Specular Hightlighting

Well, the good news is that the highlight is in the right place, and appears to be a reasonable size. I suspect the coloration issue is caused by the specular value being too large (perhaps by several magnitudes), so that when it is added to the base color, it causes the color to overflow. Can you share your code somewhere, at least the bit that computes the shading?

Thank you! Also, what values are you using for ambient, diffuse, and specular for that scene? And what is the output of those debug values when you define _DEBUG?

Yes, for debugging these kinds of issues, here's a technique that's been invaluable for me.

First, find the coordinates of a pixel in the output image that you want to inspect. In your case, just pick one somewhere in that odd specular highlight. Let's call that pixel (x, y).

Next, fiddle with your render() function so that it only actually renders the pixel at (x, y).

Now, your debug output should be significantly smaller, and should relate only to the production of that specific pixel. This can really help you zero in on issues like this, so you can see what's happening when things go wrong.

Excellent! I'm glad you were able to figure it out.

I created this account because i had this exact same issue. Thanks for this post!
Im really enjoying this book so far.

Another solution and the test pass, and the highlight shows, but that might because there is no test coverage

on page 89

if (reflect_dot_eye <= 0)

{

specular <- black;

}
changed to


if (reflect_dot_eye <= -1)

{

specular <- black;

}
Don't know if the issue is with the calculation of reflectv and reflect_dot_eye or the test of the results.

Just finished chapter 7 awesome stuff, great to spend many hours learning and writing test, thinking will this work, then getting to Putting It Together and seeing an image!!!

When the comparison relative to zero (<= 0) The logic passes the test, but fails to put a highlight on the image. Now I have tried swapping lightv with the comparison relative to 0. Test failed the image is an inverse, shadow on top and the highlight is missing. If I changed the comparison relative to -1 then I get the inverse image, shadow on top and the highlight on bottom, test don't pass.

Which test covers the comparison relative to zero (<= 0) is true? I have tools that tell me if logic has test coverage, I get no coverage for when comparison to zero (<= 0) is true, comparison is always false. Based on what I'm reading I was thinking scenario: Lighting with the eye between light and surface, eye offset 45°. I get zero for specular but not because of the comparison but because of the calculations from computing the specular contribution. With that scenario lightv is (0, 0, -1), light_dot_normal is 1, normalv is (0, 0, -1), reflectv is (0, 0, -1), reflective_dot_eye 0.707106

I guess if I know if the individual variables used are correct and/or what the value should be.

The following code is is in a Material class that works and passes test using comparison of reflect_dot_eye <= -1.
Material.cs

public RtColor Lighting(PointLight light, RtPoint point, RtVector eyeVector, RtVector normalVector)

{
   var effectiveColor = Color * light.Intensity;
   var lightVector = (light.Position - point).Normalized();
   var ambient = effectiveColor * Ambient;
   var lightDotNormal = lightVector.Dot(normalVector);

   RtColor diffuse;
   RtColor specular;

   if (lightDotNormal < 0)
   {
       diffuse = RtColor.Black;
       specular = RtColor.Black;
   }
   else
   {
       diffuse = effectiveColor * Diffuse * lightDotNormal;

       var reflectVector = lightVector.Negate().Reflect(normalVector);
       var reflectDotEye = reflectVector.Dot(eyeVector);

       if (reflectDotEye <= -1)
       {
           specular = RtColor.Black;
       }
       else
       {
           var factor = Math.Pow(reflectDotEye, Shininess);
           specular = light.Intensity * Specular * factor;
       }
   }

   var results = ambient + diffuse + specular;
   return results;

}