I went through several of your OpenGL videos last night. I have to say you present some of the most clear and correct content on the topic I have ever come across. No waffling, straight to the point and easy to understand. Thank you
Brian, I greatly appreciate the work you put into this subject and the accompanying book. It is concise, explained well and most of all touches on topics without over complicating them. Kudos to you!
If i'm not mistaken, the division by PI in the diffuse component and division by 4 * NdotL * NdotV is the normalization factor for keeping our NDF normalized (equal to 1)
Depending on the source, some use Pi, other use 4... I didn't find an explanation on why each value, but they're used. I think Disney and UE4 used 4, until Epic went to Pi in UE5
5:50 k is a function of alpha, (alpha+1)^2/8 for a direct light source and alpha^2/2 for image-based lighting Regarding the term GGX, this is so funny, it cracks me up every time I see it. Nobody knows what it means. As in not a single person on earth. My theory: the guy who invented it was drunk and wrote it down. The next morning, he couldn't remember anything but still sent the paper to Disney. That's what happened.
4:18 how can the sphere reflect light from a point of its surface that is not even exposed to light? images for α 0.1, 0.3 and 0.5 i can accept, but 0.8 and 1.0 must be incorrect?
In my opinion its right. he talking about roughness on the left its almost clear surface "mirror" this is how light should looks like in mirror and with higher roughness its starts spreading and bouncing into random direction instead of directly to "camera"
The line kD = 1.0 - metallic threw me off at first, but maybe now I understand. So kD+kS=1 is for dielectrics, however, conductors have also some kind of a grounding term (e.g. kD+kS+kConductiveGrounding=1)?
What boggles my mind: is not a N dot V always negative ? Since N points outwards of object and V inwards, so the angle is always greater than 90° isn't it ?
Not if those textures are already in the sRGB color space. GL can omit that transform of you have already the texture expressed in sRGB. But that isn't said
I went through several of your OpenGL videos last night. I have to say you present some of the most clear and correct content on the topic I have ever come across. No waffling, straight to the point and easy to understand. Thank you
Brian, I greatly appreciate the work you put into this subject and the accompanying book. It is concise, explained well and most of all touches on topics without over complicating them.
Kudos to you!
i can not put into words how much i needed this
thank you so much!
Another well detailed rendering tutorial, nice job! ;)
Excellent video and very clear presentation of PBR. Thank you.
Thanks for the explanation the shading, technique and the code.
If i'm not mistaken, the division by PI in the diffuse component and division by 4 * NdotL * NdotV is the normalization factor for keeping our NDF normalized (equal to 1)
Depending on the source, some use Pi, other use 4... I didn't find an explanation on why each value, but they're used. I think Disney and UE4 used 4, until Epic went to Pi in UE5
5:50 k is a function of alpha, (alpha+1)^2/8 for a direct light source and alpha^2/2 for image-based lighting
Regarding the term GGX, this is so funny, it cracks me up every time I see it. Nobody knows what it means. As in not a single person on earth. My theory: the guy who invented it was drunk and wrote it down. The next morning, he couldn't remember anything but still sent the paper to Disney. That's what happened.
I wish I could give more thumbs ups, thanks for this great explanation
Thanks for the videos Brian. Really informative.
Very good explanation, thanks!
4:18 how can the sphere reflect light from a point of its surface that is not even exposed to light? images for α 0.1, 0.3 and 0.5 i can accept, but 0.8 and 1.0 must be incorrect?
In my opinion its right. he talking about roughness on the left its almost clear surface "mirror" this is how light should looks like in mirror and with higher roughness its starts spreading and bouncing into random direction instead of directly to "camera"
finally. i always wondered what pbr was. was it gpu hardware, a specific shader, or just a loose guide for standardizing material properties
"Finnell effect"
I am dying
The line kD = 1.0 - metallic threw me off at first, but maybe now I understand. So kD+kS=1 is for dielectrics, however, conductors have also some kind of a grounding term (e.g. kD+kS+kConductiveGrounding=1)?
What boggles my mind: is not a N dot V always negative ? Since N points outwards of object and V inwards, so the angle is always greater than 90° isn't it ?
Confuses me too. Though I’ve noticed most shaders calculate the view vector as pointing away from the fragment instead of towards.
At the end of your video it looks like you're not correcting the gamma on the Albedo textures; I believe that should happen, no?
Not if those textures are already in the sRGB color space. GL can omit that transform of you have already the texture expressed in sRGB. But that isn't said
How would PBR work for one directional light, this is for just point lights, correct?
Use the directional light direction vector directly as the light vector. Ie, don’t subtract it from the fragment position
Pbr is so hot right now.
Based you say?
First! on Feb. 28, 2022 that is...