Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[WxTech]

Hopefully we can get the hooks issue straightened out across the board. I suspect that the mod ship makers over these many years have had no familiarity with the way the _Left, _Centre and _Right hooks are utilized. They got lucky in some respects due to simplicity of the scheme for their use, which did not suffer from instances of the L-R reversal. And the starting points of the wakes was usually masked by the hull.

Thats correct. I for one never really understood the Left Centre Right hooks.   Jus made it work by trial and error.

A short concise guide to the correct way to use _Prop _Nose _Left _Centre _Right would be useful.  I am wondering if I need to amend any of my past efforts?

I'll cook up a tutorial! It would be good to have the work to fix this issue spread out among more than one taxed shlub! 

[cbradbury]

I'm running the most up to date BAT version - Flashpoint with the two hotfixes, Glenn.

[WxTech]

Here's some Hook orientation info, just to get something out there right away...

Here is a sample [HookLoc] section, which typically resides in Hull2.msh:

[HookLoc]
0 0 1   0 -1 0   1 0 0   0.0 11.243 22  //Vapor;  L B U
0 0 1   0 -1 0   1 0 0   0.0 38.3886 22  //Vapor1
0 -1 0   1 0 0   0 0 1   0.0 -108.5 -1.9  //_Nose
0 1 0   -1 0 0   0 0 1   0.0 -35 -1.9  //_Centre
1 0 0   0 1 0   0 0 1   12 -35 -1.9  //_Left
-1 0 0   0 -1 0   0 0 1   -12 -35 -1.9  //_Right
0 1 0   -1 0 0   0 0 1   0.0 132.7 -1.9  //_Prop


The "L B U" in the top line indicates my determination of the XYZ values as increasing in value in the Left, Right and Up directions, respectively. That is, X increases to the Left, Y increases to the Back (stern) and Z increases Upward. Notice how all Z values for all the hooks in question are -1.9, sharing a common Z height on the waterline. And the Vapor hooks, for stack smoke, are at 22, which tells up that Z increases upward.

Because a ship is typically longer than it's wide, the largest difference in coordinate values will usually be along the ship's length, between the _Nose and _Prop hooks. Here that axis is the Y-axis, with figures of -108.5 and 132.7 making for a ~240m length.

If the hull mesh is at least near to being constructed symmetrically about the longitudinal axis, The _Left and _Right hooks will have similar values; one positive and the other negative. Here we have X values of 12 and -12; the _Left value being positive tells us that the X axis increases toward the left.


Below is a look-up table that should work in at least the majority of cases. It restricts to Z always increasing upward, as well as the longitudinal axis (along the ship's length) being the Y axis. [I've added for asheshouse his orientation scheme of FLU.]

The foregoing illustrates how to determine the axis orientation and the directions on those axes in which the values increase/decrease. Once you determine the XYZ axis situation as either of LBU, RBU, LFU, RFU or FLU, find the appropriate set of hook orientations and apply them accordingly. Naturally, only the 9 matrix transform values are supplied here; be careful to not overwrite the XYZ values following on!

Code: [Select]

L B U +X = LEFT +Y = BACK +Z = UP
_Nose 0 -1 0 1 0 0 0 0 1
_Left       1 0 0 0 1 0 0 0 1
_Centre 0 1 0 -1 0 0 0 0 1
_Right -1 0 0 0 -1 0 0 0 1
_Prop 0 1 0 -1 0 0 0 0 1


R B U       +X = RIGHT +Y = BACK +Z = UP
_Nose 0 -1 0 -1 0 0 0 0 1
_Left       -1 0 0 0 1 0 0 0 1
_Centre 0 1 0 1 0 0 0 0 1
_Right 1 0 0 0 -1 0 0 0 1
_Prop 0 1 0 1 0 0 0 0 1


L F U       +X = LEFT +Y = FRONT +Z = UP
_Nose 0 1 0 1 0 0 0 0 1
_Left       1 0 0 0 -1 0 0 0 1
_Centre 0 -1 0 -1 0 0 0 0 1
_Right -1 0 0 0 1 0 0 0 1
_Prop 0 -1 0 -1 0 0 0 0 1


R F U       +X = RIGHT +Y = FRONT +Z = UP
_Nose 0 1 0 -1 0 0 0 0 1
_Left       -1 0 0 0 -1 0 0 0 1
_Centre 0 -1 0 1 0 0 0 0 1
_Right 1 0 0 0 1 0 0 0 1
_Prop 0 -1 0 1 0 0 0 0 1


F L U       +X = FRONT +Y = LEFT +Z = UP
_Nose 1 0 0 0 1 0 0 0 1
_Left      0 1 0 -1 0 0 0 0 1
_Centre -1 0 0 0 -1 0 0 0 1
_Right 0 -1 0 1 0 0 0 0 1
_Prop -1 0 0 0 -1 0 0 0 1


Very briefly, in other respects...

I would suggest that the _Centre hook lie on a line between the _Left and _Right hooks. It should not be situated any noticeable distance farther forward of or behind this line.

Unless there is a reason to make a wake noticeably narrower than the hull, The _Left and _Right hooks should lie not too far inside the wall of the hull. For a full wake width, the hooks should be at least 85-90% of the distance between the centreline and the hull face.

