787 and CRJ-200 Development

[Jetman.]

That section of seats wouldn't need windows

[dany93]

That section of seats wouldn't need windows

Just keep seatbelts fastened!

To Nickyivyca:

Transparency:
With your (last) 787112210, I have exactly the same images (transparency) as yours, same locations. Useless to re-post them.
I do not have this with 787111410-V04. Everything is right even when looking at the landing lights dazzling us from front.

Main gear: (I hope, but am not sure to understand what you mean)
787112210: like your image, rear wheels alone touch the ground when rolling on the runway, from slow to high speed. They have a normal position only when strongly compressed (try landing after a brief taking-off). [EDIT : Compressed about 0.1, observe taking-off and landing]. Otherwise, they are very hard, steady.
787111410-V04: The effect is more moderate. Front wheels alone touch the ground when stopped and at slow speed, then progressively with acceleration position comes normal, then inverse, with rear wheels alone touching at higher speed. By making slight turns when rolling, it gives the impression that the front wheels and rear wheels (of this main gear) do not have the same compression factor. I presume this is to take the right position (front wheels higher) at landing just before touching down. This point seems rather difficult to adjust.... Sorry I can't help you more, I've only experimented modifying animations on very simple cases. In the Models/787.xml file, I more or less find where it is, but i'm lost in the interpolations, compressions, from which point, which parts ..... Maybe I can help a bit, but I cannot do.

my Configuration:
Dell XPS 420 with Windows Vista 6.0 (No 6002 ; Service Pack 2) 32 bits
Intel Core Quad CPU Q6600 @ 2.40GHz
RAM: 3GB
Graphics card: 512 MB ATI Radeon HD 3800 series w/ Driver V.10.9 (updated 19 sept 2010)
Keyboard with numeric keypad

FlightGear V.2.0.0
Launcher 1.5.2, in French

[Rick Ace]

After a while flying in the CRJ-200 it seems to develop an issue. I see no log errors. But here is what happens. After a good while of flying (or standing there parked doing nothing) the simulation pauses, and goes, pauses and goes. SO if you were flying...pause for sec...2 secs of flying...pauses...2 secs...uniformly

[nickyivyca]

Currently my only possible hypothesis is that since I switched the properties for the lights, their animations were messed up. I've tried to sync the properties up with their animations and the switches in the cockpit, because I'm not having this problem with the CRJ-200 or any other aircraft.

I'll also take a look at the gear animations. I do think that the damping on the front gear should definitely be increased, as well.

So matching the properties up allowed me to select away the messy lights. But it didn't solve the problem. Any suggestions from model experts? I'm using FG 1.9.1, also with an ATI card.

Note: it does work on the older .ac model. I'll try finding out the differences with the older one and change them back.

Back again: I can see no differences between the two different files in terms of their lighted parts. Eek.

[skyop]

Starting and pausing? That happened with me on my old ATI card, with the Concorde.

Nick, check to see if the model is optimized- be sure all objects using the same texture are grouped with one parent, remove any unnecessary transparency, etc.

[redneck]

After a while flying in the CRJ-200 it seems to develop an issue. I see no log errors. But here is what happens. After a good while of flying (or standing there parked doing nothing) the simulation pauses, and goes, pauses and goes. SO if you were flying...pause for sec...2 secs of flying...pauses...2 secs...uniformly

Didn't you have an issue with a nasal loop that caused the same problems? Or is this specific to that plane?

[Rick Ace]

Starting and pausing? That happened with me on my old ATI card, with the Concorde.

Nick, check to see if the model is optimized- be sure all objects using the same texture are grouped with one parent, remove any unnecessary transparency, etc.

That happens to the concorde and crj-200 too But for the concorde, it's always like that. But for the CRJ-200 it's only after an hour or something close to that.

[nickyivyca]

The main model only uses 3 different textures-UAX.png for main textures, UAX-lights.png for lighted textures, and then transparent.rgb for the lights themselves. It also uses a chrome shader and a glass shader. I haven't noticed the problem, but try commenting out the glass shader in the main model file.

