It is therefore essential to eliminate all kinds of distractions and potental problems right at the beginning. That includes tuned workflow (e.g. Marking Menus, shelfs etc.) but also the planning process that precedes modeling. You should also correctly place your background images into Image Planes. And that´s what this tutorial is about.
If you are not somebody like Bay Raitt it would be wise to use some sort of reference – your best bets are images that can be placed in viewport and that will help you to define the volume, scale etc. In the case of mechanical (hard edge) objects it´s often a necessity to have them.
Sources of reference images
Unlike blueprints, reference photos, pictures and drawings can be gathered without any problem. Just use your camera, Google or purchase them on some kind of special website.
Background pictures are much harder to get.
If you want to use photographic references for. e.g. organic models you will soon find out that the images on the Internet aren´t good enough. They lack good lighting, image quality and are not high-res. Don´t even try to set them up as Image Planes because views from different angles won´t match correctly.
Shooting your own reference
If you want to take your own reference pictures, you should be aware that you are not allowed to photograph people without their permission. It´s against the law. There is one exception though – you can take pictures of celebrities without their permission.
Problem with perspective distortion
Perspective distortion is the effect that appears when a scene is photographed from a wide angle.
It means that the subject or the part of it that is nearest the camera is bigger in the picture than in the real world.
If you take pictures of, for e.g. a head, the nose will be bigger than the other parts because it´s closer to the camera and that´s something we don´t really need for precise modeling. Lenses with a long focal length have lesser perspective distortion. Generally speaking, the longer the focal length, the lesser the perspective distortion.
So how do we lengthen the focal length? Just step back from your photographed subject and use the optical zoom – the more you zoom the more you lengthen the focal length and actually eliminate the distortion. I suggest you use a tripod.
Lots of pictures on the Internet are perspectively distorted.
You will find blueprints useful especially for modeling (of) mechanical objects. They can be found on the Internet, sometimes also in books or you can even create your own. Once centered they become great background guides for modeling.
In the case of organic models you may consider drawing your own background images if you are at least slightly into drawing. You can base your drawing on photos if desired.
You can download some blueprints at our site, in the section MODEL REFERENCES
or from these websites:
I have chosen this one for example:
Preparation in bitmap editor
In the following tutorial I am using Adobe Photoshop but I believe that tools in other editors are very similar.
First we will check the measurements of our blueprint, as it’s not a good thing to find out that something doesn´t match in the middle of the modeling process.
I like the method where I cut out the individual views and place them all into one canvas (see picture). Of course I use the layers here.
Now let´s show the rulers (View > Rulers or Ctrl-R /Apple-R). By clicking on them and dragging them to the canvas you will get Guides. They will serve for measurement and comparison.
Set up the Guides in a similar way, as in the picture. Zoom in quite a bit in order to achieve precise results.
As you can see many things don´t match correctly. It will be necessary to tweak the turn and placement of the pictures using Transformation. Make sure you are in the right layer and go to Edit > Transform > Rotate.
Now you can rotate and move the picture using one rectangular manipulator. You can also write the numeric data in the Options bar.
Again it pays to zoom in because that way the tool works better. You can use this setup:
Create three new views using Window > Arrange > New > Window. Then arrange them (Window > Arrange > Tile). In the individual views, zoom into the key areas, then click on any window with the close up (it´s important to do the transformation in close up otherwise it´s not accurate and it will begin to snap to guides), go to the menu and use the transformation. Tweak the transformation using the manipulator until you are satisfied with the results. You can see the realtime response in the other windows at the same time. Beware: before you confirm the transformation do not click on any of the other windows otherwise it will be cancelled.
Repeat the transformation on the remaining pictures. In some cases you might need to scale the object. You can do so using Scale (hold Shift for proportional scaling).
It is very important to firstly adjust one view (for exaple the top view), then adjust the side view according to the top one, and finally adjust the front view according to the side view.
When you have completed the lining up, all you have to do is to select the pictures exactly around the drawing (you can take advantage of snapping the selection to the Guides or make a new selection using the dragging marquee and subsequently apply Select > Transform Selection). Make sure the width of your top and side views are equal. The height of both the side and front view should also be equal.
Rotate the top view by 90 degrees counter clockwise (Image > Rotate > Canvas > 90 CCW), so that the nose of X-38 will be pointing down. Through doing this, the setup in Maya will be easier.
Now save the individual pictures e.g. as .jpg. Name the files according to the views – front, top, side.
Import to Maya
Finally we are getting down to Maya.
Make a new project and scene if you haven´t done so yet. Move your tuned pictures to the directory named “sourceimages” located in the project´s directory.
Before importing any picture into Maya, create new cameras for every orthographic view (I call them “import cameras”). These will contain Image Planes with pictures. It can seem weird to create new cameras but the response in the orthographical viewport is slightly better than when you create the Image Plane in one camera but look at it through the default camera. That way we can avoid this bug. There is also an additional advantage that I will discuss later.
