Project D3D11Client Development

[asmi]

Whatever you did seems to have fixed it for me.

:thumbup:

Awesome!

 

[Mandella]

Taking up AMSO's Apollo 14, all settings same as before. "J" key fails to activate tower gantry swing out. "K" key activates the sound pack and the takeoff just fine. Liftoff and autopilot to orbit work great, as do cockpit command keys (Tab to pull up info, and camera shift keys.)

Separation particle effects look blocky, which is the only visual problem I noted.

Hrm. Now that I think about it, I am also running the new hi-res KSC from Martin. I'll unload that and check to see if "J" key problems persist. Also, I see that AMSO has a 1.21 version out (I'm using 1.20). I'll download and try that also later tonight.

:tiphat:

 

[Frogisis]

It worked for me, too - Thanks!
There's still a veeeery slight "wobble" in the brightness but it's almost beneath notice, and might already be a known issue (or just my machine being weird).

It runs continuously above the refresh rate of my monitor with all the settings cranked up all the way, too, so I'm going to play around with putting in and taking out addons to see what I can get away with.

 

[NovaSilisko]

hi-res KFC

 

[asmi]

I take it that you really like chicken, aren't ya? :rofl:

 

[Mandella]

I take it that you really like chicken, aren't ya? :rofl:

Apparently, I'm the one that's hungry. :lol:

 

[asmi]

Apparently, I'm the one that's hungry. :lol:

Looks like you had a late dinner

Last edited by Mandella; Feb 4th, 2012 at 21:00. Reason: Replaced fast food chicken eatery with Space Center..

I'd prefer you to leave things as they were - it looked much more fun that way :thumbup:

 

[Mandella]

Anyway! :lol:

Let me show you those blocky particle effects:

D3D11 client:

Vanilla Orbiter Beta:



D3D11 client:

Vanilla Orbiter Beta:

Also of note, AMSO 1.21 does not like Orbiter Beta, and just won't run (incompatible version error). "J" key works fine in Vanilla Orbiter Beta, so it does seem to be a problem with AMSO and the external client.

 

[asmi]

Yes, I do have particle effects in my TODO list - there are quite some things that are to be done in that field.

 

[Mandella]

Yes, I do have particle effects in my TODO list - there are quite some things that are to be done in that field.

Right. Known issue then.

What about local lights? This is the effect of pressing "L" for navigation lights in AMSO (animation works correctly):

 

[asmi]

Right. Known issue then.

What about local lights? This is the effect of pressing "L" for navigation lights in AMSO (animation works correctly):

They should work correctly... At least they do for stock DG...
Honestly I've never used AMSO and I don't even know what it is Right now I'm more concerned about getting stock stuff to work as expected, with some popular non-stock stuff like XR vessels family as they seem to have little to no issues at all.
But if you'll provide some easy way to reproduce the issue (like saved scenario) as well as some instructions regarding how to get it to work (I mean what kind of stuff do I need to install, some config settings if they are relevant to reproducing the problem), then I would be able to diagnose and (hopefully) fix the issue. Also I will need a reference screenshot which would show how it's supposed to look.

 

[Mandella]

This is AMSO. It is an attempt to replicate the entire Apollo program in Orbiter, moon buggy and all. It may well be as popular as the XR series of vessels, although that is hard to say, since they both download from sites independent of Orbit Hanger.

But I get what you are saying. Since AMSO is a non stock app, it is probably going to take some recoding of AMSO itself to work with D3D11 (I know the AMSO devs did something similar with the D3D9 project that is in the works, so they might be willing to "meet half way" if contacted).

But in the meantime, I'll update my XR vessel collection (I love flying them too!) for future testing. Or should I stick with stock vessels for now, such as the base Delta Glider?

