New Release D3D9Client Development

[Marg]

Specs are typical Win7, GeForce-GTX760 2GB, i5 processor...
Everything else works with no problems (GTA V at high settings).
BUT... I reinstalled, and managed to get some environment. but after beginning to experiment with atmosphere sliders (because colors are weird - dark, high contrast), error happens (it is a long list, but part of it is):

000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00146906], angle=0, y=1
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.0014548], angle=1.92, y=1.00056
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00141241], angle=3.84, y=1.00225
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00134307], angle=5.76, y=1.00507
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00124871], angle=7.68, y=1.00905
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00113203], angle=9.6, y=1.0142
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00107475], angle=26.88, y=1.12012
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00123815], angle=28.8, y=1.14001
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00138655], angle=30.72, y=1.16192
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00152869], angle=32.64, y=1.18601
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00141468], angle=34.56, y=1.21277
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00139956], angle=36.48, y=1.24199
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.0013416], angle=38.4, y=1.27412
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00124522], angle=40.32, y=1.30943
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00111957], angle=42.24, y=1.34829
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00115399], angle=61.44, y=2.08123
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00101255], angle=63.36, y=2.21725
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00241006], angle=0, y=1
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00238411], angle=1.92, y=1.00056
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.0023069], angle=3.84, y=1.00225
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00218036], angle=5.76, y=1.00507
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00200766], angle=7.68, y=1.00905
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00179322], angle=9.6, y=1.0142
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00154259], angle=11.52, y=1.02056
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00126247], angle=13.44, y=1.02816
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00102665], angle=23.04, y=1.08587
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00136865], angle=24.96, y=1.10212
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00167871], angle=26.88, y=1.12013
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00195064], angle=28.8, y=1.14002
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00218155], angle=30.72, y=1.16195
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00237331], angle=32.64, y=1.18607
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00253409], angle=34.56, y=1.21255
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00234082], angle=36.48, y=1.24201
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.0022598], angle=38.4, y=1.27414
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00210416], angle=40.32, y=1.30947
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00188128], angle=42.24, y=1.34835
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00160536], angle=44.16, y=1.39118
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00129936], angle=46.08, y=1.43843
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.0011973], angle=55.68, y=1.76765
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00151835], angle=57.6, y=1.85915
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00169251], angle=59.52, y=1.96285
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.00189185], angle=61.44, y=2.08156
000080.271: D3D9: ERROR: Difference greater than 0.001 [0.0017662], angle=63.36, y=2.21772

---------- Post added at 07:26 PM ---------- Previous post was at 07:17 PM ----------

ISS station default scenario look almost OK. Orbital view with sun and its flares was impressive, I was waiting for sunset (no slider experiments) - then message appeared that Orbiter must terminate.
I found D3D9 log html file. It was very long, but the end was like that:

