Talk:Digital compositing

Untitled

This is a very technical explanation. I'd like to help put it into plain English, but don't want to get anything wrong. Howabout, for example: "It relies on replacing parts of one image with another. This is achieved by making transparent any part of the image which is a specified colour (e.g. green, blue, or black). This allows a second image to 'show through', creating a composite." --Rayray 11:34, 23 November 2005 (UTC)[reply]

Merging

In my opinion, there ought to be (at least) two articles. One (this page) should be about compositing in relation to filmmaking, covering practical techniques and products. Another, currently alpha compositing, should focus on the theory (Porter-Duff, etc.), mathematics, and its relation to software and computer graphics. The audience of the first page would be the layman/filmmaker, and that of the second page a mathematician/computer scientist. Perhaps we should work to move the math to alpha compositing. --IanOsgood 04:02, 20 July 2006 (UTC)[reply]

The article had an error in one variable name, which I fixed in the Mathematics section. By the way, I tried to fix this section which in first place has an incorrect title, "Algorithm for alpha-compositing" would be more descriptive. That section is not clear and has more mistakes. The section is about how to control the way in which the mathematical formula F=A*B*C*D*E can be evaluated. Some parts seem unnecessarily, for example: although it depends on the programming language semantics, the statement: F:=(A*B)*(C*D) is performed by the majority of compilers in the same order as {S:=A*B;F:=S*(C*D)}. Although it is true that a parallel computer can perform each parenthesized expression in different processors, there are different parallel computer architectures, that should be taken into account when the algorithm is developed.

That section describes how the formulas are transformed to optimize the calculation, but it does not explain which kinds of alpha-compositing exists, suddenly it starts to talk about the algorithm instead of explaining which kinds of alpha-compositing exists to use those definitions to explain how the algorithm is developed according to the computer architecture.

Until I entered here to talk about such modification, I saw your comment, and after a very quick and superficial sight, it seems that the other article contains a better explanation. For that reason I agree the merging of the articles, or even the deletion of this one if it is contained in the other, which it seemed to my after a very very superficial sight.

— Preceding unsigned comment added by 189.178.46.237 (talk) 08:16, 1 February 2015‎ (UTC)[reply]

This article shouldn't exist. Merge with Compositing. Create a section entitled "Digital" there, if you'd like, as that would be most appropriate.Clepsydrae (talk) 04:01, 4 January 2019 (UTC)[reply]

γ

γ is introduced without a description equivalent to the description for α, being "the opacity value of the foreground pixel" Hackwrench 07:22, 31 July 2006 (UTC)[reply]

new compositing article

The article on compositing has been replaced with new material. Jim Stinson 23:16, 10 August 2007 (UTC)[reply]

Nonlinear Light and Aliasing

From the Article:

Note that if the operations are performed in a colour space where γ is not equal to 1 then the operation will lead to non-linear effects which can potentially be seen as aliasing artifacts (or 'jaggies') along sharp edges in the matte.

I very much doubt that linear light and gamma have much to do with aliasing here. How could this possibly lead to visibly jagged lines after compositing? Even if we were talking about rendering, I would expect that other effects (eg. light sources appearing too dark) will be much more visible. Also, the uninformed reader is tempted to assume that linear light is the most common choice for computer graphics. It should be made clear that this is not the case. --Matumio (talk) 18:08, 23 September 2008 (UTC)[reply]