Fast swapping framebuffers OpenGL - opengl

I recently read that simply switching the render targets of a framebuffer object is much faster than switching framebuffer object.
As extreme as it sounds, does this this mean I should only ever use one framebuffer object and only switchout it's targets?
EDIT: I changed 'swapping' to 'switching' to avoid confusion. By switching I mean binding a new framebuffer in place of the old one. Not to be confused with the SwapBuffers() call used to swap the front- and backbuffers.

EDIT: this answer is probably wrong. Read the comments below.
It's faster to switch framebuffer-attachable textures than switching between framebuffers (FBOs). More here http://www.songho.ca/opengl/gl_fbo.html
There are limits to how many attachments a FBO can have though.

Related

Can we sample texture attached to inactive FBO slot?

Can I attach two textures to one FBO, and switch between them using glDrawBuffers, binding the inactive one as shader input? This seems much more efficient than switching FBOs for multipass effects.
If we're assuming you don't have access to OpenGL 4.5/ARB/NV_texture_barrier, no you cannot. The part of the OpenGL specification that forbids feedback loops on framebuffer attached images does not care whether the image can be written to or not. This is also true for array layers or mipmap levels; reading from one layer while writing to another layer will not save you.
All that matters is attachment. You must either bind a new FBO that doesn't have the texture attached, or remove the attachment from the current FBO.
Though again, texture barrier functionality makes everything I said irrelevant. And considering how widespread it is, it's really not something you should be concerned about.

How to draw Renderbuffer as Texturebuffer in FBO?

I succeeded in render to texture with Texturebuffer, using VAO and shaders.
But FBO has another options for color buffer, it's Renderbuffer. I searched a lot on the internet, but cannot found any example related to draw Renderbuffer as Texturebuffer with shaders
If I ain't wrong, Renderbuffer is released in OpenGL 3.30, and it's faster than Texturebuffer.
Can I use Renderbuffer as Texturebuffer? (stupid question huh? I think it should be absolutely, isn't it?)
If yes, please lead me or give any example to draw render buffer as texture buffer.
My target is just for study, but I'd like to know is that a better way to draw textures? Should we use it frequently?
First of all, don't use the term "texture buffer" when you really just mean texture. A "buffer texture"/"texture buffer object" is a different conecpt, completely unrelated here.
If I ain't wrong, Renderbuffer is released in OpenGL 3.30, and it's faster than Texturebuffer.
No. Renderbuffers were there when FBOs were first invented. One being faster than the other is not generally true either, but these are implementation details. But it is also irrelevant.
Can I use Renderbuffer as Texturebuffer? (stupid question huh? I think it should be absolutely, isn't it?)
Nope. You cant use the contents of a renderbuffer directly as a source for texture mapping. Renderbuffesr are just abstract memory regions the GPU renders to, and they are not in the format required for texturing. You can read back the results to the CPU using glReadPixels, our you could copy the data into a texture object, e.g. via glCopyTexSubImage - but that would be much slower than directly rendering into textures.
So renderbuffers are good for a different set of use cases:
offscreen rendering (e.g. where the image results will be written to a file, or encoded to a video)
as helper buffers during rendering, like the depth buffer or stencil buffer, where you do not care anbout the final contents of these buffers anyway
as intermediate buffer when the image data can't be directly used by the follwoing steps, e.g. when using multisampling, and copying the result to a non-multisampled framebuffer or texture
It appears that you have your terminology mixed up.
You attach images to Framebuffer Objects. Those images can either be a Renderbuffer Object (this is an offscreen surface that has very few uses besides attaching and blitting) or they can be part of a Texture Object.
Use whichever makes sense. If you need to read the results of your drawing in a shader then obviously you should attach a texture. If you just need a depth buffer, but never need to read it back, a renderbuffer might be fine. Some older hardware does not support multisampled textures, so that is another situation where you might favor renderbuffers over textures.
Performance wise, do not make any assumptions. You might think that since renderbuffers have a lot fewer uses they would somehow be quicker, but that's not always the case. glBlitFramebuffer (...) can be slower than drawing a textured quad.

Techniques for drawing tiles with OpenGL

I've been using XNA for essentialy all of my programming so far and would like to move on to OpenGL (along with SFML for IO, creating the window etc.) with C++ . For starters I'd like to create a tile-based game and I've mostly looked at LazyFoo's tutorials.
I just have a two questions:
How should I draw the tiles? Should I use immediate drawing, arrays, VBOs or what? VBOs feel like overkill for this but I'm not sure. It's very tempting to use immediate drawing but apparently it's deprecated. Maybe it's fine for this purpose since it's 2D and only for a bunch of quads.
I'd like a lot of different tiles and thus all of my tiles will not fit into a single texture without making it massive. I've read that using bindTexture isn't very cheap and thus I should avoid as many calls as I can. I thought that maybe I can create a manager for my textures and stitch them all together into one big texture and bind that but then the dimensions of that is an issue.
Don't use immediate mode! It's cumbersome to work with and has been removed from recent OpenGL versions. Use Vertex Arrays, ideally through VBOs. In the end they're much easier to use, believe me.
Regarding that switching of textures. We're talking about optimizing the texture switch patterns in very complex scenes. In your case it will hardly matter at all.
Update
Right now you worry abount things without having even used them. That's worse than premature optimization. I suggest you first get a good grip on OpenGL, then start worrying about state switch management.
With regards to the texture atlas; this is usually done by stitching textures into groups of power-of-two sized textures. For example in a tile-based game you might have a particular tile set (say, tiles for an ice world) grouped together on 2 or 3 textures. When you want to render them you would determine what tiles are visible, then you bind each texture once and render the tiles from that texture for any tiles that are visible on screen.
This requires quite a lot of set-up time to get right; you need keep information on each sub-texture of the atlas so you can find the right texture and render the appropriate region of that texture whenever a tile is referenced. You also need a good way of grouping rendering operations so that they occur when the appropriate texture is bound.
Like datenwolf said, I wouldn't focus too much on complicated texture systems early on; eager binding of textures will be plenty fast enough until you get further down the road.