[COLOR=Gray](50473: 104.2s 02.75ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atms.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50474: 104.2s 00.27ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atmo.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50475: 104.3s 10.51ms)(0x1420)[/COLOR][COLOR=Olive] Deleting TileManagerBase 0x9429FD40 ...[/COLOR]
[COLOR=Gray](50476: 104.3s 00.01ms)(0x1420)[/COLOR][COLOR=Olive] Recycling Pool Status nVtx=2, nIdx=2[/COLOR]
[COLOR=Gray](50477: 104.3s 00.28ms)(0x1420)[/COLOR][COLOR=Olive] Deleting TileManagerBase 0x9429F338 ...[/COLOR]
[COLOR=Gray](50478: 104.3s 00.01ms)(0x1420)[/COLOR][COLOR=Olive] Recycling Pool Status nVtx=2, nIdx=2[/COLOR]
[COLOR=Gray](50479: 104.3s 03.36ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atms.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50480: 104.3s 00.30ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atmo.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50481: 104.3s 11.26ms)(0x1420)[/COLOR][COLOR=Olive] Deleting TileManagerBase 0x9429F338 ...[/COLOR]
[COLOR=Gray](50482: 104.3s 00.02ms)(0x1420)[/COLOR][COLOR=Olive] Recycling Pool Status nVtx=2, nIdx=2[/COLOR]
[COLOR=Gray](50483: 104.3s 00.25ms)(0x1420)[/COLOR][COLOR=Olive] Deleting TileManagerBase 0x9429FD40 ...[/COLOR]
[COLOR=Gray](50484: 104.3s 00.01ms)(0x1420)[/COLOR][COLOR=Olive] Recycling Pool Status nVtx=2, nIdx=2[/COLOR]
[COLOR=Gray](50485: 104.3s 02.78ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atms.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50486: 104.3s 00.28ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atmo.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50487: 104.3s 10.17ms)(0x1420)[/COLOR][COLOR=Olive] Deleting TileManagerBase 0x9429FD40 ...[/COLOR]
[COLOR=Gray](50488: 104.3s 00.02ms)(0x1420)[/COLOR][COLOR=Olive] Recycling Pool Status nVtx=2, nIdx=2[/COLOR]
[COLOR=Gray](50489: 104.3s 00.24ms)(0x1420)[/COLOR][COLOR=Olive] Deleting TileManagerBase 0x9429F338 ...[/COLOR]
[COLOR=Gray](50490: 104.3s 00.01ms)(0x1420)[/COLOR][COLOR=Olive] Recycling Pool Status nVtx=2, nIdx=2[/COLOR]
[COLOR=Gray](50491: 104.3s 03.33ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atms.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50492: 104.3s 00.28ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atmo.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50493: 104.3s 11.47ms)(0x1420)[/COLOR][COLOR=Olive] Deleting TileManagerBase 0x9429F338 ...[/COLOR]
[COLOR=Gray](50494: 104.3s 00.02ms)(0x1420)[/COLOR][COLOR=Olive] Recycling Pool Status nVtx=2, nIdx=2[/COLOR]
[COLOR=Gray](50495: 104.3s 00.74ms)(0x1420)[/COLOR][COLOR=Olive] Deleting TileManagerBase 0x9429FD40 ...[/COLOR]
[COLOR=Gray](50496: 104.3s 00.01ms)(0x1420)[/COLOR][COLOR=Olive] Recycling Pool Status nVtx=2, nIdx=2[/COLOR]
[COLOR=Gray](50497: 104.3s 02.77ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atms.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50498: 104.3s 00.28ms)(0x1420)[/COLOR][COLOR=Olive] Loading Atmospheric Configuration file [GC/Earth.atmo.cfg] Handle=0x11880FC0[/COLOR]
[COLOR=Gray](50499: 113.2s 8948.41ms)(0x1420)[/COLOR][COLOR=Olive] --------------ExitModule------------[/COLOR]
[COLOR=Gray](50500: 113.2s 00.54ms)(0x1420)[/COLOR][COLOR=Olive] Log Closed[/COLOR]

 

[jarmonik]

000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00146906], angle=0, y=1
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.0014548], angle=1.92, y=1.00056
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00141241], angle=3.84, y=1.00225
000080.257: D3D9: ERROR: Difference greater than 0.001 [0.00134307], angle=5.76, y=1.00507

That is a result from tuning the atmospheric scale height lower then the math can handle. So, this is a result from your experiment (attempt to fix dark/over contrasted Earth) I guess ?

 

[Marg]

Hm, but I did not change anything in ISS orbital scenario... value was default 8.0 K. But colors on Earth are weird, too dark, etc, like sliders messed up, but it was before I began to change them. And I quickly pressed "LOAD SETTINGS" to their default values.

 

[SolarLiner]

Here's a build with the corrected Earth visual creation behavior. Copy the DLL to /Modules/Plugin and overwrite the dll.

 

[Marg]

First impression - currently no runtime errors... orbital ISS sunrise scenario (in 2016 standard package) runs normally.
KSC scenery still dark, but currently no errors also.

 

[DaveS]

Here's a build that should work with RC1 - I have merged my changes with the trunk.
To reduce file attachment size, the source is downloadable here.

While the glare and lens flares look good, the lighting model in this is completely broken. Everything is way dark.

 

[Marg]

After 30 mins -no errors. I noticed, that Moon does not hide flare and terrain seems to have no effect.

 

[SolarLiner]

While the glare and lens flares look good, the lighting model in this is completely broken. Everything is way dark.

That might be because of the tonemapping formula I put in - however I should have had updated all relevant parts of the code to reflect that change - mainly disabling brightness clamping to allow the tonemapper to be used at its full potential. I also tried to convert everything to linear color space (especially color textures), so that everything would blend more naturally and then the gamma correction would be applied in the end.

 

[DaveS]

Is this correct:

This is extracting all the files from both your archives and letting them overwriting the existing D3D9Client files.

 

[SolarLiner]

This is not normal, but it still is behaving normally given the state of the code. Right now compared to the average luminosity of the Earth, the panel is emitting at only ~30% brightness, and because there is no dynamic adaptation, the panel appears dark. So I'm working on basically "enabling HDR" for every emissive surface and light source so that they can make use of the full dynamic range.

