Beta Testing 2.07

[duzenko 564]

 

Did you have a chance to take a look at it? When I use maskDepth in the program stage, it cancels the refraction effect. Maybe I'm doing something wrong?

Do you have a test map/quick save?

[peter_spy 1355]

Nope, that's a 1,5 gig WIP project. But, you can apply textures/darkmod/glass/clear_warp material on a brush to see that. Just remember to use larger scale, like 1.

[duzenko 564]

Nope, that's a 1,5 gig WIP project. But, you can apply textures/darkmod/glass/clear_warp material on a brush to see that. Just remember to use larger scale, like 1.

For that I would need DR that I don't have and clueless about anyway

[peter_spy 1355]

 

For some reason that effect vanishes when you turn on in-game AA:

 

The material code is as follows:

textures/darkmod/glass/clear_warp
{

	// Written by Tels, mainly based on 
	// http://www.iddevnet.com/doom3/materials.php

	noSelfShadow
	noshadows
	twoSided
	translucent
	forceoverlays
	sort decal
	glass
	qer_editorimage	textures/darkmod/glass/clear_warp_ed

	description "A reflecting, translucent material suited for large, flat glass panes"

    // this block makes the glass distort what you see through it
    // one unwanted side-effect is that this effect does not stack
	// and distorbs objects in front of the glass, too
    {
                vertexProgram heatHazeWithDepth.vfp
                vertexParm  0  0 , 0  // texture scrolling
                vertexParm  1  .8  // magnitude of the distortion

                fragmentProgram heatHazeWithDepth.vfp
                fragmentMap 0  _currentRender
                fragmentMap 1  textures/sfx/vp1 // the normal map for distortion
                fragmentMap 2 _currentDepth
				maskDepth
    }

	// this stage sets up the alpha mask (it doesnt render anything)
	// we use another glass texture for the alpha map, so that the reflection
	// shows more clearly and different from what the alpha map of our glass
	// says, that makes it more interesting 
	{
		maskcolor				
		map makealpha(textures/glass/glass1)
	}
	// now use the alpha mask to add some fake reflections
	{
		blend gl_dst_alpha, gl_one
		maskalpha
		cubeMap env/gen2
		// tone down the reflection a bit
		red     Parm0 * .2
		green   Parm1 * .2
		blue    Parm2 * .2
		texgen  reflect
	}
	// add our texture on top
	{
		blend   gl_dst_color, gl_one
		map	textures/darkmod/glass/crystal_milky
		red     Parm0 * 0.2
		green   Parm1 * 0.2
		blue    Parm2 * 0.2
	}

	// This is the code required for frob highlighting this texture
	{
		if ( parm11 > 0 )
			blend       gl_dst_color, gl_one
			map         _white
			rgb         0.40 * parm11
   	}
	{
		if ( parm11 > 0 )
			blend       add
			map         textures/darkmod/glass/crystal_milky
			rgb         0.15 * parm11
	}


}

[duzenko 564]

Any chance I could have a map to test this?

[VanishedOne 479]

delete your tdm_base01.pk4 and re-run tdm_updater

Still not fixed, I'm afraid. It got slightly better after I also deleted the config file again, but slightly better is still very broken.

[peter_spy 1355]

Any chance I could have a map to test this?

 

https://we.tl/t-yGJAWpZuqg

[stgatilov 652]

Still not fixed, I'm afraid. It got slightly better after I also deleted the config file again, but slightly better is still very broken.

Could you please do the following:

 

1) Provide the contents of your darkmod.cfg.

 

2) Show how exactly it looks. Including the "press attack to start" screen.

Video would be perfect (e.g. record with OBS). Set of screenshots would be OK. If the issue is not present on screenshots, take at least some photos.

 

3) Try to toggle various parameters.

I see you already tried r_useFBO 0. I suggest setting r_fboResolution 1, disable soft shadows and all experimental settings, set shadows to "Stencil".

Also try toggling ambient/interaction shaders (simple/enhanbed), switch bloom off. Disable antialiasing just in case.

If nothing of this helps, then try to disable parts of lighting:

r_skipInteractions 1 --- completely disables interactions (i.e. lighting be lamps, fires, etc.)

r_skipAmbient 3 --- completely disables ambient light (sort of global and scene-baked light)

If one of this removes crazy flickering, please tell us which one.