Create a new equivalent camera in every orthographic viewport (so for example, for the top view it would be Panels > Orthographic > New > top). Now you have got cameras with suffix 1. Rename them to e.g. front_import, side_import, top_import using Channel Box, Outliner or Hypergraph.
There are several ways to create the Image Plane in Maya. I personally use the following way: Choose the viewport where you want to create the Image Plane, go to View > Camera Attribute Editor. Unroll Environment section in the camera´s Attribute Editor. Click on “Create” next to the “Image Plane” section . Now you are in the Image Plane node´s attributes. In the “Image Plane Attributes” section click on the icon of directory located in the subsection “Image Name”. Import the corresponding picture. Repeat for all remaining views.
Switch to the corresponding default cameras (top, side, front) in every orthographic view (for example quickly via Hotbox). Your scene should now look like this:
Image Planes Adjustment
Image Planes need to be scaled and centred. They can be selected in the perspective view or now also in the orthographic view (assuming you are looking at it through another camera – that´s what we have previously done). This is the second advantage to creating new cameras that was mentioned earlier. After the selection of the Image Plane in the Channelbox, click on INPUTS. The attributes of the Image Plane will now appear. The important ones to focus on are: Center X, Center Z, Center Y, Width a Height (see picture).
Use Center X, Y, Z for the translation along individual axes. Click on e.g. Center X
Then move the cursor to the viewport and drag it to the right or left while holding down the middle button. It will activate the so-called Virtual Slider.
You will soon discover that the Virtual Slider is relatively inaccurate, and that is because it´s set to the highest speed by default. In the Channelbox go to Channels > Settings > Slow to adjust this. You can also enable the Hyperbolic for a hyperbolic curve.
You can change the speed using the icons on top of the Channelbox. Doubleclick on the icon with the manipulators to as to be able to switch to the two remaining icons on the right. When the setup is complete, click on the first icon to return it to the previous status.
Use Width and Height for scaling the Image Plane. Change the Width and Height simultaneously to ensure that the scaling is proportional.
Tip: It pays to slow down the virtual sliders when moving inserted edgeloops using the Weight parameter. I created a Marking Menu for that case.
First, line up the top view. The (imaginative) center line of the blueprint should match the grid axis. So the aircraft should be “divided” in two by the grid axis. If preferred, you can use some geometry to do that – e.g. a Plane with a colored wireframe. Now center the front view, and finally center the side view according to the top and front view.
If you want to use your Image Planes in the perspective view it´s a good idea to push them back so that they won´t get in your way during modeling (see picture).
A final revision needs to be done using the Planes (Create > NURBS > Primitives > Plane) and/or locators (Create > Locator).
There are just a couple more adjustments and then we are ready to go.
If you want to darken or prime the Image Plane, select it, hit Ctrl+A for the Attribute Editor and click on the Image Plane´s tab. In the Image Plane Attributes section, move the Color Gain slider or click on the white oblong found on the side and change the prime colour.
To ensure that the Image Planes cannot be selected during modeling, switch off the selectability of the cameras in the Status Line (see picture below).
If you want to hide the Image Planes in the perspective view, switch off the visibility of the cameras (Show > Cameras).
Now everything is prepared for modeling.
Back blueprint setup
If you have got a blueprint of the objects back side (e.g. car blueprints often contain them) some extra setup steps will be required.
Create a new camera in the front viewport, name it “back_import”. Create an Image Plane and import the picture. Adjust it. Then select the camera “back_import”, rotate it by 180 degrees along the Y (vertical) axis, then move it to the other side along the Z axis behind your future model. Move the Image Plane using Center Z in front of the model (see picture).
Finally create one more camera named “back”. Copy the Translate and Rotate parameters of the “back_import” camera to the “back” camera so that they will be situated in the same place. I personally use the Koshigaya plugin – Copy and Paste function for this. “Copy” copies the attributes of the selected object, and after selecting another object and clicking on “Paste”, it will cause the second object to adopt the same parameters as the first object.
In order to switch between the front and “back” view, all you have to do is hide one import camera using Outliner and Channelbox (Visibility: “0” or “off” – invisible; Visibility: “1” or “on” – visible). You can also use the Hypergraph for this operation.
Therefore for modeling according to the back blueprint, hide the “front_import” camera and make the “back_import” camera visible. Then in the orthographic front view switch to the “back” camera.
Conversely for modeling according to the front blueprint, hide the “back_import” camera, make the “front import” visible and in the orthographical view switch to the “front” camera.
The whole setup may seem tedious but it is very useful in focusing on the very process of modeling instead of fighting with badly set up Image Planes. You can even automate some of the steps using MEL.