5 - Symmetry and Sub-division
This tutorial describes how you can use Symmetry, so that you can make changes to one side of the model, and it affects both sides; also how to subdivide your mesh in order to smooth it off.
There are several ways to mirror your model.
Mirror tool
Select the object – and select Mirror tool from the tool bar. Choose which axis that you want to mirror the object in. You can then choose from the following:
Copy – a unique object
Instance – provides information between the clone and the original
Reference – One way communication, the clone mimics the original with changes but you cannot modify the clone.
This way of mirroring applies an inverse scale matrix, so that if the model is merged into another programme the normals could end up becoming inverted, which causes problems.
Clone
Select the object > right click > clone > reference.
When you select reference, Max places a grey bar above edit poly in the stack, suggesting that anything below the grey bar is duplicated, anything above will be specific to the clone.
Mirror modifier
This is the export friendly way to mirror your object.
Select the object and apply the mirror modifier and check ‘Flip face’. If the axis of the object is too close and it looks odd, select the + sign next to symmetry in the stack and it will reveal ‘Mirror’. With this selected you can move your clone and orientate it.
If you click on ‘Show end result’ you can see the changes you are making to the mesh without having to toggle from mirror to edit poly.
Subdivision Surfaces
In order to smooth your surfaces within editable poly you can choose ‘nurbs sub-division’. Here you can choose the amount of iterations you would like, and you can also specify whether you require those iterations on rendering.
Choosing sub-division can highlight problems in your modelling.
There are several other subdivision modifiers such as ‘Turbosmooth’, ‘Meshsmooth’, ‘HSDS’, ‘Subdivide’ and ‘Tessilate’.
Tuesday, 2 September 2008
Tutorial 4b - Using Splines to model a head
4b Using splines to model a head
Eric Maslowski
Giving the model shape.
First we need to fix problems that arise when we converted these quads to a surface.
First, select object, click on vertex in sub object level, zoom in on forehead, turn off snaps chose move tool, Click on quads and you will see that they are not attached. We need to weld these together to create one continuous mesh
Click and drag over all vertices, in the modify panel > edit vertices > Weld. To check that you have welded all the vertices check the amount of vertices before and after as you may weld or collapse two vertices together. Now decrease to 0 increase by 1 click and check the vertices. It should be the same for ‘after’ each time. Click OK. Now the vertices are attached.
Go into both views, now we start giving image some depth. Take the points and align them to key features. This is why it is important that your images line up.
Firstly select all the points which are on the border straight down the middle. Select > edit geometry > make planar > click on x, which will align in the x axis.
Now switch to left view and click and drag the vertices across on the x axis.
Reposition the left plane if it is overlapping. Unfreeze the reference plate and move to side where it won’t interfere with the object. Re-freeze.
Select vertices and take individual points and match the profile. Just use the x axis.
It is difficult to see what is going on. If you right click > properties, see through object, it makes the face semi transparent. (Or Alt x).
Choose edged faces. Align as best you can.
Now go to next row. Continue to drag the vertices over. The forehead is a fairly flat area, so the vertices of the front line will be very close to those in the next line. Work in layers working from right to left.
Eric Maslowski
Giving the model shape.
First we need to fix problems that arise when we converted these quads to a surface.
First, select object, click on vertex in sub object level, zoom in on forehead, turn off snaps chose move tool, Click on quads and you will see that they are not attached. We need to weld these together to create one continuous mesh
Click and drag over all vertices, in the modify panel > edit vertices > Weld. To check that you have welded all the vertices check the amount of vertices before and after as you may weld or collapse two vertices together. Now decrease to 0 increase by 1 click and check the vertices. It should be the same for ‘after’ each time. Click OK. Now the vertices are attached.
Go into both views, now we start giving image some depth. Take the points and align them to key features. This is why it is important that your images line up.
Firstly select all the points which are on the border straight down the middle. Select > edit geometry > make planar > click on x, which will align in the x axis.
Now switch to left view and click and drag the vertices across on the x axis.
Reposition the left plane if it is overlapping. Unfreeze the reference plate and move to side where it won’t interfere with the object. Re-freeze.
Select vertices and take individual points and match the profile. Just use the x axis.