[duzenko 564]

 

https://we.tl/t-yGJAWpZuqg

This is how it looks for me on Intel (practically OK) but I guess it's broken for you on nVidia because, as stated above, the glass surface writes to depth buffer at the same time as it reads which Opengl considers undefined behavior.

 

Since I can't repeat it on my side due to implementation specific behavior I have to rely on you for experimenting.

 

How does it look with and without the maskDepth for you? I was wrong - it's actually a stage modifier (although it does not make much sense for me)

 

Also, try to make the material front/back faces only. It seems to be both sides currently.

[peter_spy 1355]

MaskDepth doesn't change a thing on my nvidia card, it looks like this:

 

 

 

Also, try to make the material front/back faces only. It seems to be both sides currently.

 

This is one sided brush, but this glass material does make it two sided by default (and without twosided keyword).

[VanishedOne 479]

Well, this is weird: not only could I not capture the problem in screenshots, it actually vanishes when OBS has TDM as a source. I'll have to use my 'phone camera instead.

 

r_skipInteractions 1 and r_skipAmbient 3 both remove the glitches.

 

Generated Darkmod.cfg is attached.

 

The problem seems to manifest itself mainly near (some) lights. In this map, when I move near the hanging oil lamp this happens: link

 

The HUD is also affected: link

Darkmod.cfg.txt

[stgatilov 652]

I quickly booted up my missions to see any changes. I noticed both are super dark even with the gamma and brightness turned up?

I have all the graphic bells and whistles turned to OFF It looks like when I turn ambient lighting model to simple the ambient works but in advanced mode, the renderer doesn't seem to use the ambient lighting at all

Could you also say what's the value of image_useNormalCompression cvar?

 

I have briefly looked at the changes since 2.06, and there are quite a lot of them.

I'm afraid that substantial help is needed from your side.

 

I have attached an instrumented version of ambientInteraction.fs.

I'm pretty sure that you know how to override stock shader with this one (replace it inside pk4 / uncompress one pk4 and change the file / etc).

The file has several #if 0 and #if 1 blocks, each one has a comment at the end of the line.

With their current values, the shader works normally for me (at least at the beginning of St.Lucia map).

 

I suggest going through all the #if-s. For each of them, do the following:

1) Toggle the digit (change 0 to 1 or vice versa).

2) Execute reloadGLSLPrograms in console.

3) Check if the picture has changed, and how it has changed.

4) Toggle the digit back.

5) Proceed to next #if.

Then write back exactly which changes made a difference and what was the difference.

If there are no changes, then the whole effort is useless

 

This is rather bruteforce approach, but I would definitely do it myself if I could reproduce the problem.

Probably Duzenko would suggest something less stupid.

ambientInteraction.fs.txt

[stgatilov 652]

r_skipInteractions 1 and r_skipAmbient 3 both remove the glitches.

Just to clarify: are the glitches removed when you set one of these (either one, in both cases issue goes away), and only when you set both of these?

 

The HUD is also affected: link

I see the White Letters of Glitch here

I have always wondered what they are: I have seen that often in the part related to other FBO bugs.

 

Also could you please try again (just in case) to disable FBO.

Note that you have to restart game after setting r_useFBO 0, and be sure to check that it is still zero after restart.

This cvar is sometimes hard to disable

[nbohr1more 1887]

Well, this is weird: not only could I not capture the problem in screenshots, it actually vanishes when OBS has TDM as a source. I'll have to use my 'phone camera instead.

 

r_skipInteractions 1 and r_skipAmbient 3 both remove the glitches.

 

Generated Darkmod.cfg is attached.

 

The problem seems to manifest itself mainly near (some) lights. In this map, when I move near the hanging oil lamp this happens: link

 

The HUD is also affected: link

 

Hmm I thought that r_fboSeparateStencil was created for nvidia bugs like this. Does the behavior go away when set to 1?

[VanishedOne 479]

No luck with r_fboSeparateStencil. (It's an AMD chipset, incidentally.)

 

Just to clarify: are the glitches removed when you set one of these (either one, in both cases issue goes away), and only when you set both of these?

