Phaser: [Discussion] Better way for Multi-pass rendering

[jcyuan]

I'm doing post efx, but I found if I need to use the camera buffer, I have to call camera.setRenderToTexture, and if with some shader efx, I have to create a new pipeline which actually takes a large block of memory (for vertex batching) which is unnecessary.

and, if I need to create 2 texture for multi-pass ping-pong rendering, I can't find a better way to do this (or I don't know how)
if I use rendertexture, this one can only draw things on it without shader.
if I use shader, this one actually is a quad display object, so I can't reuse this with the renderer in the same frame ticking.

and, if I use camera.setRenderToTexture() without a new pipeline to have my final framebuffer, what is the right next step to apply my shaders with it (for example in a handler function of camera's postRender event)?

my current solution is in the camera's postRender event use native gl code to handle the rest steps... obviously that's a bit awkward.

[photonstorm]

I can finally close this thanks to the new post-pipeline and render target support available in 3.50 Beta 10 :)

For more details see: https://www.patreon.com/posts/43706661 (I know you've seen this link, just posting it here for anyone else)