Just something cool from the blog of Leadwerks CEO Josh Klint, an early view of Leadwerks Engine 3 running on the first of several mobile platforms. OpenGLes2 seems to be able to manage the more advanced shaders quite well. Android and Mac cross-compilation too. Being able to cover these platforms in the non too distant future is quite exciting.
Blog entry here...
http://www.leadwerks.com/werkspace/blog/1/entry-674-multiplatform-madness/
Wednesday, 29 June 2011
Tuesday, 14 June 2011
AD updates his diary on SimHQ
AD added some screen-shots of new assets and general polishing that's going on for the USA firing-range facility being built as a pre-pay beta/support training/prologue release mini-game thing.
You can read it here...
Oil pipeline and animated pumps |
Particle Ribbons
My news is more technical. However I've been working on my own particle system that will serve to improve rocket motor smoke and as a side-effect will (later) work as a backbone for an imposter/3D particle cloud system and possibly even dynamic ground vegetation. I'm really pleased with results so far.
Particle ribbon - 75m long smoke trail |
Advanced Flight Model
Also the advanced flight-model has had a lot of love and is feeling really nice. Fred has done a great job. Some issues relating to anti-torque control and stabilisers to resolve but I'm really pleased with it. It's certainly responsive to small inputs, to get the absolute best flying experience I would recommend a full sized helicopter cyclic replica. Given the amount of controls on a real Apache cyclic any advanced HOTAS comes highly recommended. Although an XBOX joypad still works we'll need to fix up some damping.
Some details need to be worked out on migrating the new flight model into the main project source. The engine instruments need to be hooked into the FM power-plant outputs as well as hooking up the inputs.
It's all one step closer to the testing phase.
Thursday, 9 June 2011
Screenshot of the day
Time for another one of these filler posts. Hey, you don't have to read'em but I have to keep posting or I begin to lapse.
Spent the morning testing the new Leadwerks Engine 2.43 update which improved weapons fire by having collision hulls generated at load time. And a problem with recursive AABB bounding box calculation. AOE weapons use bounding boxes to begin damage testing.
Problem with clustering small objects such as rockets so close together in a scene with a large camera range are the z-buffer errors that arise, these appear as clipping or z-fighting. What I've done to mitigate this is apply a proportional z buffer offset to 'enclosing' entities.
This takes the distance from the vertex and camera, then applies this as an offset to the depth buffer. It's useful only for decals or large hulls, any error in overdraw is masked by camera distance. Annoying stuff like this takes time away from game internals but z-fighting bugs me even more. There are still parts of the Apache that exhibit this, the glass parts of the TADS for example but you only notice them with the night vision system.
*update*
US road-markings...
Spent the morning testing the new Leadwerks Engine 2.43 update which improved weapons fire by having collision hulls generated at load time. And a problem with recursive AABB bounding box calculation. AOE weapons use bounding boxes to begin damage testing.
Problem with clustering small objects such as rockets so close together in a scene with a large camera range are the z-buffer errors that arise, these appear as clipping or z-fighting. What I've done to mitigate this is apply a proportional z buffer offset to 'enclosing' entities.
float dist = length(modelvertex.xyz - cameraposition ) / 512.0; float z = DepthToZPosition(gl_FragCoord.z); z = z - dist; gl_FragDepth = ZPositionToDepth( z );
This takes the distance from the vertex and camera, then applies this as an offset to the depth buffer. It's useful only for decals or large hulls, any error in overdraw is masked by camera distance. Annoying stuff like this takes time away from game internals but z-fighting bugs me even more. There are still parts of the Apache that exhibit this, the glass parts of the TADS for example but you only notice them with the night vision system.
*update*
US road-markings...
If you have access to BBC iPlayer....
Our War - 1. Ambushed
Concerns 3 Platoon 1st Battalion Royal Anglian regiment in Helmand province 2007. Using their own helmet cams the MoD granted the BBC permission to use their footage in this uncompromising account.
http://www.bbc.co.uk/iplayer/episode/b011vngx/Our_War_Ambushed/
For those without BBC iPlayer access (because to the BBC the world wide web doesn't actually mean world wide) you might find this available as a torrent somewhere.
Concerns 3 Platoon 1st Battalion Royal Anglian regiment in Helmand province 2007. Using their own helmet cams the MoD granted the BBC permission to use their footage in this uncompromising account.
http://www.bbc.co.uk/iplayer/episode/b011vngx/Our_War_Ambushed/
For those without BBC iPlayer access (because to the BBC the world wide web doesn't actually mean world wide) you might find this available as a torrent somewhere.
Monday, 6 June 2011
Rocket Man.....
...burning out his fuse up here alone. Or something like that with apologies to Elton John.
I'm having problems sourcing decent internal and external 70mm rocket audio FX. Going quietly insane browsing online videos with enough quality, no music or talking, background noise. Might have to see about recording something, internally its similar to someone banging on a metal shutter.
Made some big improvements to stores loading, adding of preset load-outs and I nudged one of the shorter hydra rocket types so it sits forward in the M261 pod to make them visible (to facilitate zone loading).
I'm having problems sourcing decent internal and external 70mm rocket audio FX. Going quietly insane browsing online videos with enough quality, no music or talking, background noise. Might have to see about recording something, internally its similar to someone banging on a metal shutter.
Made some big improvements to stores loading, adding of preset load-outs and I nudged one of the shorter hydra rocket types so it sits forward in the M261 pod to make them visible (to facilitate zone loading).
First attempt at rocket motor burn effect |
Sunday, 5 June 2011
Wind - 5MW generator turbines
AD has been tinkering again. He's been making noises about wind-farms on and off for a while and surprised me with this one today.
