- Each canvas (and the screen) has an invisible stencil value associated with each pixel. Stencil values are between [0, 255].
- love.graphics.stencil takes a function argument which draws things, and makes the geometry of what's drawn set the stencil values of pixels instead of coloring the pixels.
- Its second argument is a 'stencil action' which determines what happens to the stencil values of pixels. The possible stencil actions are 'replace' (the default), 'invert', 'increment', 'decrement', 'incrementwrap', and 'decrementwrap'. If the stencil action is 'replace', the third argument of love.graphics.stencil is a number between [0, 255] which determines what the stencil value of each pixel that touches drawn objects is replaced with.
The fourth argument is a boolean which determines whether the previous stencil values of all pixels on-screen should be kept (true), or cleared to 0 (false, the default.) love.graphics.clear also clears the stencil values.
- love.graphics.setStencilTest allows for anything drawn to be affected by a comparison between the arguments to setStencilTest and the existing stencil value of the pixels that are touched by what's drawn. Its first argument is a compare mode which can be 'equal', 'notequal', 'less', 'lequal', 'gequal', or 'greater', and the second argument is an integer value (between 0 and 255) used in the comparison with the value in the stencil buffer.
For example, love.graphics.setStencilTest("greater", 0) will cause any geometry drawn afterwards to only appear on pixels whose stencil value is greater than 0.
Gamma-correct blending and shader math can be enabled globally via the new 't.gammacorrect' boolean flag in love.conf.
The new function love.graphics.isGammaCorrect will return true if it was requested in love.conf and is supported on the system.
When gamma correct rendering is enabled, colors (including the colors of pixels from images) are automatically converted from sRGB to linear RGB before use. When drawing to the main screen or to a canvas with the 'normal' or 'srgb' format, the final output colors of pixel shaders are automatically converted from linear RGB to sRGB after blending and before the color is stored in the pixel.
This lets the rendering pipeline do math using linear RGB values for colors rather than sRGB values, so the math is correct, without making users of the APIs manually linearize their colors with the love.math.gammaToLinear function (which still exists). The final output of the screen is encoded as sRGB, which is what systems expect. Canvases (except when otherwise requested) store their contents with sRGB encoding for increased precision with darker colors.
the 'srgb' window setting flag has been removed, as well as the 'srgb' image flag. A new image flag 'linear' has been added, which when set to true will cause the colors of the image to always be treated as linear RGB rather than sRGB, when gamma-correct rendering is enabled.
A new function 'Shader:sendColor' has been added, which has the same argument structure as 'Shader:send' but expects colors in the range of [0, 255]. When gamma-correct rendering is enabled it automatically gamma-corrects the given colors (by applying gammaToLinear to them.)
New shader code functions have been added: gammaToLinear, linearToGamma, gammaCorrectColor, and unGammaCorrectColor. When gamma-correct rendering is enabled, the LOVE_GAMMA_CORRECT #define is set and gammaCorrectColor and unGammaCorrectColor are aliases for gammaToLinear and linearToGamma respectively, otherwise the functions do nothing.
The new shader functions have 'precise' and 'fast' variants. If the LOVE_PRECISE_GAMMA define is set, then the normal functions default to the precise variants, otherwise they default to the fast variants. Currently the define is set in vertex shaders and not set in pixel shaders.
The default per-vertex color is automatically gamma-corrected by LÖVE when gamma correct rendering is enabled, but any custom named per-vertex attribute specified with the new custom attribute Mesh functionality won't be, unless 'gammaCorrectColor' or similar functions are explicitly used (preferably inside the vertex shader code that has the custom attribute.)
Removed the 'premultiplied' blend mode, since setBlendMode("alpha", false) accomplishes the same thing now.
The 'multiply' blend mode is unaffected by the new flag, since fixed-function hardware blending can't do the equivalent of setBlendMode("multiply", true).
Text objects are Drawable objects that represent text on the screen. They decouple text drawing from text parsing and vertex uploading, in a similar manner to SpriteBatches. They also optionally allow efficient batching of text from the same Font in different relative coordinate transformations, without having many draw calls.
Text objects currently have the following methods:
Text:set(textstring) -- replaces any existing text in the object with the new string.
