although Inkscape isn't as sophisticated as Illustrator, most of the options we use for the conceptual architectural visualization can be found in Inkscape as well. In this tutorial a few basic options are explained. The Adobe Illustrator file (ai) that is created by Maya Vector Render can be opened in Inkscape without any problems. More information can be found in the Inkscape user manual

Explode the drawing

'Exploding' the drawing

Line width and fill color can be adjusted just like in Illustrator. However, if you want to adjust the width of individual lines, you need to take a few extra steps. First of all we need to 'explode' the drawing. We do this by using the 'Edit paths by nodes' button, to be able to select the edit points of each line and break the line at each of these edit points. To finalize this step we go to:

Path » Break Apart

We can now select every individual line and adjust it's appearance.

Adjust line width

Changing line width

To adjust the line width we select the line(s) we want to adjust and we go to:

Object » Fill and Stroke

And in the 'Fill and Stroke' menu we can adjust the line width.

Add color

Adding color

To color the areas between the lines we select a color in the bottom of the screen and go to the 'Fill bounded areas' button. This function isn't as precise as the 'Live Paint Bucket' in Illustrator, but we have some option to make it look good enough. To find out more about these options go to the corresponding Inkscape help page.

Photoshop, adding material, light and shadows

The image we have now in Illustrator is fairly clean and sterile, we now want to add some atmosphere by using light, material and shadows. To express the architectural concept without getting too much into detail.

Photoshop (or Gimp) is the preferred program to do so, as it offers a certain freedom in drawing shadows and light. In the following steps we will discuss setting up the photoshop part, how to create shadows and light, and how to add materials.

Setting up the Photoshop part

Copy paste the vectors from Illustrator to Photoshop

If we want to create an image that we later on can use again in our Illustrator file, we need to make sure the images are aligned, in other words: making sure everything fits together when we combine the different files. The easiest method would be to make sure that in both Illustrator and Photoshop you are using the same paper size, A4 for example.

Now simply copy and paste the rendered line drawing from the Illustrator file to your Photoshop file.

Layer structure

Creating a layer group

The use of layers in Photoshop seriously increases the workflow, the ability to change things, and dis/enable certain parts. Therefore we want to have our background linedrawing function as a background only. Thus lock the layer by using the small lock icon the the layer menu (see screenshot).

Also, when working in Photoshop, the large amount of layers can be really inefficient. Therefore, make sure you're using layer names (double click a layer) and layer groups (folder icon in the layer menu), these really help to structure your Photoshop file. Have a look at the screenshot to get an impression of a possible layer structure, where you can see the layers are placed together in groups, that have names so they are easily distinguishable from each other.

Using layer groups and name to structure the file

When using layers in a structured manner make sure you are working in seperate layers. For example, if you want to draw shadows in the image, do that in in a different layer, not in the background one (which will not work anyway because its locked).

A lot of Photoshop tools have the option to "sample all layers" and not just the one you have selected. Its advantage is that you can use the tool in your new empty layer, but it will take into account the other layers. In the shadow example, you can use the "Magic wand tool" to select certain parts of the locked background layer, and start drawing in the new empty layer. This method is very important if you want to work "non destructive", which means that you only add information (by using new layers), and do not remove or overwrite any. This will help you to more easily make changes later on.

Sample All Layers

Adding materials

Applying textures, before and after

Next we will add materials to our scene. The workflow is as followed:

• Use a (selfmade) texture that covers a large physical area. A texture that is 2 by 3 bricks stretched over 30 meter walls look ridiculous
• Duplicate the texture's layer, hide the original
• Hit CTRL+T in the new layer. This will give you the transformation tool
• Hold CTRL and drag any of the corner points. You can easily distort the texture to fit the perspective and the surface you want to attach it to
• Take into account the scale of the texture and the laws of perspective
• Remove the parts that you don't need, by using the magic wand tool and the sample all layers option to select the parts where the texture should be hidden. Use a clipping mask on the layer to do this (see below).

Creating soft shadows

Drawing the first shadows

Adding textures to the emtpy scene increases the desired effect, which is to communicate the design. But it still looks somewhat flat. Therefore we will add some soft shadows. These shadows are not cast from the direct sunlight, but are a result of the diffuse light that is all around. These soft shadows give depth to corners and the image overall.

In 3D this is often referred to as Ambient Occlusion or a dirt pass. In this situation we will discuss making one ourselves with the use of Photoshop, as this method is not related to any 3D program, as we will draw in a image that not necessarily has to come from for example Maya. However it is possible to let Maya render an Ambient Occlusion pass, that offers similar result, and does not require any "handdrawing", information on this topic can be found in the Mental Ray Ambient Occlusion tutorial.

The first step is:

• Create a new layer for your shadows, name it, put it in a new shadow group
• Grab a soft brush (hardness 0), and drop the opacity to 10-30 percent
• Make sure you're drawing with a black color
• Draw thick fat shadows in the corners, don't worry about them being too much
• Don't forget the 'corners between floor and furniture

Masking

Drawing the first shadows

Note the thin rectangle around the active layer or mask

Obviously you will not get it right the first time, so corrections are necessary. But we are not going to use the eraser tool, as that will permanently erase our information. Instead of that we will use the masking function. As shown in the simple example, a mask is a greyscale image that is attached to a layer. This greyscale image controls what parts of the layer are visible, and which parts are not. The black mask parts are made invisible, the white parts will be visible, by using greytones you can make something partially visible.

The big advantage of masking is that it leaves our layer intact. So all the information is still in there, it is just made invisible. This method is preferred over the eraser tool because by simply using a white color in the mask you can make parts visible again.

