Once you have loaded DRM_Nuthead.lwo, you will notice it is a subdivision object so we will need to freeze the mesh at a desired subdivision level before we proceed. For the following illustrations, I chose a subdivision level of 4 by choosing the File->Modeler->Options->General Options pull-down menu and then set the Patch Divisions parameter to 4. Now choose the Construct tab along the top of Modeler and click on the Freeze button in the Convert section along the left side of Modeler. This should result in an object with 6288 polygons. The object has been assigned a total of 7 vertex maps, of which 4 of those are Weight maps (body, head, leg_left, and leg_right) and 3 Color maps (1 RGB map named SQ_vertmap and 2 RGBA maps, one also named SQ_vertmap and the other named Vertex_ColorSQ).

As mentioned earlier, the bulk of this tutorial deals with Nuthead's Weight maps, however before we start talking about them I would like to emphasize a point about the Color maps and discontinuous vertex maps in qemLOSS3. If you run qemLOSS3 and click on the RGB tab, you will find the "Constrain Discontinuous RGB Map Vertices" parameters ghosted, even though I mentioned above that there was a single RGB map assigned to the object. The ghosted parameters is an indication to you  that the SQ_vertmap RGB map does not contain any discontinuous vertices. Now click on the RGBA tab, and you will find the  "Constrain Discontinuous RGBA Map Vertices" parameter is not ghosted, meaning that there is at least one of the two RGBA maps that contains discontinuous vertices. Temporarily click on the "Constrain Discontinuous RGBA Maps Vertices" selection box, and then pull down the RGBA Maps menu. Only the Vertex_ColorSQ map shows up in this menu, meaning that RGBA map has some discontinuous vertices and the SQ_vertmap RGBA map only contains continuous vertex values. I wanted to use the Color maps to emphasize the fact that not all vertex maps will show up in the "Constrain Discontinuous Vertices" pull-down menu, only those vertex maps with discontinuous vertices will appear (this is true for all vertex map types, not just the RGB and RGBA maps discussed in this paragraph). Make sure you unselect the previously enabled "Constrain Discontinuous RGBA Maps Vertices" for the remainder of this tutorial, although the only area of discontinuous vertices in the Vertex_ColorSQ map is around the nose and I don't imagine it would cause much difference in the results seen in this tutorial.

Finally. let's start exploring the use of Nuthead's Weight maps as qemLOSS3 reduction constraints. Click on the qemLOSS3 WGHT tab, and you should immediately notice that the "Constrain Discontinuous Weight Map Vertices" parameters are ghosted. From the previous paragraph, you should realize that this means that the 4 Weight maps assigned to the object do not have any discontinuous vertices, and we won't have to worry about that set of parameters. However, the "Use Weights as Reduction Constraint" parameter is unghosted and available for use. We will select this parameter shortly (but don't do it yet), and find that all Weight maps assigned to the object will be listed in the associated pull-down "Weight Maps" menu. But for now just leave this parameter unselected and reduce the object to a Reduction Goal (Percent) of 50%. Any weight values assigned to vertices in the object have been ignored, and the reduction takes place as if there were no Weight maps assigned to the object at all. We will use this reduction as a basis of comparison for the remainder of the reductions done in this tutorial.

Switch back to the original object in layer 1 and run qemLOSS3 again. This time select the "Use Weights as Reduction Constraint" under the WGHT tab. Now the "Weight Maps" menu and "Multiplier" parameter have been unghosted and are available for use. If you pull down the "Weight Maps" menu you will notice that the 4 Weight maps assigned to this object are listed in this menu, and if you select each weight map you will also see that the "Multiplier" for each individual Weight map will still show the default of 100. We'll leave the "Multiplier" value of each Weight map at 100 for this example, so go ahead and click on OK to run qemLOSS3. The reduced object that results should be different than the first reduction we performed, and if you display the first reduction in the background layer you will see a Perspective OpenGL view similar to the following image.