Optimizing OpenGL: VAOs vs Shader Organization

I'm writing some code that will be drawing a number of 3D models to the screen, ultimately for a sort of 3D world. Each model could have meshes within it that use different shaders for rendering. In terms of making things run efficiently, I know that you should try to minimize OpenGL calls, namely those switching VAO and those switching shaders as they're supposed to have a lot of overhead.
In my implementation, each model has its geometry organized into a VAO. VAOs can have meshes that will differ in material settings but not necessarily shaders.
My question: Is it better to bind a VAO and render it entirely in one go, switching shaders if needed along the way, or to bind one shader and run through all of the (subsets of the ) VAOs that use that question. The first would be redundantly binding shaders, the second would be redundantly binding VAOs, so is it better to minimize shader binding calls or VAO binding calls?
I realize this is rather open ended and seems like it could depend on the application, however I would also be interested to know how I could find out how much gl operations cost (overhead? dependencies?), or how I could figure them out, as that would provide the information necessary to construct a proper response for whatever the scenario is. Also, if there's a better way to organize my project to begin with, that insight would be greatly appreciated.

What is the point of an SDL2 Texture?

I'm kind of stuck on the logic behind an SDL2 texture. To me, they are pointless since you cannot draw to them.
In my program, I have several surfaces (or what were surfaces before I switched to SDL2) that I just blitted together to form layers. Now, it seems, I have to create several renderers and textures to create the same effect since SDL_RenderCopy takes a texture pointer.
Not only that, but all renderers have to come from a window, which I understand, but still fouls me up a bit more.
This all seems extremely bulky and slow. Am I missing something? Is there a way to draw directly to a texture? What are the point of textures, and am I safe to have multiple (if not hundreds) of renderers in place of what were surfaces?
SDL_Texture objects are stored as close as possible to video card memory and therefore can easily be accelerated by your GPU. Resizing, alpha blending, anti-aliasing and almost any compute-heavy operation can harshly be affected by this performance boost. If your program needs to run a per-pixel logic on your textures, you are encouraged to convert your textures into surfaces temporarily. Achieving a workaround with streaming textures is also possible.
Edit:
Since this answer recieves quite the attention, I'd like to elaborate my suggestion.
If you prefer to use Texture -> Surface -> Texture workflow to apply your per-pixel operation, make sure you cache your final texture unless you need to recalculate it on every render cycle. Textures in this solution are created with SDL_TEXTUREACCESS_STATIC flag.
Streaming textures (creation flag is SDL_TEXTUREACCESS_STREAMING) are encouraged for use cases where source of the pixel data is network, a device, a frameserver or some other source that is beyond SDL applications' full reach and when it is apparent that caching frames from source is inefficient or would not work.
It is possible to render on top of textures if they are created with SDL_TEXTUREACCESS_TARGET flag. This limits the source of the draw operation to other textures although this might already be what you required in the first place. "Textures as render targets" is one of the newest and least widely supported feature of SDL2.
Nerd info for curious readers:
Due to the nature of SDL implementation, the first two methods depend on application level read and copy operations, though they are optimized for suggested scenarios and fast enough for realtime applications.
Copying data from application level is almost always slow when compared to post-processing on GPU. If your requirements are more strict than what SDL can provide and your logic does not depend on some outer pixel data source, it would be sensible to allocate raw OpenGL textures painted from you SDL surfaces and apply shaders (GPU logic) to them.
Shaders are written in GLSL, a language which compiles into GPU assembly. Hardware/GPU Acceleration actually refers to code parallelized on GPU cores and using shaders is the prefered way to achieve that for rendering purposes.
Attention! Using raw OpenGL textures and shaders in conjunction with SDL rendering functions and structures might cause some unexpected conflicts or loss of flexibility provided by the library.
TLDR;
It is faster to render and operate on textures than surfaces although modifying them can sometimes be cumborsome.
Through creating a SDL2 Texture as a STREAMING type, one can lock and unlock the entire texture or just an area of pixels to perform direct pixel operations. One must create prior a SDL2 Surface, and link with lock-unlock as follows:
SDL_Surface surface = SDL_CreateSurface(..);
SDL_LockTexture(texture, &rect, &surface->pixels, &surface->pitch);
// paint into surface pixels
SDL_UnlockTexture(texture);
The key is, if you draw to texture of larger size, and the drawing is incremental ( e.g. data graph in real time ) be sure to only lock and unlock the actual area to update. Otherwise the operations will be slow, with heavy memory copying.
I have experienced reasonable performance and the usage model is not too difficult to understand.
In SDL2 it is possible to render off-screen / render directly to a texture. The function to use is:
int SDL_SetRenderTarget(SDL_Renderer *renderer, SDL_Texture *texture);
This only works if the renderer enables SDL_RENDERER_TARGETTEXTURE.