Either way it's gonna be fixed with as much backward compatibility as possible.

---------- Post added at 22:39 ---------- Previous post was at 22:36 ----------

After 30 mins -no errors. I noticed, that Moon does not hide flare and terrain seems to have no effect.

That is normal given the current state of the code - the obstruction checks only take the planet radius into account and its atmosphere - and only of the closest body. Wether or not we need to change that will depend on how frequent are lunar eclipses during gameplay. As for terrain, I need to get a better obstruction check system going; but that is more on jarmo's territory here.

 

[jedidia]

I was just wondering if anyone ever noticed that bases render through the terrain?

---------- Post added at 09:02 PM ---------- Previous post was at 08:51 PM ----------

Also, my microtextures aren't working anymore with RC1 and appropriate D3D9 client. I've skiped a few revisions in between, so I suppose something changed. How do microtextures have to be named and where must they be located now?

 

[Eduard]

I was just wondering if anyone ever noticed that bases render through the terrain?

Yes I noticed it too, especially around Brighton Beach. You can see the launch pads through the mountains.
But this problem also exists within the stock client, not only in the D3D9Client.

Also, my microtextures aren't working anymore with RC1 and appropriate D3D9 client. I've skiped a few revisions in between, so I suppose something changed. How do microtextures have to be named and where must they be located now?

See my comments and Jarmo's answer on this thread: http://orbiter-forum.com/showthread.php?t=36637&page=20

 

[jedidia]

But this problem also exists within the stock client, not only in the D3D9Client.

whoops. I actually tried it in MOGE to see whether it's a client problem, and thought I didn't see it there. Tried again, now I see it. Must be one of these days.

See my comments and Jarmo's answer on this thread

I was refering to the microtextures, that thread looks like it's dedicated to standard textures.

 

[jarmonik]

I was just wondering if anyone ever noticed that bases render through the terrain?

No valid solutions there. It is either seeing landing pads and runways being partially under a terrain or seeing them through the terrain.

Also, my microtextures aren't working anymore with RC1 and appropriate D3D9 client. I've skiped a few revisions in between, so I suppose something changed. How do microtextures have to be named and where must they be located now?

There was a microtexture format change a long ago. There is a conversion utility in the D3D9DebugControls. Click "Open" to open an old texture which should definitely be uncompressed and it must have mipmaps. Then click "Execute" to save it in a proper format. It's also possible to assign it in a specific texture slot for a realtime preview with an assign option selected from a combo box. The utility is pretty slow so a delay of a few seconds can happen after clicking "Execute".

We should prepare an official microtexture distribution package so would you care to share the textures you have ?

 

[DaveS]

Something I have been meaning to ask about is how the rest of the texture generation tools in D3D9DebugControls are coming along? I would love to update the SSU textures so that they're ready for 2016.

 

[jarmonik]

Something I have been meaning to ask about is how the rest of the texture generation tools in D3D9DebugControls are coming along? I would love to update the SSU textures so that they're ready for 2016.

The rest of the tools ? There are no other tools scheduled than ability to bake base tiles to planetary textures and some other terrain editing tools.

The ground work is laid and ready, we have a surface snap and a proper terrain/mouse cursor interaction implemented. Which was likely the hardest part. But no progress other than that.

 

[jedidia]

We should prepare an official microtexture distribution package so would you care to share the textures you have ?

Sorry, I have only what has been distributed in the past. Not much of an image editor...

 

[DaveS]

The rest of the tools ? There are no other tools scheduled than ability to bake base tiles to planetary textures and some other terrain editing tools.

I was thinking of the material texture generators (specular, roughness, reflection etc).

 

[jarmonik]

I was thinking of the material texture generators (specular, roughness, reflection etc).

On that sector everything is ready and implemented as far as I can tell. I am going to leave the "advanced renderer" (i.e. Vessel.fx) for Felix24 so he can implement that by his own preferences. The client can switch between Vessel.fx and PBR.fx in a per meshgroup basis.

I guess I should focus in writing some documentation regarding the PBR system and client architecture in general. Right now the client is in a good condition and ready for Orbiter 2016 release. Except the documentation of course.

 

[DaveS]

On that sector everything is ready and implemented as far as I can tell. I am going to leave the "advanced renderer" (i.e. Vessel.fx) for Felix24 so he can implement that by his own preferences. The client can switch between Vessel.fx and PBR.fx in a per meshgroup basis.

As far as I can tell, it doesn't save the modifications you do. That is what I talking about.