New Release D3D9Client Development

[DaveS]

Onward to the next issue, this time with the latest Orbiter 2015 Beta (Rev. 13) and the latest D3D9Client for the beta (Rev. 12).

This concerns an old Orbiter bug report from 2011: http://www.orbiter-forum.com/project.php?issueid=550

This is very obvious in the D3D9Client:

Is there any way to address this? Or would it require core changes?

 

[jarmonik]

It's possible to address the missing fog issue from a client side. Current atmospheric model is an experiment and it shows that this kind of model could be a potential candidate for a "final" model especially after a few improvements. Vessel/Structure fog issue will be fixed then.

 

[jarmonik]

D3D9Client R15.1

Here is a new development snapshot from D3D9Client for Orbiter 2010-P1. I don't really remember what's been changed since the last release.

But at least:
- DGIV CTD has been fixed.
- Normal Map related UV mirror issue has been fixed. (Pay attention to normal mapping issues)
- Geometry instancing has been removed. (Since it was not working as expected).

Are there any high priority bugs or problems existing for 2010-P1 release ?

 

[Cras]

gonna drop this in and try this later today, I'll let you know how it goes.

 

[softmaster]

Here is a new development snapshot from D3D9Client for Orbiter 2010-P1. I don't really remember what's been changed since the last release.

But at least:
- DGIV CTD has been fixed.
- Normal Map related UV mirror issue has been fixed. (Pay attention to normal mapping issues)
- Geometry instancing has been removed. (Since it was not working as expected).

Are there any high priority bugs or problems existing for 2010-P1 release ?

Thanks for this!

 

[Weirdo Earthtorch]

Latest version works fine.

Something I've been meaning to ask: full-screen mode (True, Window with Taskbar, Full screen window) always uses monitor native resolution (1920x1200 in my case) despite what's set in the resolution drop-down - is this what should happen? I was trying to test full-screen, but with lower resolutions to see if there was a framerate improvement.

Cheers, WE.

 

[jarmonik]

Something I've been meaning to ask: full-screen mode (True, Window with Taskbar, Full screen window) always uses monitor native resolution (1920x1200 in my case) despite what's set in the resolution drop-down - is this what should happen? I was trying to test full-screen, but with lower resolutions to see if there was a framerate improvement.

"Window with Taskbar" and "Full screen window" are currently locked to a desktop resolution and changing the desktop resolution would be a bad idea. However, the "True Fullscreen" mode is working fine with a resolution selected from a drop-down menu (at least on my computer). So, could you check that again ?

 

[Weirdo Earthtorch]

I'm running an extended desktop. (Laptop screen = 2nd display, primary display is 1920x1200 monitor).

If I disable extended desktop (single display) True full screen works as described (non-native resolutions work fine).

If I check the multiple displays box on the client tab, with extended desktop enabled, then orbiter starts, but I get a blank screen on the primary display, and have to alt-tab out of Orbiter. If I don;t select multiple displays, then I just get a blank screen with nothing happening.

Win 7 on i3 laptop (HD3000).

If it helps, I will post D3D9Client logs for the different scenarios in a short while.

Thanks, WE.

 

[jarmonik]

If I check the multiple displays box on the client tab, with extended desktop enabled, then orbiter starts, but I get a blank screen on the primary display, and have to alt-tab out of Orbiter. If I don;t select multiple displays, then I just get a blank screen with nothing happening.

I am assuming that this happens in a "True Fullscreen" mode ?

I can confirm that "True Fullscreen" mode doesn't work with extended desktop, and there's nothing I can do about it. (Should work with cloned display)

I can also confirm that "Multiple displays" check-box has no effect in True Fullscreen mode and it's likely not working in any mode, if the display height resolutions are not equal.

 

[Weirdo Earthtorch]

I am assuming that this happens in a "True Fullscreen" mode ?

Yes.

I can confirm that "True Fullscreen" mode doesn't work with extended desktop, and there's nothing I can do about it. (Should work with cloned display)

I can also confirm that "Multiple displays" check-box has no effect in True Fullscreen mode and it's likely not working in any mode, if the display height resolutions are not equal.

