effect won't stay aligned to aircraft

[Voodoo]

I have a simple planar image effect that I want to align with an aircraft so that it remains parallel to the aircraft wing at all times.

The problem is that the effect will pitch ok with the aircraft but always seems to want to rotate and align itself to the ground when the plane is banked. (see attached images)

Altering Pitch, Bank and Heading parameters in the .fx file makes no difference (other than when the plane is sitting level on the ground).

I do not have the orginal model so I cannot do anything with attachpoints or suchlike in gmax or 3ds etc.

Is it possible to achieve what I want through simple .fx files?

I'd be very grateful for any help.

 

[Foiter]

I'd like to join up to this question which is presumably addressed to the developers team responsible for the effects subsystem in FSX. I don't know whether it is a known bug or a "designed feature" but it seems that whenever the effect is attached to the aircraft via AttachTool in GMAX/MAX its rotation matrix used to render the effect is not multiplied by aircraft's attitude matrix (Roll, Pitch, Yaw).

 

[Misho]

I have a simple planar image effect that I want to align with an aircraft so that it remains parallel to the aircraft wing at all times.

The problem is that the effect will pitch ok with the aircraft but always seems to want to rotate and align itself to the ground when the plane is banked. (see attached images)

Altering Pitch, Bank and Heading parameters in the .fx file makes no difference (other than when the plane is sitting level on the ground).

I do not have the orginal model so I cannot do anything with attachpoints or suchlike in gmax or 3ds etc.

Is it possible to achieve what I want through simple .fx files?

I'd be very grateful for any help.

Indeed, this aspect of the effects is broken and sorely in need of fixing. I have gone through all possible permutations and trials, defining orientation in CFG and FX files, without success. The only way to have the effect behave correctly is to have it in its default orientation (upright, facing forward/back) without having any rotations defined. Introuducing any kind of an angle to your effect will "break" the orientation relationship with the aircraft. This is unfortunate omission on part of MS, and I'm sad to notice it hasn't yet been fixed.

The only thing I can suggest is to use multiple instances of "fuzzballs" (a cloudy ball with whispy edges) to create a volumetric effect of what you're trying to accomplish. This way, you won't have to introduce any rotations - you only need correct positions of the fuzzballs. This technique was used to create volumetric light cone effects, some afterburner effects, and I used it to create thruster plumes on my lunar landers. Combining them with animation (for instance, cycling a fuzzball through a series of brighter/lighter/coarser/fluffier versions) will give you additional eyecandy that, with some careful attention, might turn out to look amazing.

You'd be surprized what can be built out of a bunch of fuzzy balls

 

[Foiter]

The only thing I can suggest is to use multiple instances of "fuzzballs" (a cloudy ball with whispy edges) to create a volumetric effect of what you're trying to accomplish.

I guess that's exactly what people do trying to overcome the problem while creating realistic afterburner or engine exhaust effects. This actually works untill one tries to attach an effect to some geometry in its aircraft model. And THIS is where the problem is revealed in its nature. Either the stack of rotation matricies used to render the geometry itself is ignored when the effect is rendered or the order in which those matricies are multiplied is wrong.

I have an aircraft model with animated thrust vectoring nozzles and I'm "attaching" the engine exhaust effect to the nozzle itself. The effect is triggered by the SMOKE key.
Whatever nozzle angle I set the exhaust direction is unchanged. The particle system used in FSX effects uses the s.c. "particle velocity" property (which is a 3D vector) and exactly this vector stays unchanged with the respect to the aircraft (local) reference system. The system where the Pitch, Yaw and Roll angles are zeros.

I'd really love someone from FSX team to comment on this, because I guess this issue deserves enough attention already since the times of FS2004 and maybe earlier. If this is not a bug, rather than an "intentional feature" then the only explanation I could possibly find is a perfomance issues that may arise during the effects rendering IF the whole matrix stack is taken into account for every single FX particle.
....BUT, I'm sorry, maybe I missed something, why would one need the "Attach Tool" for the aircraft model if the only way to use it seems to be a static scenery?

Has anyone tried to use Attach Tool to hang lights on the aircraft? Maybe I do something wrong I don't know...

 

[Misho]

I guess that's exactly what people do trying to overcome the problem while creating realistic afterburner or engine exhaust effects. This actually works untill one tries to attach an effect to some geometry in its aircraft model. And THIS is where the problem is revealed in its nature. Either the stack of rotation matricies used to render the geometry itself is ignored when the effect is rendered or the order in which those matricies are multiplied is wrong.

I have an aircraft model with animated thrust vectoring nozzles and I'm "attaching" the engine exhaust effect to the nozzle itself. The effect is triggered by the SMOKE key.
Whatever nozzle angle I set the exhaust direction is unchanged. The particle system used in FSX effects uses the s.c. "particle velocity" property (which is a 3D vector) and exactly this vector stays unchanged with the respect to the aircraft (local) reference system. The system where the Pitch, Yaw and Roll angles are zeros.

I'd really love someone from FSX team to comment on this, because I guess this issue deserves enough attention already since the times of FS2004 and maybe earlier. If this is not a bug, rather than an "intentional feature" then the only explanation I could possibly find is a perfomance issues that may arise during the effects rendering IF the whole matrix stack is taken into account for every single FX particle.
....BUT, I'm sorry, maybe I missed something, why would one need the "Attach Tool" for the aircraft model if the only way to use it seems to be a static scenery?

