XEffects - Reloaded - New Release (V 1.4)

[BlindSide]

No, you shouldn't need anything else. I see nothing wrong there.

Can you give me more info? Does it work when compiling the example app by itself?

[Prott]

No, the example application doesn`t work (same error). Other Irrlicht programs compiles succesfully.
I`m including only effectWrapper.h , other files are just inserted into project - maybe that could be wrong.

[BlindSide]

I will include an Example MSVC 2008 (It shouldn't be too much trouble updating, right? ) project with the next release. Hopefully that will sort out any issues like this. Expect it within a few days.

[Prott]

Ok, thanks. That would be very usefull. I`m looking forward to it ...

[Prott]

Problem solved - it was solved by turning runtime library into MT DLL.

[BlindSide]

Very interesting, I'll have to run some tests with static runtime library.

[nickle]

I have been playing with XEffects for a while. It is awesome.. just awesome, however, there are some potential revisions that should be taken in consideration.

1. In the function "removeShadowFromNode", the while loop statement should be changed to

Code: Select all

while(i < ShadowNodeArray.size() && ShadowNodeArray[i].node != node)

Original is

Code: Select all

while( ShadowNodeArray[i].node != node && i < ShadowNodeArray.size() )

The original one will cause memory leak when node is not in the shadow array and one is try to remove it.

2. The user defined clipping plane will also clip the screen quad. Is there a way to solve this problem?

Again... The work is brilliant. Thanks for the package, BlindSide.
[/code]

[BlindSide]

1. In the function "removeShadowFromNode", the while loop statement should be changed to

Hmmm, even better, I will change it to a binary search in the coming version.

2. The user defined clipping plane will also clip the screen quad. Is there a way to solve this problem?

Again... The work is brilliant. Thanks for the package, BlindSide.
[/code]

You can enable the clipplane while drawing the scene, and then disable it before the screenquad is drawn. You will have to edit effectWrapper.cpp yourself and disable the clipplane after smgr-drawAll(). Another alternative is that you can implement the clipplane yourself for every material in the pixel shader using EMT_TRANSPARENT_ALPHA_CHANNEL_REF as a base, but this is cumbersome.

Cheers

[nickle]

Cannot wait for the coming version... I will try the 1st method of the clipping... Thanks, genius

[BlindSide]

New release, check first post.

[Halifax]

Very nice release, I like the new features that I am seeing. Great job on the SSAO too!

By the way, I'm just curious, and I know nothing about it, but does that whole "assigning different shadow map resolutions to lights" help with PSSM? And are we going to see PSSM any time soon?

[BlindSide]

By the way, I'm just curious, and I know nothing about it, but does that whole "assigning different shadow map resolutions to lights" help with PSSM?

Not necessarily, the PSSM guys just use the same shadow map sizes for each split in the paper, it's the fact that the further away ones are spread over a larger area that affects the quality.

And are we going to see PSSM any time soon?

No comment. (By that I mean I just made a new release, get off my back!)

[Halifax]

No comment. (By that I mean I just made a new release, get off my back!)

Ah, okay, sorry, I was just wondering.

[m_krzywy]

Great work with SSAO. Love it <3

[9YkKsvXM]

-