It is difficult to see what is going on. If you right click > properties, see through object, it makes the face semi transparent. (Or Alt x).
Choose edged faces. Align as best you can.
Now go to next row. Continue to drag the vertices over. The forehead is a fairly flat area, so the vertices of the front line will be very close to those in the next line. Work in layers working from right to left.
Tutorial 4 - Using Splines to model a head
4 - Using splines to model a head
Eric Maslowski
We are now starting the modelling process.
Set up your view ports. Right click on perspective in the corner of the view port - configure and chose layout - two side by side. Change by clicking on name to change view port > Ok. Front view port is allocate to the left, And the left view port on the right.
Change to smooth and highlights and turn off show grid from the same menu.
Zoom into viewport with the zoom extents button.
Drawing the topology.
Make front view port full screen. Draw each quad using the line tool, that we originally drew in Photoshop, which will be converted to a mesh.
Create > shapes > line. Creation method – no Bezier as this will add triangles. We need to use Hard corners., so initial and drag types should be set to corner.
Click once to define a starting point, click to form your quad and returm back to start in order to close the spline Say yes to the dialogue box.
Turn on snap so that when you reach existing spline it will snap to it. Right click on 3D Snap, click on vertex. Continue to do this process of making quads. This can actually save you time in the long run. Do this over the entire surface.
Area around eye. If five sided area, approximate certain areas and simplify the number of points. Click on line and delete if you make a mistake. Back space on keyboard will remove the last vertex that you have made.
If the point won’t select particularly in difficult areas, you can change the snap type and go to options and adjust the radius. If too sensitive drop the number of pixels.
Keep it simple as detail can be added later on.
Do not fill in the nostrils straight away.
Creating a surface
Get out of line mode. Select shape > modify panel > right click on shape , convert to editable poly. This produces a single polygon. Go to ‘Edit geometry’ in modify panel > attach > click small box and a dialog box lets you select the different quads. Click all, attach and have one complete surface.
Chose Edged faces can now see the topology.
Eric Maslowski
We are now starting the modelling process.
Set up your view ports. Right click on perspective in the corner of the view port - configure and chose layout - two side by side. Change by clicking on name to change view port > Ok. Front view port is allocate to the left, And the left view port on the right.
Change to smooth and highlights and turn off show grid from the same menu.
Zoom into viewport with the zoom extents button.
Drawing the topology.
Make front view port full screen. Draw each quad using the line tool, that we originally drew in Photoshop, which will be converted to a mesh.
Create > shapes > line. Creation method – no Bezier as this will add triangles. We need to use Hard corners., so initial and drag types should be set to corner.
Click once to define a starting point, click to form your quad and returm back to start in order to close the spline Say yes to the dialogue box.
Turn on snap so that when you reach existing spline it will snap to it. Right click on 3D Snap, click on vertex. Continue to do this process of making quads. This can actually save you time in the long run. Do this over the entire surface.
Area around eye. If five sided area, approximate certain areas and simplify the number of points. Click on line and delete if you make a mistake. Back space on keyboard will remove the last vertex that you have made.
If the point won’t select particularly in difficult areas, you can change the snap type and go to options and adjust the radius. If too sensitive drop the number of pixels.
Keep it simple as detail can be added later on.
Do not fill in the nostrils straight away.
Creating a surface
Get out of line mode. Select shape > modify panel > right click on shape , convert to editable poly. This produces a single polygon. Go to ‘Edit geometry’ in modify panel > attach > click small box and a dialog box lets you select the different quads. Click all, attach and have one complete surface.
Chose Edged faces can now see the topology.
Tutorial 3 - Building of 3D Reference Plates
3 - Building of 3D Reference Plates
Eric Maslowski
Create > tab > geometry > plane. Create plane that will hold the front view of the image.
Set up units of scale. Click and drag plane in view port.
Modify panel > change the length and width segments down to 1.
Position plane – in top view port. We created plane at 0, but the plane needs to be positioned behind what we are modelling. Click and drag plane to the side and move back and over to the left.
