问题描述
应该使用哪些Direct3D渲染状态来实现Java的Porter-Duff合成规则(CLEAR，SRC，SRCOVER等)?
What Direct3D render states should be used to implement Java's Porter-Duff compositing rules (CLEAR, SRC, SRCOVER, etc.)?
推荐答案
我使用Java的程度不是很高，但是基于白皮书，它应该是渲染状态混合模式的相当简单的映射.
I'm haven't used Java too much, but based on the white paper from 1984, it should be a fairly straightforward mapping of render state blend modes.
当然，除了这些之外，您还可以做更多的事情，例如普通的alpha混合(SourceAlpha，InvSourceAlpha)或加性(One，One)等等. (我假设您是在问这些问题，因为您正在移植某些现有功能?因此您可能不在乎其他组合...)
There are of course more that you can do than just these, like normal alpha blending (SourceAlpha, InvSourceAlpha) or additive (One, One) to name a few. (I assume that you are asking about these specifically because you are porting some existing functionality? In that cause you may not care about other combinations...)
无论如何，这些都假定Add的BlendOperation且AlphaBlendEnable为true.
Anyway, these assume a BlendOperation of Add and that AlphaBlendEnable is true.
清除
SourceBlend = Zero
DestinationBlend = Zero
A
SourceBlend = One
DestinationBlend = Zero
B
SourceBlend = Zero
DestinationBlend = One
A胜过B
SourceBlend = One
DestinationBlend = InvSourceAlpha
B超过A
SourceBlend = InvDestinationAlpha
DestinationBlend = One
A in B
SourceBlend = DestinationAlpha
DestinationBlend = One
A中的B
SourceBlend = Zero
DestinationBlend = SourceAlpha
A out B
SourceBlend = InvDestinationAlpha
DestinationBlend = Zero
B出A
SourceBlend = Zero
DestinationBlend = InvSourceAlpha
A在B之上
SourceBlend = DestinationAlpha
DestinationBlend = InvSourceAlpha
A上的B
SourceBlend = InvDestinationAlpha
DestinationBlend = SourceAlpha
A xor B
SourceBlend = InvDestinationAlpha
DestinationBlend = InvSourceAlpha
束缚这些稍微复杂一点，并且需要向着色器多次传递或多次输入纹理.
Chaining these is a little more complex and would require either multiple passes or multiple texture inputs to a shader.
这篇关于在Direct3D中实施Porter-Duff规则的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！