Edit: So I've done some testing on the transparency issue. So if I copy the 787-8 from Dany's release and put it into my folder, I get the problem. It's also showing up in FGRun. Next up is to test it out on my laptop, though I expect the same results because the problem happened to Dany.

As for the CRJ-200, I heard that the latest version doesn't work correctly with 2.0. Anyone using 2.0 having this problem?

[dany93]

I have no transparency issue with CRJ-200 (CRJ-200111310 with FG 2.0, Vista, ATI card). But its landing lights are not so powerful than 787-8's, and they have neither halo nor impact on the fuselage. Part of the problem appears at this light impact or in the halo on 787's.
For the gear compression, (knuckle? or strut?) the four wheels together odd-rotation-with-weight issue, you can easily observe their animation by merely filling and emptying your fuel tanks (more visible on 787111410's than on 787112210's which are very hard). I feel this could be adjusted by a several-step interpolation, freezing the rotation above some compression, but I can't find the parts involved. Or maybe I haven't understood...

[Calico]

With a Mac with 10.5.8 and an Intel chip, the newest CRJ strangely doesn't seem to work for me. Here's a link to the older one I've been using: https://sites.google.com/site/shsater/

It seems to have a reasonable FDM. With 47000 pounds gross it'll approach with no complaint at Vref 141, and in the flare stall warning horn goes of at around 125 kt.

[Rick Ace]

The main model only uses 3 different textures-UAX.png for main textures, UAX-lights.png for lighted textures, and then transparent.rgb for the lights themselves. It also uses a chrome shader and a glass shader. I haven't noticed the problem, but try commenting out the glass shader in the main model file.

Edit: So I've done some testing on the transparency issue. So if I copy the 787-8 from Dany's release and put it into my folder, I get the problem. It's also showing up in FGRun. Next up is to test it out on my laptop, though I expect the same results because the problem happened to Dany.

As for the CRJ-200, I heard that the latest version doesn't work correctly with 2.0. Anyone using 2.0 having this problem?

I'm having the problem with the stop-go lag after a while of flying

[nickyivyca]

For the transparency problem...

Someone had the EXACT same thing happen on 1.9.1. Their solution? Remove the alpha layer of the base livery. So I'll try that out. It seems to work in FGRun, as the problem could show up in that when it was happening.

Guess how I found that? I was google searching "landing lights on fuselage" to try to see how landing lights really reflect off of a fuselage. That image in the thread was one of the ones that popped up. I think that only the original livery (what the model is mapped to itself) is what matters.

By the way, I should have a CRJ-200 release out tomorrow. Just some little things, all of the stuff with known causes that have been mentioned. If the 787 lightfix is successful, I'll do some polishing of systems and model and then make a 787 release.

As promised:
New CRJ-200 Release!

http://www.mediafire.com/?ae7kja0aeayp72m

This release is mostly to correct small things that I or other users saw.

-Fixed pushback animations-both the wheels, and the steering
-Turned down the overpowered pushback thruster
-Slightly changed light mappings
-Modified the autobrakes to use a separate prop for the autobrakes, it's bound to the gears and allows the script to not mess with the keyboard or joystick brakes, which it sometimes did.
-Added lightkit and painting instructions with instructions for lights.
-Scaled down light objects (the things mapped to transparent.rgb)

Some questions:
-For 2.0 users, are there any problems other than those already mentioned, like sound acting weird while switching the views? The sound and selection of stuff is currently synced with 1.9.1, which doesn't have model view.
-What do you think of the light textures included in this release? Mostly the tail (logo lights), but also the other lights.
-ANYTHING else that could be fixed. Those little things. I'm looking forward to a front-page release in time for FG 2.1, so I want to fix the little bugs.

And also...
New 787 Release!

http://www.mediafire.com/?hbzhmk0c5fcuck6

This is mainly to combine Dany's FDM touchups and AP with my systems and model.

