This is an exercise in learnopengl and it provides a solution using texture by rendering the scene in an offline buffer and then use it as a texture for the mirror.
I tried to implement it using stencil buffer and got a strange behavior that I could not explain. Here are the relevant parts of my code:
shader.use();
glEnable(GL_STENCIL_TEST);
glClearColor(0.1f, 0.1f, 0.1f, 1.0f);
glEnable(GL_DEPTH_TEST);
glClear(GL_COLOR_BUFFER_BIT | GL_DEPTH_BUFFER_BIT | GL_STENCIL_BITS);
glDisable(GL_DEPTH_TEST); // disable depth test so screen-space quad isn't discarded due to depth test.
glColorMaski(0, GL_FALSE, GL_FALSE, GL_FALSE, GL_FALSE);
glStencilOp(GL_KEEP, GL_KEEP, GL_REPLACE);
glStencilFunc(GL_ALWAYS, 1, 0xFF); // all fragments pass the sten
glStencilMask(0xFF); // enable writing to the stencil buffer
shader.setMat4("model", glm::mat4(1.0f));
shader.setMat4("view", glm::mat4(1.0f));
shader.setMat4("projection", glm::mat4(1.0f));
glBindVertexArray(quadVAO);
glDrawArrays(GL_TRIANGLES, 0, 6);
This sets the stencil buffer for the mirror. Then I draw the scene while enabling the stencil test:
glEnable(GL_DEPTH_TEST); // enable depth testing (is disabled for rendering screen-space quad)
// make sure we clear the framebuffer's content
glColorMaski(0, GL_TRUE, GL_TRUE, GL_TRUE, GL_TRUE);
glClear(/*GL_COLOR_BUFFER_BIT |*/ GL_DEPTH_BUFFER_BIT);
glStencilOp(GL_KEEP, GL_KEEP, GL_KEEP);
glStencilFunc(GL_EQUAL, 1, 0xFF);
glm::mat4 model = glm::mat4(1.0f);
camera.Yaw += 180.0f; // rotate the camera's yaw 180 degrees around
camera.ProcessMouseMovement(0, 0, false); // call this to make sure it updates its camera vectors, note that we disable pitch constrains for this specific case (otherwise we can't reverse camera's pitch values)
glm::mat4 view = camera.GetViewMatrix();
camera.Yaw -= 180.0f; // reset it back to its original orientation
camera.ProcessMouseMovement(0, 0, true);
glm::mat4 projection = glm::perspective(glm::radians(camera.Zoom), (float)SCR_WIDTH / (float)SCR_HEIGHT, 0.1f, 100.0f);
shader.setMat4("view", view);
shader.setMat4("projection", projection);
// cubes
glBindVertexArray(cubeVAO);
glActiveTexture(GL_TEXTURE0);
glBindTexture(GL_TEXTURE_2D, cubeTexture);
model = glm::translate(model, glm::vec3(-1.0f, 0.0f, -1.0f));
shader.setMat4("model", model);
glDrawArrays(GL_TRIANGLES, 0, 36);
model = glm::mat4(1.0f);
model = glm::translate(model, glm::vec3(2.0f, 0.0f, 0.0f));
shader.setMat4("model", model);
glDrawArrays(GL_TRIANGLES, 0, 36);
// floor
glBindVertexArray(planeVAO);
glBindTexture(GL_TEXTURE_2D, floorTexture);
shader.setMat4("model", glm::mat4(1.0f));
glDrawArrays(GL_TRIANGLES, 0, 6);
glBindVertexArray(0);
This draws the view in the mirror and while commenting out the cleaing of the color buffer I get a strange behavior and I have to uncomment it for avoiding it but don’t know why (need explanation).
Finally I draw the scene outside the mirror using the negation of the above stencil test:
glClear(GL_DEPTH_BUFFER_BIT);
glStencilOp(GL_KEEP, GL_KEEP, GL_KEEP);
glStencilFunc(GL_NOTEQUAL, 1, 0xFF);
model = glm::mat4(1.0f);
view = camera.GetViewMatrix();
shader.setMat4("view", view);
// cubes
glBindVertexArray(cubeVAO);
glActiveTexture(GL_TEXTURE0);
glBindTexture(GL_TEXTURE_2D, cubeTexture);
model = glm::translate(model, glm::vec3(-1.0f, 0.0f, -1.0f));
shader.setMat4("model", model);
glDrawArrays(GL_TRIANGLES, 0, 36);
model = glm::mat4(1.0f);
model = glm::translate(model, glm::vec3(2.0f, 0.0f, 0.0f));
shader.setMat4("model", model);
glDrawArrays(GL_TRIANGLES, 0, 36);
// floor
glBindVertexArray(planeVAO);
glBindTexture(GL_TEXTURE_2D, floorTexture);
shader.setMat4("model", glm::mat4(1.0f));
glDrawArrays(GL_TRIANGLES, 0, 6);
glBindVertexArray(0);
Now here is an image of the strange behavior I get:
Note that the mirror has the color of the background and when I move the mouse traces of the 2 containers keep accumulating despite the glClear before drawing the mirror.
When I uncomment the GL_COLOR_BUFFER_BIT these traces disappear and I don’t understand why.