Create material that will hold the photos
Material editor> select material slot and assign to diffuse channel > bitmap > ok Find our topology file which is a PSD file. Open as this is a Photoshop file, we can collapse all layers or individual layer. Use collapsed layers if you have several layers in the image.
Click and drag image on to the plane. You will notice that the map doesn’t show up, which is default behaviour in Max, in order to see the texture go to the blue and white chequered box, click to see image.
The image looks severely squashed; we need to maintain the same aspect ratio. To do this we need to modify texture co ordinates through the modify panel > UVW map > to modify the texture co-ordinates for the plane. We could decrease the U tile amount but not precise enough, so use a tool called ‘bitmap fit’ under ‘alignment’ in the modify panel.
On the modify panel where you see UVW mapping in the stack, click on the plus sign – this reveals a gismo, this can move rotate or scale image it is applied to. Move tool, click on x axis and move to fit the face on to the plane properly.
As you rotate image around it gets darker. We want to see the reference plane clearly behind the object.
The Material editor is hierarchal, when assigning a map you are assigning a child, so need to ‘go to parent’, back to bitmap, need to remove lighting by self illuminating the material. In the self illumination panel change the illumination to 100. Close material editor.
Copy our existing object. We need to tweak to show the left side. Switch to rotate shift click and drag it makes a copy to 90 degrees. Press A on key board which is angle snap if you are having trouble.
Move to new location set up left plane. You will notice that the image is facing the wrong direction and the plane isn’t big enough to hold profile image.
Therefore we need to flip the image – flip is next to UV tile.
In order to add space to accommodate the entire side of the head to left and right – jump back to plane increase the width to whatever is needed.
Tweek the UV mapping gismo and adjust further. Now we have two nicely positioned reference planes.
Select one ref planes – right click freeze selection the panel turns grey. You need to modify the behaviour of the plane. Therefore left click > properties – Check freeze then uncheck ‘show frozen in grey’ this preserves the appearance of the object. Do the same to the profile view.
Eric Maslowski
Create > tab > geometry > plane. Create plane that will hold the front view of the image.
Set up units of scale. Click and drag plane in view port.
Modify panel > change the length and width segments down to 1.
Position plane – in top view port. We created plane at 0, but the plane needs to be positioned behind what we are modelling. Click and drag plane to the side and move back and over to the left.
Create material that will hold the photos
Material editor> select material slot and assign to diffuse channel > bitmap > ok Find our topology file which is a PSD file. Open as this is a Photoshop file, we can collapse all layers or individual layer. Use collapsed layers if you have several layers in the image.
Click and drag image on to the plane. You will notice that the map doesn’t show up, which is default behaviour in Max, in order to see the texture go to the blue and white chequered box, click to see image.
The image looks severely squashed; we need to maintain the same aspect ratio. To do this we need to modify texture co ordinates through the modify panel > UVW map > to modify the texture co-ordinates for the plane. We could decrease the U tile amount but not precise enough, so use a tool called ‘bitmap fit’ under ‘alignment’ in the modify panel.
On the modify panel where you see UVW mapping in the stack, click on the plus sign – this reveals a gismo, this can move rotate or scale image it is applied to. Move tool, click on x axis and move to fit the face on to the plane properly.
As you rotate image around it gets darker. We want to see the reference plane clearly behind the object.
The Material editor is hierarchal, when assigning a map you are assigning a child, so need to ‘go to parent’, back to bitmap, need to remove lighting by self illuminating the material. In the self illumination panel change the illumination to 100. Close material editor.
Copy our existing object. We need to tweak to show the left side. Switch to rotate shift click and drag it makes a copy to 90 degrees. Press A on key board which is angle snap if you are having trouble.
Move to new location set up left plane. You will notice that the image is facing the wrong direction and the plane isn’t big enough to hold profile image.
Therefore we need to flip the image – flip is next to UV tile.
In order to add space to accommodate the entire side of the head to left and right – jump back to plane increase the width to whatever is needed.
Tweek the UV mapping gismo and adjust further. Now we have two nicely positioned reference planes.
Select one ref planes – right click freeze selection the panel turns grey. You need to modify the behaviour of the plane. Therefore left click > properties – Check freeze then uncheck ‘show frozen in grey’ this preserves the appearance of the object. Do the same to the profile view.