Anyways, here is that picture of what the lights look like in the stock client:

 

[asmi]

Anyways, here is that picture of what the lights look like in the stock client:

I can tell you right away what most likely a problem is. As far as I understand, authors use beacons to display these "lights". And since stock client doesn't actually render lighting from beacons - just beacons themselves - that's an end story for them. But D3D11 client does actually process and render lighting caused by beacons, and this data is being retrieved from beacons' definitions. Since stock client doesn't care much about pretty much every beacon parameter except for a few that are used to render "the light bulb" - such as light color, light size and position - so developers could pass in pretty much anything without any impact on the visual appearance, while our client requires full set of beacon parameters to be set up accordingly. So if I'm right (and I will check this tomorrow if I'll manage to get it working), no one but AMSO devs can fix that - but the way I see it this fix should be very easy and straightforward for them - they will probably need to use D3D11 client to tune up beacons' parameters, as IIRC this client is the only one that renders beacon lighting.

I still need a saved scenario to confirm aforementioned diagnosis, so please provide me with one, but I'm pretty confident with it.

---------- Post added at 22:39 ---------- Previous post was at 22:35 ----------

But in the meantime, I'll update my XR vessel collection (I love flying them too!) for future testing. Or should I stick with stock vessels for now, such as the base Delta Glider?

Well you can use whatever you want - we will need the client to be tested with all kinds of additions sooner or later - just keep in mind that our first priority is to get stock stuff to work. So if you use some third-party vessel and find a visual bug (like the one with AMSO), please report it here anyways. If it's an easy one (or a "not our bug"-type of thing which is what AMSO issue seems to be all about) - we'll get it fixed right away, while in other cases please understand that we might prefer to defer it to a later time due to a presence of higher-priority things in our current TODO list.

 

[Mandella]

Understood! These are just bug reports, not "you gotta get em fixed right now!!" reports.
:lol:

 

[Namahage]

Hmm, that's weird. I think we gonna have to ask Glider to take a look, since I've run out of ideas.

Okay. Could it possibly be the crusty drivers from Lenovo? The latest posted version is Sept. 2010. I also attempted to install the Intel generic driver from circa Nov. 2011 but the installer wouldn't allow it due to proprietary code requirements on the machine.

I will be changing domiciles in the near future so my forum-checking will be spotty for a few months. I'd still like to keep testing, though. If you need to reach me a PM will be most effective.

 

[asmi]

Okay. Could it possibly be the crusty drivers from Lenovo? The latest posted version is Sept. 2010. I also attempted to install the Intel generic driver from circa Nov. 2011 but the installer wouldn't allow it due to proprietary code requirements on the machine.

I will be changing domiciles in the near future so my forum-checking will be spotty for a few months. I'd still like to keep testing, though. If you need to reach me a PM will be most effective.

It *could* be the reason since we have at least one confirmed case of CTDs due to outdated drivers, when updating them resolved the issue...

---------- Post added at 22:36 ---------- Previous post was at 19:50 ----------

Understood! These are just bug reports, not "you gotta get em fixed right now!!" reports.
:lol:

Ok it was partially my fault as diffuse lighting did not take falloff into account - it was calculated for specular lighting only. This is now fixed and AMSO looks way better. Allthough I would still recommend AMSO devs to review beacons' parameters, possibly increasing falloff value. Here is a formula and some graphs for different falloff parameter values (horizontal axis is distance from light source to a pixel, vertical axis is attenuation multiplier, the formula is final_color ~= color * attn):

Please get latest and check it out.

 

[DaveS]

What is the latest workings on the project?

 

[asmi]

What is the latest workings on the project?

As far as I know Glider is working on heightmaps implementation. I'm working on figuring out a way to implement self-shadowing. But since I went on vacation it's going kinda slow , but I will be back at full speed in 3 weeks.

 

[dumbo2007]

Nice, would it be possible to get heightmaps to be as detailed and look as good as Orulex with shadowing and bump mapping ?

 

[Robrutus]

Not working, crashes on launch.

Hi, it is commendable a group is working on a DirectX 11 client.

I've done a clean install of Orbiter and used the 319192b7a89f version of the d3d11 client.

Orbiterng.exe loads, but crashes on launch scenario or video tab selection, it may sound a silly question, and I apologise, but what have I done wrong? Thanks for your time.

P.S. I've attached install.log, Orbiter.log and dxdiag.log files for you to browse.

View attachment dxdiag.log

View attachment install.log

View attachment Orbiter.log