 

I see the White Letters of Glitch here

I have always wondered what they are: I have seen that often in the part related to other FBO bugs.

 

Also could you please try again (just in case) to disable FBO.

Note that you have to restart game after setting r_useFBO 0, and be sure to check that it is still zero after restart.

This cvar is sometimes hard to disable

 

When I initially tried those cvars in a test map, it seemed either would remove all the glitching I was seeing. However, I just went and stood near a light source in a map with a 0 0 0 global ambient, and only r_skipInteractions 0 had any effect.

 

r_useFBO 0 + restart doesn't remove the glitching.

Edited December 21, 2018 by VanishedOne

[nbohr1more 1887]

Hmmm...

 

Let's try forcing off fence sync. You'll lose FPS but it may reveal an issue with VBO behavior:

 

r_useFenceSync 0

[VanishedOne 479]

Hmmm...

 

Let's try forcing off fence sync. You'll lose FPS but it may reveal an issue with VBO behavior:

 

r_useFenceSync 0

Ah, now we're getting somewhere. No more glitching so far.

[nbohr1more 1887]

(It's an AMD chipset, incidentally.)

If I googled this correctly:

 

OpenGL version: 4.3.12615 Compatibility Profile Context 13.250.31.0

This appears to be a driver from 2014?

 

Have you tried getting the latest Adrenaline drivers?

 

https://www.amd.com/en/support/previous-drivers/apu/amd-series-processors/amd-a6-series-apu-for-desktops/a6-5200-radeon-hd-8400

[duzenko 564]

Ah, now we're getting somewhere. No more glitching so far.

Heh, I remember @cabalistic wtf'ing me for adding that cvar, LOL

 

 

Hmm I thought that r_fboSeparateStencil was created for nvidia bugs like this. Does the behavior go away when set to 1?

That trick only works on Intel and now even I don't turn it on

 

 

This appears to be a driver from 2014?

AMD has a habit of screwing the unsuspecting buyers of the mobile IGP's when it comes to driver support

Really shitty habit of pushing the noncompetitive hardware and then cutting driver support due to poor sales, now in its Nth iteration.

[VanishedOne 479]

Okay: the installer froze my laptop so I ended up rebooting and can't be sure it installed cleanly, but (1) I have a 2016 driver date in the Device Manager, and (2) I can't reproduce the glitching any more. Cheers.

[nbohr1more 1887]

@Judith: Try this one

 

{    
surftype15
    description glass
	noSelfShadow
    noshadows
	translucent
	forceoverlays
    sort decal

	qer_editorimage		textures/do/mesh/ar_cor_window_d
	bumpmap				textures/do/mesh/ar_cor_window_n
    diffusemap       	textures/do/mesh/ar_cor_window_d
    specularmap			textures/do/mesh/ar_cor_window_s
	
	
    {
        vertexProgram heatHazeWithDepth.vfp
        vertexParm  0  0 , 0 // texture scrolling
        vertexParm  1  1  // magnitude of the distortion

        fragmentProgram heatHazeWithDepth.vfp
        fragmentMap 0  _currentRender
        fragmentMap 1  textures/do/mesh/glass_warp01_n // the normal map for distortion
        fragmentMap 2 _currentDepth
        maskDepth
   }
      
       

[peter_spy 1355]

Already tried that, it doesn't work. MaskDepth is also present in the clear_warp material, and it doesn't work either.

[Spooks 643]

I've reopened #4686. Lest you forget, I had already provided a test map in there. r_fboseperatestencil plainly disables r_useFBO and prevents it from being on. In-game AA is the only way for the heathaze materials to appear correct with FBO on.

[duzenko 564]

I have an issue with my WIP in 2.07.

 

Here's the scene using 2.06:

 

206.jpg

 

Here's the scene using 2.07:

 

207.jpg

 

Looks like a lighting problem.

I think I fixed it and merged to the 2.07 branch but someone need to test the merge result

[duzenko 564]

I've reopened #4686. Lest you forget, I had already provided a test map in there. r_fboseperatestencil plainly disables r_useFBO and prevents it from being on. In-game AA is the only way for the heathaze materials to appear correct with FBO on.

Not exactly "disables" FBO but your driver does not support it and thus TDM falls back to the non-FBO path.