Tutorial 2 - Defining topoloy lines on reference images
2 - Defining topology lines on reference images
Eric Maslowski
Topology is the flow of surfaces over an environment or object.
We want to break the face in to key areas. What is meant by key areas?
If you look at an anatomy drawing, you can see how the muscles flow round key features, so our edge loops should have a similar flow to those muscles.
If you looking at a simple shape, such as a plane with a line bisecting it, if you drag the vertices up the shape will crease. If you move that bisecting line you get a different surface. This shows how edges play a role. We want and edge wherever we want a crease or protrusion in the surface that we need to control.
On your image, add edges at the wrinkles. Create a new layer, chose a bright colour, brush size 3, start drawing in certain areas, just to define the basic areas where there is a large crease or wrinkle or protrusion.
Just go through and try to isolate all the key areas.
Around eyes, eyebrows, crow’s feet bridge of nose, tip of nose, cheek bone. The cheek bone needs a lot of control as it gives definition to the face. Also make sure that you draw in the indentation under cheek bone and nasal folds. Think about how these join up. And draw in the topology. This is a good planning stage until we start modelling the character.
Now choose a new colour zoom in and have a smaller brush size and start tracing lines and placing quads. This gives flexibility, subdivision algorithms depend on quads and when you subdivide you get no pinching effects – if there is a triangle you will get pinching which is difficult to get rid off.
First stage is definition of eye and nose and side of nostrils. Place a small set of quads in between the nasal folds which gives flexibility later on when modelling for expression.
Add the cheek, and nose. You may get a direction changing quad, which can continue down the cheek or round the eye so topology can flow in any direction needed.
Chin area and side of mouth. Smile line on mouth is important so dimples can be added or puffy cheeks.
Forehead and upper lip edge flow from the cheek bone, try to approach the upper lip vertically. All muscles coming out of the lip are coming up vertically.
Nose – it can be difficult to see how this area can be broken into quads, try to avoid triangles. Think how surface should flow.
Sit back and think about your approach. If you create another layer and play around with your quad building you will become more aware of the process.
Now plot guidelines on the side image. Put on layers that you can toggle these on and off.
Eric Maslowski
Topology is the flow of surfaces over an environment or object.
We want to break the face in to key areas. What is meant by key areas?
If you look at an anatomy drawing, you can see how the muscles flow round key features, so our edge loops should have a similar flow to those muscles.
If you looking at a simple shape, such as a plane with a line bisecting it, if you drag the vertices up the shape will crease. If you move that bisecting line you get a different surface. This shows how edges play a role. We want and edge wherever we want a crease or protrusion in the surface that we need to control.
On your image, add edges at the wrinkles. Create a new layer, chose a bright colour, brush size 3, start drawing in certain areas, just to define the basic areas where there is a large crease or wrinkle or protrusion.
Just go through and try to isolate all the key areas.
Around eyes, eyebrows, crow’s feet bridge of nose, tip of nose, cheek bone. The cheek bone needs a lot of control as it gives definition to the face. Also make sure that you draw in the indentation under cheek bone and nasal folds. Think about how these join up. And draw in the topology. This is a good planning stage until we start modelling the character.
Now choose a new colour zoom in and have a smaller brush size and start tracing lines and placing quads. This gives flexibility, subdivision algorithms depend on quads and when you subdivide you get no pinching effects – if there is a triangle you will get pinching which is difficult to get rid off.
First stage is definition of eye and nose and side of nostrils. Place a small set of quads in between the nasal folds which gives flexibility later on when modelling for expression.
Add the cheek, and nose. You may get a direction changing quad, which can continue down the cheek or round the eye so topology can flow in any direction needed.
Chin area and side of mouth. Smile line on mouth is important so dimples can be added or puffy cheeks.
Forehead and upper lip edge flow from the cheek bone, try to approach the upper lip vertically. All muscles coming out of the lip are coming up vertically.
Nose – it can be difficult to see how this area can be broken into quads, try to avoid triangles. Think how surface should flow.
Sit back and think about your approach. If you create another layer and play around with your quad building you will become more aware of the process.
Now plot guidelines on the side image. Put on layers that you can toggle these on and off.