These are US sized, 80 meter tall 5 megawatt wind generators complete with rumbling wind audio effects, anti-col lights and animation of course. Can't express enough how massive these are, such structures fill up screen real-estate economically which has traditionally made them a background object of choice for racing games (along with Ferris wheels and Space Needles).
Here in Yorkshire we have quite modestly sized wind-farms which caused some issues with radar returns at RAF Fylingdales if memory serves.
My first attempt at flying through the blades in the AH64D wasn't a great success. AD broke up the model to allow individual blades to come off when damaged.
Headtracking Software
Cyclic brought to my attention some alternative head tracking software: FaceTracknoIR
I'll be investigating how we can include this alongside other input devices. I like this one as it's within my budget (free with optional donation).
Alternative hardware we were looking at was Kinect for hand-tracking for operation of virtual MFDs, at least until the "non-commercial" nature of Microsofts Kinect SDK came to light. So until there's clarification on Microsoft's SDK usage, plans for virtual cockpit operation via Kinect will be shelved.
These are US sized, 80 meter tall 5 megawatt wind generators complete with rumbling wind audio effects, anti-col lights and animation of course. Can't express enough how massive these are, such structures fill up screen real-estate economically which has traditionally made them a background object of choice for racing games (along with Ferris wheels and Space Needles).
Here in Yorkshire we have quite modestly sized wind-farms which caused some issues with radar returns at RAF Fylingdales if memory serves.
My first attempt at flying through the blades in the AH64D wasn't a great success. AD broke up the model to allow individual blades to come off when damaged.
Headtracking Software
Cyclic brought to my attention some alternative head tracking software: FaceTracknoIR
I'll be investigating how we can include this alongside other input devices. I like this one as it's within my budget (free with optional donation).
Alternative hardware we were looking at was Kinect for hand-tracking for operation of virtual MFDs, at least until the "non-commercial" nature of Microsofts Kinect SDK came to light. So until there's clarification on Microsoft's SDK usage, plans for virtual cockpit operation via Kinect will be shelved.
Labels:
audio scripting,
face tracking,
wind farms
Saturday, 4 June 2011
APKWS - Laser guided Hydra rockets
We've discussed this before and is something I might add to the line-up of available munitions after the Hellfire.
http://www.defenseindustrydaily.com/apkws-ii-hellfire-jr-hydra-rockets-enter-sdd-phase-02193/
I was scratching my head over the launch order of Hydra rockets but found a post at a military photos website.
Hand typing all these position offsets within the pod was a bit of a pain. A certain amount of trial and error was involved. These will only be visible in the pods if Max Object Detail is turned on in graphics settings. As they are I'm not fond of having them in the tubes due to potential scale/fp issues later. But I needed the offset data for the ballistics.
*edit* Photo added for comment below...
http://www.defenseindustrydaily.com/apkws-ii-hellfire-jr-hydra-rockets-enter-sdd-phase-02193/
I was scratching my head over the launch order of Hydra rockets but found a post at a military photos website.
Hand typing all these position offsets within the pod was a bit of a pain. A certain amount of trial and error was involved. These will only be visible in the pods if Max Object Detail is turned on in graphics settings. As they are I'm not fond of having them in the tubes due to potential scale/fp issues later. But I needed the offset data for the ballistics.
MPSM almost largest of available 70mm rockets |
Not a loadout you would see, 19 Illumination rockets |
*edit* Photo added for comment below...
FFAR Launch |
Labels:
FFAR,
Hydra rockets,
launch order
Friday, 3 June 2011
Update for this week
You can tell it's been a busy week since my blog update frequency dropped. Usually a good sign that lots of cool stuff is getting done. The tool-set has some save problems to be fixed this weekend. The biggest mind bender has been the math behind trajectory solutions both predictive (against moving objects) and static. I'm quite impressed by how much of this had to be solved by the lead Longbow 2 programmers amongst all the other things they had to contend with. You can research these things on-line, Wikipedia is often a good first look but once you throw moving helicopters into the mix it gets complicate fast.
This week we have laid foundations for content tools, auto-update software, added the FFAR rockets (still working on these) and the ballistics solver. Looking through the source there's also the cubic spline stuff for multi-player positioning which is now in the main project but awaiting interface between that and the Raknet network library.
Pylons which carry munitions under the wings got their elevation code so they track up and down a small amount to help stabilise rockets during launch. DoF was also tweaked, something I'll use for screen-shots and vids (CTRL X to toggle).
Working on rocket trails and zones. Once that's taken care of there's the remaining two HUD modes and the advanced flight-model to finish. Once those milestones have been reached it's time for polishing prior to first stage testing.
It's been a lot of fun working on this, never get tired of tooling around at tree-top height. I will miss working on it when it's over.
One last thing, AD made a short video of the Firing Range map (still un-named) which a nice lonely piano piece. The calm before the storm.
This week we have laid foundations for content tools, auto-update software, added the FFAR rockets (still working on these) and the ballistics solver. Looking through the source there's also the cubic spline stuff for multi-player positioning which is now in the main project but awaiting interface between that and the Raknet network library.
Pylons which carry munitions under the wings got their elevation code so they track up and down a small amount to help stabilise rockets during launch. DoF was also tweaked, something I'll use for screen-shots and vids (CTRL X to toggle).
Working on rocket trails and zones. Once that's taken care of there's the remaining two HUD modes and the advanced flight-model to finish. Once those milestones have been reached it's time for polishing prior to first stage testing.
It's been a lot of fun working on this, never get tired of tooling around at tree-top height. I will miss working on it when it's over.
One last thing, AD made a short video of the Firing Range map (still un-named) which a nice lonely piano piece. The calm before the storm.
Subscribe to:
Posts (Atom)