Notice that by comparing the 2 different reductions, the only significant changes to the mesh is in the arms and tail. The polygons in the arms and tail of the object have been reduced much more than the other parts of the object. This is because the vertices in the body, head, leg_left, and leg_right Weight maps  had all been assigned a value of 1.0 (or 100%). So when we assigned a "Multiplier" of 100 to every Weight map, we increased the geometric error weight for all those vertices by 100 (1.0 x 100). The arms and tail of the object have not been assigned any weight values at all, and therefore their geometric error weight was not increased at all. So the vertices that were assigned weights were less likely to be removed than those vertices without weight values.

Now let's try changing the "Multiplier" value for individual Weight maps. Switch back to the original object in layer 1, run qemLOSS3 and select te WGHT tab again. Pull down the "Weight Map" menu and choose the leg_left Weight map. Change the value in the "Multiplier" requester to 0. Follow the exact same procedure for the leg_right Weight map. Now if you select all the different maps in the "Weight Map" menu, the "Multiplier" field will show the multipler value assigned to each individual Weight map. In other words if you select the body or head Weight map, their multipliers should still be set to 100, and the leg_left and leg_right maps should be set to 0. Click on OK to perform the new reduction and if you display that reduction with the first reduction in layer 2 displayed in the background, you will see a Perspective OpenGL view similar to the following image.

By comparing this image to the first one on this page, you can see that there has now been some reduction of the polygons in the legs of the object. And fewer reductions have taken place in the arms and tail because fewer polygons needed to be removed from those areas in order to meet the reduction goal. The head and body maps have still prevented many changes in those areas because their multipliers were still set to 100.

In the existing Weight maps, all the weight values are set to 1.0, but that doesn't need to be the case. To illustrate this, create a new Weight map called Tail. The following image is a Weight Shade view of the Tail Weight map that I created. The vertices at the top of the tail have been assigned weight values of 5.0 (500%), the middle of the tail has values of 1.0 (100%), and the values gradually reduce to 0 at the base of the tail. This way, when I set the Tail multiplier to 20 the next time I run qemLOSS3, the vertices at the top of the tail will have added geometric error weights of 100 (5.0 x 20), the vertices in the middle of the tail will have added geomteric weights of 20 (1.0 x 20), and will gradually reduce to vertices with additional geometric error weights of 0 at the base of the tail.

So let's run qemLOSS3 one last time, and set the weight multipliers to these values for each Weight map: body - 50, head - 100, leg_right - 0, leg_left - 0, Tail - 20. Since there are 2 maps that will be set to 0, we can use that as the default initial value for all weight maps. So while the "All Weight Maps" is shown in the "Weight Maps" pull-down menu, type a 0 in the "Multiplier" requester. Now select the body map from the pull-down and change the "Multiplier" requester to 50, select the head map and change its multiplier to 100, and select the Tail map and change its multiplier to 20. We shouldn't need to set the leg maps individually, since those were already set to 0 by changing the default multiplier value for "All Weight Maps". Again, if you re-select all the different Weight maps, you should see the "Multiplier" value that is assigned to each map. After  the reduction proces is complete you will end up with an OpenGL Perspective view similar to the following image (the first reduction in layer 2 is displayed as the background layer). As you can see, the top of the tail has not been reduced as much as it was in previous reductions, and the base of the tail has been reduced more than the top.

By using Weight maps along with the Weight Map Reduction Constraint parameters, you can gain tremendous control over the reduction process and the amount of reduction taking place on different parts of your object. The other vertex map types that have Reduction Constraint parameters are the Relative and Absolute Endomorph maps (MORF and SPOT). The "Multipler" parameter for those map types is very similar to the Weight map multiplier used in this tutorial, except that instead of using the weight map values, the Endomorph multipliers are multiplied by the length of the Endomorph vector. The next tutorial will therefore be very similar to this one, and will reinforce the notion of a "Multiplier" in the context of Endomorphs.