Tutorial 1 - Preparing photos to use as reference plates
1 - Preparing photos to use as reference plates in Photoshop CS2
Eric Maslowski
You will need two images of a face, a front view and a side view – ‘3D Data SK’ is an on-line resource which has hundreds of reference photos which can be used rather than having to take your own photographs.
You need to remove any distortion introduced by the lens, camera or photographer – the lens causes barrel distortion, so you need to remove this distortion to produce a flat image.
Lens Distortion is a filter. Go to Filter > Distort >Lens correction.
Tweek the ‘Remove distortion’ slider, which is similar to the spherize filter in previous versions of Photoshop, you can also use vertical perspective or rotate the image.
To use the ‘Straightening tool’ click and drag between two areas that are horizontal. Then release left mouse button and the image straightens up.
Select – filter > distort > lens correction.
Look at symmetry of the image, so make sure that the person is fairly symmetrical. Unfortunately most people are asymmetrical naturally so you need to look at markers which may imply that the lens has tilted the image. Look at the position of pupils, ear lobes, nostrils and edges of the mouth.
Click on the ‘Straighten’ tool and drag between these different areas to see if the image is straight.
Using the ‘Remove distortion’ slider, slide back until fisheye effect is removed. The image flattens out.
Do the same for the side image.
Merge the two images together and make sure the key features line up.
This gives you two points of reference when modelling later.
Control A, Control C, then Control V to paste the image and move it to the side.
Go to ‘Image canvas size’ and extend canvas 200% to the right.
Take layer one and set opacity to 60-70 % so you can see the face behind it. Pick common points for lining up such as the side of eye or mouth. ‘Control T’ will put you into free transform mode, rotate the image slightly, move the centre to your reference point, so can line up all ref points at same time.
Return opacity to 100% and the clear up background with the eraser tool.
Double check results by using guide lines – rulers, click and drag guide lines to the centre of the pupils, under nose, through mouth and under ears or the top of the ears or eyebrows and you can then see how well you are lined up.
Crop the image
Save as Jpeg.
When using the lens distortion if you have complete control ptlens plug in which removes barrel distortion.
Eric Maslowski
You will need two images of a face, a front view and a side view – ‘3D Data SK’ is an on-line resource which has hundreds of reference photos which can be used rather than having to take your own photographs.
You need to remove any distortion introduced by the lens, camera or photographer – the lens causes barrel distortion, so you need to remove this distortion to produce a flat image.
Lens Distortion is a filter. Go to Filter > Distort >Lens correction.
Tweek the ‘Remove distortion’ slider, which is similar to the spherize filter in previous versions of Photoshop, you can also use vertical perspective or rotate the image.
To use the ‘Straightening tool’ click and drag between two areas that are horizontal. Then release left mouse button and the image straightens up.
Select – filter > distort > lens correction.
Look at symmetry of the image, so make sure that the person is fairly symmetrical. Unfortunately most people are asymmetrical naturally so you need to look at markers which may imply that the lens has tilted the image. Look at the position of pupils, ear lobes, nostrils and edges of the mouth.
Click on the ‘Straighten’ tool and drag between these different areas to see if the image is straight.
Using the ‘Remove distortion’ slider, slide back until fisheye effect is removed. The image flattens out.
Do the same for the side image.
Merge the two images together and make sure the key features line up.
This gives you two points of reference when modelling later.
Control A, Control C, then Control V to paste the image and move it to the side.
Go to ‘Image canvas size’ and extend canvas 200% to the right.
Take layer one and set opacity to 60-70 % so you can see the face behind it. Pick common points for lining up such as the side of eye or mouth. ‘Control T’ will put you into free transform mode, rotate the image slightly, move the centre to your reference point, so can line up all ref points at same time.
Return opacity to 100% and the clear up background with the eraser tool.
Double check results by using guide lines – rulers, click and drag guide lines to the centre of the pupils, under nose, through mouth and under ears or the top of the ears or eyebrows and you can then see how well you are lined up.
Crop the image
Save as Jpeg.
When using the lens distortion if you have complete control ptlens plug in which removes barrel distortion.
Subscribe to:
Posts (Atom)