Has anyone tried to use Attach Tool to hang lights on the aircraft? Maybe I do something wrong I don't know...

Someone from MS DID comment on the Effects engine on this very forum... and I paraphrase: "the effects engine is a mess".

Therefore: All effects attached to the aircraft will properly rotate with the aircraft, but will not rotate if attached to the rotating geometry of the aircraft. So, as far as the AC frame reference, they are fixed.

In voodoo's case, it would be possible to assemble fuzzballs to create a leading edge vapour effect (which is what I think the intention is) but creating an exhaust from a variable pitch exhaust nozzle would not be possible...well, not quite:

It would be possible to create a series of FX files (lets say, for every 5 degrees), each with a different angle of exhaust. Then, you could activate and deactivate those by writing a gauge that checks for the pitch variable and uses the appropriate range fx. I've done this before for my lander thrusters, except it was the scale, not the pitch, but the same principle applies.

 

[n4gix]

Someone from MS DID comment on the Effects engine on this very forum... and I paraphrase: "the effects engine is a mess".

Therefore: All effects attached to the aircraft will properly rotate with the aircraft, but will not rotate if attached to the rotating geometry of the aircraft. So, as far as the AC frame reference, they are fixed.

That's not always true either. I have several a/c with 'Attached' .fx nav and strobe lights. They are perfect as long as the a/c is on the ground.

As soon as the a/c is > 50' AGL, the "lights" are shifted inboard of the wingtips...

 

[Misho]

That's not always true either. I have several a/c with 'Attached' .fx nav and strobe lights. They are perfect as long as the a/c is on the ground.

As soon as the a/c is > 50' AGL, the "lights" are shifted inboard of the wingtips...

Ack! that must be as frustrating as your icon indicates... I have never attached lights for my AC... I always define them in CFG file, where you can specify the x,y,z pos. What's the advantage of "attaching" them? (probably something obvious that I'm missing)

Another tidbit of wisdom that I discovered the hard way is, if you are positioning effects in cfg file (for example, "smoke" section) and you declare them to be at 0,0,0 and then do your offsets in the fx file, all kinds of problems crop up (misalignments and unexpected shifts). However, if you introduce a minute values, as in 0.01,0.01,0.01, everything works properly...

 

[Foiter]

It would be possible to create a series of FX files (lets say, for every 5 degrees), each with a different angle of exhaust. Then, you could activate and deactivate those by writing a gauge that checks for the pitch variable and uses the appropriate range fx. I've done this before for my lander thrusters, except it was the scale, not the pitch, but the same principle applies.

This will do, but... I have 2DOF animation for the thrust vectoring, so it is not only a pitch, it is yaw as well, just like a swivel nozzle of F-35B. Actually no matter how complicated the thing is, it could be done with multiple FX files, I agree, but I wouldn't go that way either, since it gets too complex whenever you introduce a second degree of freedom....

well, what can I say, I'm a programmer as well, I know it is not easy to write a code covering FSX's complexity level. Looks like the only way to fix all FSX issues we have so far is to go "open source", at least on modular basis.
As far as I see FSX team is hardly capable of making this product the way it's meant to be.

 

[Misho]

This will do, but... I have 2DOF animation for the thrust vectoring, so it is not only a pitch, it is yaw as well, just like a swivel nozzle of F-35B. Actually no matter how complicated the thing is, it could be done with multiple FX files, I agree, but I wouldn't go that way either, since it gets too complex whenever you introduce a second degree of freedom....

well, what can I say, I'm a programmer as well, I know it is not easy to write a code covering FSX's complexity level. Looks like the only way to fix all FSX issues we have so far is to go "open source", at least on modular basis.
As far as I see FSX team is hardly capable of making this product the way it's meant to be.

I wouldn't agree... they admitted the FX engine is a mess. Everything else is pretty darn good... the amount of work needed to make it work with add-ons and user-generated content multiplies the complexity by a factor of 10. I have my fingers in most of the FSX aspects and I am still amazed at the flexibility and the power... For example, using SimConnect, you can program FSX to behave as almost anything you want... ship simulator, tank simulator... you name it.

 

[Foiter]

I wouldn't agree... they admitted the FX engine is a mess. Everything else is pretty darn good... the amount of work needed to make it work with add-ons and user-generated content multiplies the complexity by a factor of 10. I have my fingers in most of the FSX aspects and I am still amazed at the flexibility and the power... For example, using SimConnect, you can program FSX to behave as almost anything you want... ship simulator, tank simulator... you name it.

good or bad, it depends on what the product is aimed at. For the most of purposes FS had from its very begining it does quite well, until someone tries to push the thing beyond its "flight envelope".

you see, most of people here have already learned how to take off and land without "mission reload" in between, but that's it, switching to another craft and doing the same with different instrument constellations or under "bad" weather conditions helps a while, but not for long. Those who learned how to fly now wish to learn recovering from spin, land on the carrier, or fire "fox three" at the "bogie"... which someone would probably name a silly gameplay rather than pure simulation. "Want to fight in close combat?... go for LOCK-ON, Falcon4 or CFS, MSFS is not intended for that...." Well, I see it different, and I know how to make it different.

I see the potential this programm has to evolve into highly customizable tool, but unfortunatelly still not at the core.
Making it safe and satisfying all third-party needs is hard but it is already moving that way, only... not as fast as we need.

This is what I'm doing, tweaking that thing at its core.