For best fitting of wake and side waves, The _Left and _Right hooks should be neither too far forward nor too far aft. In most cases, a position about 1/3 to 1/2 the distance between _Nose and _Prop is fine. Or a position at about where the hull initially reaches its widest behind the bow--if this point is well aft! Otherwise the 1/3 rule is good.


Hopefully this will provide assistance in dealing with hook fixing for wakes.

[WxTech]

Unfortunately, yes. If you just copy over the BigShipGeneric folder from 1.3 over the CV shows in FMB, but not in the mission i.e. attach aircraft to carrier, run the mission, no flattop, splash in the drink!

Clive,
I forgot about the other implementation of sidewave effects in v1.3. You were missing these effects which were in use with my older scheme in bigShipsGeneric.class for v1.3:

3DO/Effects/Tracers/ShipTrail/SideWaveLarge.eff
3DO/Effects/Tracers/ShipTrail/SideWaveMedium.eff
3DO/Effects/Tracers/ShipTrail/SideWaveSmallm.eff
3DO/Effects/Tracers/ShipTrail/SideWaveTiny.eff

Here are the necessary files to copy over into the requisite destination in the 'main' Effects v1.4 folder (SideWaveC.tga should already be present; no need to keep this one):


https://www.mediafire.com/file/qke098lam8vk9z2/Old_sidewave_effects_v1.3.7z/file


These files can be left in place even if you go back to the v1.4 bigShipsGeneric.class; there's no conflict.

[asheshouse]

My models are orientated x front, y left, z up. So i guess i need to vary the matrices.

[WxTech]

Ashe,
I'll work out a set of appropriate matrices for your coordinate system orientation. Stand by...

I've added the matrices for the F L U coordinate orientation to the end of the table in reply #26.

Here it is by itself:

Code: [Select]

F L U    +X = FRONT          +Y = LEFT          +Z = UP
_Nose 1 0 0 0 1 0 0 0 1
_Left   0 1 0 -1 0 0 0 0 1
_Centre -1 0 0 0 -1 0 0 0 1
_Right 0 -1 0 1 0 0 0 0 1
_Prop -1 0 0 0 -1 0 0 0 1


[cbradbury]

That fixed it, Glenn, thanks.

Just to clarify for anyone else following this increasingly convoluted thread, I am now running effects v 1.4, but with the BigShipGeneric folder from v1.3, which gets around the wake issues.

But to do this, you must also add the files Glenn attached to Post #27.

[WxTech]

Here's a visual guide to handling hook re-orientation. (Click for larger.) This screen capture is taken from the mesh viewer utility. It shows the hooks as cubes, with the +X face being bright red, The +Y face being bright green, and the +Z face bright blue. I hope this makes it easy for you to construct appropriate transformation matrices.

I'm using here the common stock ship model orientation, with X increasing to the left, Y increasing to the back, and Z increasing upward (the 'LBU' scheme).

For every ship having these same 5 hooks under consideration, the axis orientations will always be as seen here. The hook +Z-axes (blue faces) all point up. The +X-axes (red faces) point as follows:
_Nose - forward
_Left - left
_Centre - back
_Right - right
_Prop - back (this is always identical to _Centre, handily)
The +Y-axes (green faces) are rotated 90 degrees CCW as seen from the +Z-axis direction (from above.)

The transformation matrix comprises 3 groups of 3 numbers. The first triplet orient the object's X-axis with respect to the model, the second triplet the object's Y-axis and the last triplet the object's Z-axis. Because all these hooks are given purely orthogonal orientations, the values will always be 0, 1 or -1. And a triplet will always have two zeros.

In the treatment below I separate each triplet for clarity. The fully constructed transformation matrix will have the upper triplet followed by the middle triplet followed by the bottom triplet, as a series of 9 numbers in sequence. You can have several spaces, or use TAB characters, as separators if you feel this aids readability in the [HookLoc] section of the mesh.

It's hoped the illustration will obviate any additional verbiage. If you have questions, please ask!

[asheshouse]

Great guide. Thanks.

[WxTech]

I forgot to mention that the X-axis is the 'action' axis if, say, the emission of particles is tied to a hook. In the ship model above, note the two guns at the bow. Their hooks, tied to the muzzles, have the X-axis directed in line with the barrels.

Any muzzle fire and smoke effect assigned to them will have the emitted particles directed symmetrically about the X-axis. If the effect were to have  EmitTheta 0.0 0.0, the particles would be emitted exactly along this axis. If  we had, say, EmitTheta 5.0 15.0, the particles would be emitted also symmetrically about this axis, but always between 5 and 15 degrees divergent from the axis, never less and never more. If we had EmitTheta 90.0 90.0, the particles would be emitted perpendicular to this axis, in a random direction, collectively forming a kind of circular 'sheet' perpendicular to the X-axis/gun barrel.

[WxTech]

Here are the 29 ships in BAT 4.0 WAW that I've hauled out of the line for a hooks refit. 

[WxTech]

I'm already hitting the halfway point on fixing the hooks for those 29 ships.

The Warspite presents a problem. It does show up in game, but it has issues, or there's something not being spit out in the dump, and it crashes the mesh viewer utility. That means I'll have to bring the spreadsheet to bear, and plot the 3 hull sections, suitably positioned to as to join into a whole, so that I can visualize where hooks must be moved.