I didn't check cloned mode. What this has done is made me realise that even though the " True Fullscreen" mode works for the primary monitor in extended desktop, (but only at native resolution), if any dialogs are opened (eg F3 ship selection etc.) this causes the same effect as alt-tabbing. It works fine in single display.

Maybe something worth noting in the docs? I'm not sure how many people might be operating with a (laptop/)additional monitor in extended mode.

 

[Marg]

I just noticed pretty annoying thing, when using Orbiter BETA and latest Jarmonik's D3D9 - during launch, smoke is not cut off by ground - is rendered below ground level. It does not happen with inline client, that was one of the features I noticed quickly in new Orbiter BETA. I have a feeling that earlier D3D9 versions also had this OK. But now it is back

 

[jarmonik]

I just noticed pretty annoying thing, when using Orbiter BETA and latest Jarmonik's D3D9 - during launch, smoke is not cut off by ground - is rendered below ground level. It does not happen with inline client, that was one of the features I noticed quickly in new Orbiter BETA. I have a feeling that earlier D3D9 versions also had this OK. But now it is back

I wasn't aware of that. Looks like a surface base terrain visibility hack broke it. Let's hope it's a simple z-buffer issue.

EDIT: Fixed

 

[Abloheet]

umm, where is the fixed version? i still get smoke below the ground..

 

[kuddel]

umm, where is the fixed version?

It's in the "trunk" https://d3d9client.svn.codeplex.com/svn/trunk

Revision: 36958
Author : SND\jnikkanen_cp
Date : Samstag, 1. August 2015 01:17:19
Message : - Particle effect terrain collision fix

You have to build it yourself of course.
An official compiled version will be released by jarmonik I think.
Unfortunately I can't currently attach any ZIP (forum problem I think).

 

[kuddel]

Attached yo'll find the latest development build (for Orbiter BETA)

 

[Kyle]

Having some issues with AMSO in the latest D3D9 and Orbiter 2010-P1.

For whatever reason whenever I launch the Saturn V, the FPS drops down to 1 FPS. I tried this in the inline client and I wasn't able to replicate it. Anyone else having this problem?

 

[Marg]

I just thought is it possible to make water in Orbiter 2010 P1 to look like in Orbiter BETA 2015?

 

[kuddel]

Having some issues with AMSO in the latest D3D9 and Orbiter 2010-P1.

For whatever reason whenever I launch the Saturn V, the FPS drops down to 1 FPS. I tried this in the inline client and I wasn't able to replicate it. Anyone else having this problem?

Hi Kyle,

I have tried but could not reproduce your issue.
Here's my test-setup (maybe you find some differences to yours that might give a hint to the cause):

Orbiter||2010-P1| orbiter100830.zip [1]
D3D9Client||D3D9Client R15| D3D9ClientR15.zip [2]
AMSO||AMSO-121| AMSO-121.zip [3]
OrbiterSound||OrbiterSound 4.0| OrbiterSound40_20121120_setup.exe [4]
Scenario||Apollo 08|Scenarios\AMSO\All missions lift off\Apollo 08.scn

I first tried without OrbiterSound, but it did not make any difference (with or without it, the Saturn launch did not drop much the framerate)

/Kuddel

[1] http://orbithangar.com/orbDownload.php?file=orbiter100830.zip
[2] http://download-codeplex.sec.s-msft...07512&FileTime=130654428836630000&Build=21029
[3] http://www.acsoft.ch/AMSO/AMSO-121.zip
[4] http://orbiter.dansteph.com/downloadround/orbitersound40/OrbiterSound40_20121120_setup.exe

 

[Kyle]

Strange, I'll try it on a clean install and get back to you.

EDIT: Just ran it on a clean install. Issue was repeatable. I only had AMSO, Orbiter 2010 P1, and the new D3D9. I was also able to duplicate it on my substantially higher performing laptop, same issue. Only addon that has this issue.

FWIW, I just switched to Windows 10 a few days ago on both my PC and my laptop. Any idea if that might be related to the issue?

 

[DaveS]

Is there something that can be done to fix this uneven lighting issue on meshes?