The method of using a spot light to create a mental ray area light requires higher sample levels and has known quality and performance issues. The recommended method of creating a mental ray area light is to use a Maya area light. See mental ray Area Light Attributes (created using area light) for attributes for mental ray area lights created using a Maya area light.
To obtain soft shadows for point and spot lights, tune the Light Radius and Shadow Rays attributes under the Raytrace Shadow Attributes section of the spot light’s shape node.
Area light background info: Area lights simulate light coming consistently and continuously from the entire surface of the area. Low sample values produce lower quality area lights because there are few lights interspersed in the area light surface, producing uneven distribution of light. High sample values produce area light surfaces that are densely populated with lights, and light is more evenly distributed over the area light surface.
Spot lights are basically point lights except that light is limited to a specific direction. The concept of a spot light is the same as a point light with a housing that blocks some of the light, but rendered much more efficiently. When using area spot lights, you get the effect of light being blocked between the samples because light is limited to within the spotlight cone. With area point lights, however, you get the effect of omni-directional light at the samples which better simulates light coming from the entire surface area. If you use area spot lights, you require a higher sample value to achieve continuous and even lighting from the entire surface of the light.
The following are attributes for mental ray area lights created using a Maya spot light.
When on, creates a mental ray for Maya area light. This option is disabled by default.
This is the shape of the area light source. Select from the drop-down list: Rectangle, Disc, Sphere, Cylinder and User. The default is Rectangle.
Select User if you wish to use your own light shader to define the light shape.
This represents the number of sample points emitted from the light (X and Y). The default is 3 and 3. Values greater than 3 increase the quality by reducing graininess, but may increase rendering cost.
To reduce rendering cost and time, it is not necessary to use a high number of sample points after a light ray has been reflected or refracted several times. This attribute represents the number of bounces that the light ray should go through before it uses the Low Samples values instead.
The default is 2 and 2.
This can be used if you want the area light object to be visible when rendered. The default is off.