To make a mask, select a layer and click the mask icon in the layermenu. Select either a grey, white or black brush, and make sure the mask (and not the layer) is active, you can check around which one there is a thin rectangle. If the mask is not active, simply click it, and you can start drawing. You will not see the black/white/grey in your view, but the result of you drawing the mask on the layer, it is just like erasing, but with a brush instead of the eraser tool.

So in the render we now want to make the shadows less fat. Create a mask for your shadow layer, and start drawing black in it. If you are using a soft transparent brush (10% opacity) you will have the best results. If you feel that you removed too much shadow, simply change the color of your brush to white, and draw in the mask to bring back parts of the shadow.

This method for creating the soft shadows is fairly simple, but can be quite a lot of work, as you will constantly have to remove and add shadows, to see if the result is what you want.

The result

Creating light and color correction

Now we have upgraded our render with textures and shadows. But we really want to emphasize the light and the ambiance in the space. Again Photoshop to the rescue, we will correct the color and add some light rays coming from the rooflights.

Using exposure

The image we have now is quite evenly lit. But some should be a little darker than others because they catch less light coming from the roof. So what we want to do is make the area's that are further away from the middle a little bit darker, nothing extreme, but just to highlight the middle area more, we call this localization of an effect.

Underexposure, normal, overexposure

Normal, underexposed, localized underexposure (sides)

For this we are going to use the exposure function from photoshop. And keeping in mind that we want to work non-destructive, we are going to use so called adjustment layers.

With the exposure function you can make an image lighter or darker, as would increasing the exposure time on a camera. Compared to the levels, curves and brightness functions the exposure has less effect on the contrast, it just makes the image brighter or darker.

In the layer menu click the icon for the adjustment layers. These are seperate layers that can be changed afterwards (unlike the functions used via Image > Adjustments > ...). Also when using adjustment layers photoshop automatically attaches a mask to the new adjustment layer, letting you localize the effect of the adjustment layer (see the mask section on this page).

Now for our project; Simply make 2 new exposure adjustment layers, one under- and one overexposed. By drawing with a soft transparent brush (10% opacity) in the overexposure layer one can make light rays (the whole mask is black, but the rays, which are white, thus overexposed). And in the underexposed layer we can make the sides of the image darker by using the same brush (the middle of the mask will be black, thus normal, the sides white, thus underexposed). See the images for clarification.

Begin, general overexposure, localized overexposure

Color corrections

As a final step,one could say that our image lacks a bit of warmth. In the previously mentioned adjustment layers you can find the so called photo filters. These can make an image cooler or warmer, select one of your preference.

Photofilters; warmer, normal, cooler

Using Gimp instead of Photoshop

Gimp is a free 'Photo Editing' software available from gimp.org. Although Gimp isn't as sophisticated as Photoshop, most of the options we use for the conceptual architectural visualization can be found in Gimp as well. In this tutorial a few basic options are explained. More information can be found in the Gimp user manual

Distort textures

Distort textures

To create the effect of a perspective view we use the 'Perspective tool'. With this tool the corners can be grabbed and dragged in such a way that a perspective illusion is created.

Distort textures

Brushes

Gimp also has a brush which can be used to create the 'ambient occlusion'. After clicking on the brush tool, the brush options show. Here we can change the opacity and the scale (only till 10.0). But we can't change the hardness of the brush. To change the hardness we actually need to change the brush type to a softer brush.

Layer masks

To make a layer mask in Gimp we need to select the layer we want to add the mask to and go to:

Layer » Mask » Add Layer Mask

A window pops up with several options. Leaving it on 'White (full opacity)' is usually good. We can now draw with black to hide things within the layer.

Adjustment layers

Adjustment layers don't exist in Gimp. Therefore we have to find a work around. In this tutorial we either want to make parts of the image darker or lighter. What we can do is create a new layer above all the other layers, for instance a white layer to make things look lighter, and use different 'modes' and 'opacity' on that layer to change the appearance of the layers underneath. A bit of trial and error should get you close to what you want. Off course layer masks will still work to localize the effect.

Photo filters

Photo filters need the same approach as adjustment layers. To make your image appear warmer, add a layer with a warm color above all the other layers and play with the 'opacity' and 'mode' of the layer.

Finalization

Now we have completed our image in Photoshop (or Gimp), it is time to combine them in Illustrator (or Inkscape).

From Photoshop to Illustrator

Do not just copy/paste or open your newly created photoshop image directly into your Illustrator file. If you do this, it will be harder to make any adjustments to the file. Instead of opening it directly, you should link the file. Which means the file is not imported into Illustrator, but merely loaded. If we change the original file, it gets updated in Illustrator. For more information about linking, see Advanced Illustrator.

Now we have our image in Illustrator, we need to think of our layer structure, and the how to use our image with the colored line render. The recommended order is putting the colored line render in the back, the photoshop image in the middle and the scale figures (people) on top.

Creating variations

This layer structure offers us the possibility to try different ways of mixing the photoshop image with the colored line render. If you select the image, and open the transparency menu (next to the stroke menu, or via Window > Transparency), we can use the opacity to mix it, or we can use different blending modes, of which multiply is a very powerful one. Depending on your wishes, a opacity of 75% and multiply blending can give the best results.

Mixing the Photoshop image with the colored line render

Because what we can now do is lock the image layer, and use the paint bucket tool to change the colored line drawing that is in the back (remember the triple clicking shortcut). Within a mere matter of seconds we can create numerous color variations for our Conceptual Architectural Visualisation. And if you want to change the photoshop image, just open it, alter it, save it, and it will automatically be updated in your Illustrator file.

Color variations

Color variations

A quick adjustment of the original image results a in totally different space

This concludes the tutorial. And as a last note, using Photoshop and Illustrator isn't exact science, intuitive 'accidents' can result in the best and most interesting images, do not be afraid to play around with settings to see what they do.