Talk:Parallax mapping

The original stinks of machine translation, perhaps from a copyrighted source. Please look at the old versions of Offset Mapping for more. --Cmprince 03:51, 24 Dec 2004 (UTC)

Wouldn't it be better to merge all rendering techniques that gives 3D depth to textures in one article? —Preceding unsigned comment added by 201.81.199.100 (talk) 13:08, 29 October 2007 (UTC)[reply]

Moved the article to "Parallax mapping"

I wrote Terry Welsh, the author of Parallax Mapping with Offset Limiting: A PerPixel Approximation of Uneven Surfaces (which is the first result that appears when you search for parallax mapping on Yahoo), for permission to use the images in that PDF to improve this article. He agreed, and said that we should get rid of the name offset mapping, since offset mapping was just something that he incorrectly called it before he thought of the name parallax mapping and the name offset mapping only seems to confuse people any more. So I've moved it, fixed all redirects, etc. This also means that soon this article will be all prettied-up with images from that PDF (thank you again if you read this, Terry).Tommstein 08:59, 30 October 2005 (UTC)[reply]

Relief Mapping

I heard that F.E.A.R. uses relief mapping, not parallax mapping. Would someone investigate this? Also, what is the difference between relief mapping and parallax?--68.100.150.117 00:09, 16 February 2006 (UTC)[reply]

Parallax mapping is somewhat more dinamic and less performance hungry. It also gives far more issues, especially at steep angles. Relief mapping is based on pre-warping the textures in a compute-intensive way, but also provides some self-shadowing.
I hardly believe a game could use it: it simply does not look attractive.
MaxDZ8 talk 08:12, 1 August 2006 (UTC)[reply]

Parallax mapping don't store alpha channel in "bump" texture - thats seems only difference. Parallax mapping can have shadows from long objects and there would not be difference if object is flat and long (big, like big flat plane). Weakness of parallax mapping compare with bump mapping is that it can be made only from high polygon models and not from photographs. So Parallax mapping if used on the ground or walls without end, then impossible to say difference compare with high polygon model wall. Only if you will reach wall end then see that edge of wall is made of two flat planes without holes between stones or bricks (or if you look at 0 degrees angle, if assuming wall is almost infinitly long; of course tree can't grew from wall, but some long cone will not make difference if you do not standing in place of it and if angle is at least 1 degree and wall is hundreds meters long). — Preceding unsigned comment added by Versatranitsonlywaytofly (talk • contribs) 17:08, 5 July 2011 (UTC)[reply]

Patent

The old patent link worked, the one with the patent template doesn't. I've removed the current patent link, hopefully someone can fix it or alternativly link to the old link --80.56.0.49 22:50, 7 June 2006 (UTC)[reply]

The patent doesn't seem to describe anything different from what Gamecube was capable of. It wasn't much more than a specific (custom) hardware implementation of the "Emboss Style Bump mapping" that has been around since ~DirectX6. Nintendo probably just refreshed it for the Wii. (Assuming the patent document's URL doesn't change) Here's a cache of the old one.

And here's the DirectX equivalent Swapnil 404 (talk) 04:33, 9 November 2009 (UTC)[reply]

Parallax mapping may be used in GameCube game "Wave Race: Blue Storm" (2001) for water effects. Also parallax mapping possible used in 3Dmark 2000 water effects, but there can be DirectX 7.0 CPU programing effects (no pixels or vertex shaders at that time, only texture drawing...). — Preceding unsigned comment added by Versatranitsonlywaytofly (talk • contribs) 17:24, 5 July 2011 (UTC)[reply]

Sample renderings

Some sample renderings with comparisons VS plain bumpmapping which are completely WP-compatible are available on a prototype for new page version. I never had the time to complete it however...
MaxDZ8 talk 10:10, 2 October 2007 (UTC)[reply]

More detail

Heres more detail maybe able to put here.

The effect uses typical stretching, similar to that of the taking of a row of pixels and using it for the next rows, as well as squishing. It uses a height to determine the amount, lighter is more, darker is less or begins to squish. Lastly the cube map which is applied to all is added changes it per cube map pixel, visable under low resolution, this is the actual basis of the parallax or perspective. BobtheVila —Preceding comment was added at 17:28, 25 February 2008 (UTC)[reply]

I don't get it that way. I don't see this squishing thing, I don't see where the cube map comes in and I find debatable to suggest an heightmap representation. Could you please re-phrase it? Thank you!
MaxDZ8 talk 08:00, 26 February 2008 (UTC)[reply]