Text:setf(textstring, wraplimit, alignmode) -- replaces any existing text with formatted text using the new string.
Text:getFont() -- gets the Font object used for this Text object.
Text:getWidth() -- gets the width in pixels of the most recently added text in the Text object.
Text:getHeight() -- gets the height in pixels of the most recently added text in the Text object.
Text:add(textstring, x, y, angle, sx, sy, ox, oy, kx, ky) -- adds a new string to the text object using the specified relative coordinate transformation, without replacing existing text. This behaves much like SpriteBatch:add.
Text:addf(textstring, wraplimit, alignmode, x, y, angle, sx, sy, ox, oy, kx, ky) -- adds a new formatted string to the text object using the specified relative coordinate transformation.
Text:clear() -- clears all text from the object.
--HG--
branch : minor
drawcalls: The number of internal draw calls made so far in the current frame.
canvasswitches: The number of times the active Canvas has been changed so far in the current frame.
texturememory: The approximate amount of video memory (in bytes) used by OpenGL textures created by Images, Canvases, and Fonts.
images: The number of active Image objects.
canvases: The number of active Canvas objects.
fonts: The number of active Font objects.
Draw to the stencil buffer (rather than the color buffer) using love.graphics.drawStencil with a function which draws geometry. Once the stencil buffer has been drawn to, use love.graphics.setStencilTest before drawing to let drawn geometry be affected by it.
love.graphics.clear (and Canvas:clear) will clear both the color and stencil buffers by default. love.graphics.clear("stencil") will clear only the stencil buffer.
--HG--
branch : stencil-functions
Removed the 'hdrcanvas', 'dxt', and 'bc5' graphics feature enums now that they're deprecated in 0.9.x in favour of getCanvasFormats and getCompressedImageFormats.
--HG--
branch : minor
The two new functions return tables containing the names of all relevant formats as keys, and boolean values indicating whether the formats are supported.
Removed love.graphics.isSupported("canvas", "shader", "npot", "subtractive", and "mipmap"), since those features are all guaranteed by the minimum system requirements.
--HG--
branch : minor
love.graphics.newImage(path, “srgb”) creates a new Image whose texels are treated as being in the sRGB color space, so they are linearized when drawing/sampling from the image.
love.graphics.newCanvas(w, h, “srgb”) creates a new Canvas whose texels are treated as being in the sRGB color space, so drawing to the Canvas does a linear->sRGB conversion (but blends linearly), and sampling from it (drawing it or using it in a shader) converts from sRGB to linear space.
The "srgb" window flag does the same as Canvases for the main screen.
love.graphics.draw(mesh) will draw the mesh instancecount times, using hardware instancing when available. The only way to draw individual instances differently from each other is to use the love_InstanceID variable in a vertex shader.
Added love.graphics.isSupported(“instancing”). Hardware instancing is supported if true, otherwise a (slower) pseudo-instancing fallback is used internally when drawing instanced meshes.
Added:
love.graphics.setLineJoin(linejoin)
linejoin = love.graphics.getLineJoin()
where linjoin is one of:
- none
- miter
- bevel
Bevel and miter drawing is slightly faster than the 0.8 default (miter), while
the `none` join mode will be slower when overdraw is enabled.
- added a new CompressedData type to love.image
- added love.image.isCompressed(file or data)
- love.graphics.newImage automatically tests for compression when loading a file
- added love.graphics.isSupported strings for DXT, BC5, and BC7
Compressed textures meant to be used on the GPU offer huge performance gains when
- loading the texture from a file
- loading the texture from RAM to VRAM
- loading mipmaps
- rendering
--HG--
branch : image-CompressedData
- Simultanious rendering to multiple canvases requires all canvases to be the same size (limitation of pre-GL3.0 framebuffers)
- love.graphics.isSupported("mrtcanvas") was added
- love_Canvases[n] (0-based array of active canvases) can be written to in pixel shaders in the 'effects' callback function
- everything drawn to multiple canvases at once will be duplicated to all canvases if no shader is active or the standard 'effect' shader function is used
--HG--
branch : MRTs
Alex Szpakowski
vrld
Alexander Szpakowski
Bart van Strien