-FIXED THE TRANSPARENCY PROBLEM! Yes, removing the alpha channel fixed it.
-Slightly modified the engine startup script, so that people like Skyop can't cheat the engine startup
-Modified the gear rotation animations a bit
-Pushback animations work, similar to what I did to CRJ
-Added the modified CRJ autobrakes script
-Scaled down light objects (the things mapped to transparent.rgb)

But there are some things, mostly for Dany, to work on. On approach with 15% fuel and full flaps, I had a negative AoA. This was the most alarming. I also saw that the AP took a bit of time to pull me to the right position during the climb. Also, I added a couple things to the FDM:

Code: Select all

      <control-input axis="/controls/gear/autobrake" control="BRAKE"/>

Needs to be added to each main gear in case you've done any unposted FDM work. I changed this because the autobrakes script kept on disengaging my joystick/keyboard activated brakes, so now these can independently be kept at 0.

The only thing left to fix among BOTH airplanes is the autobrakes, or it has something to do with FG 1.9.1. Any time a prop is set to .6, it automatically resets it to .59999999999999999999999999999999999999999 (as it shows up in the property browser). Since this value is used by the autobrakes, it's a problem because they won't disengage properly. I'll have to play around with stuff, maybe putting (ABRAKE < 0.6 and ABRAKE > .59) in place of ABRAKE == 0.6 will help.

Also gotta talk about my new FG blog: [url]nickfg.blogspot.com[/url]. It'll have most of the releases and also some random other things that I'm doing with FG.

[boeing 787-8]

[skyop]

-For 2.0 users, are there any problems other than those already mentioned, like sound acting weird while switching the views? The sound and selection of stuff is currently synced with 1.9.1, which doesn't have model view.

They all seem good to me, but the CRJ200 uses internal sounds in Fly-by View- I always found that a bit weird.

-ANYTHING else that could be fixed. Those little things. I'm looking forward to a front-page release in time for FG 2.1, so I want to fix the little bugs.

If you switch to Copilot View in the 787, all the instruments disappear! Same with the CRJ200, although only the PFD is affected.

-Slightly modified the engine startup script, so that people like Skyop can't cheat the engine startup

As for the autobrake 0.6 problem, it seems to work perfectly under Git. I guess it's 1.9.1-specific. Perhaps you could consider using more standard setting values? Usually it's:

Code: Select all

-2  RTO
-1  Off
0   Disarmed
1   Power level 1
2   Power level 2
3   Power level 3
4   Power level 4
5   Maximum power

boeing 787-8's problem is completely absent on my machine. All pushback animations are working properly.

Oh, and one question: how the heck do you turn off the engines in the 787-8? They are running on startup on my FG version, and without a usable cutoff switch in the cockpit I can't figure out any other way to switch them off...

[nickyivyca]

Oh, and one question: how the heck do you turn off the engines in the 787-8? They are running on startup on my FG version, and without a usable cutoff switch in the cockpit I can't figure out any other way to switch them off...

Click the button in between the two engine start knobs. But make sure to spool up the APU first, and click on the generator. Otherwise, the cockpit will go black. As for the autobrake thing, it's a lot easier with the way mine does-that way I can just set the script to set the autobrakes prop to what the autobrakes setting is, except for RTO, which is separate. I could switch around the order, but then I'd have to switch around a lot of stuff in the script. I might do that after I get these on the front page.

As for boeing-787-8's problem, that happened in the release before this one. But I fixed it.

The 787 cockpit problem is a known thing-I just have to modify the select animations, or switch to a distance one. Or I was thinking...So in the views in the setfile, I was thinking that you could add a couple props to them. There already is an internal property, but I don't know what else that's used for.

Code: Select all

<!--cockpit view-->
<view n="0">
 <cockpit type="bool">1</cockpit>
 <internal type="bool">1</interior>
</view>
<!--helicopter view-->
<view n="1">
 <cockpit type="bool">0</cockpit>
 <internal type="bool">0</interior>
</view>
<!--coach class, over the wing-->
<view n="101">
 <cockpit type="bool">0</cockpit>
 <internal type="bool">1</interior>
</view>

I could animate and do sound with that, I just need to know if it'd work.