Scaling Transformation // OpenGL Tutorial #8

[@BillXu-hh5ux]

Great! You are not only a good OpenGL programmer but also an excellent teacher!

[@OGLDEV]

Thank you!

[@staticebr2263]

Thank you for these tutorials, many simply tell you how to implement something but do not discuss the maths behind it.

[@OGLDEV]

Thanks a lot, this is exactly how I define my niche - explain both the math and the implementation.

[@Levendo]

These are really good tutorials! Very easy to understand.

[@OGLDEV]

Thank you :-)

[@hodayfa000h]

Great tutorial as always, thanks for your efforts!

[@OGLDEV]

My pleasure!

[@weifang2218]

Greate video, It's the best OpenGL tutorial I have seen.

[@OGLDEV]

Thanks :-)

[@bladman9700]

W series mate, i've been learning or at least "try" to learn opngl and it's been an entire year, by myself thru opengl book, leanngl or by someone's else book and many series on yt, i never got to this point, obv. by learning for this much time i do know basics of VAO, VBO, EBO and shaders b4 starting this series, But this is just better. Will definetly follow your new vulkan's series after finsih opengl, hope that till that time that series will be finsihed, and if you are still using linux for vulkan's series and please try to cover vulkan on wayland, as wayland's "feature" if something isnt committed the windows will not be shown/drawn, a security feature, and it's complicated for my noodle brain up there. Have a Nice Day!

[@OGLDEV]

Thanks! Right now I'm only building the Vulkan code on Windows but I think it should work on Linux as well. Isn't Wayland the default on Ubuntu? I only have time to support a single Linux distro and only the default installation.

[@antoinethomas3633]

Why isn't the matrix multiplication done in the shader (since shader = executed by the gpu = better performance for matrix multiplication)?

[@OGLDEV]

In the vertex shader we multiply each vertex by a single matrix. If you can prepare something on the CPU before the draw call you will save operations per vertex since it is done only once (per draw call at the most). This is why it's so common to prepare the WVP matrix (which we will talk about later on) on the CPU as a combination of world-view-projection matrices and still do a single vector-matrix multiplication in the vertex shader.