damage textures again

[Callahan]

Just looked after this.
The first 24 bytes of scar data seem to position them, the second 24 bytes are for shape and size.
(intermixed them from 2 vanilla models into your test model)

Did you add the slots in the Creator before calculating the BSP nodes ?
I think this is needed to get them displayed on top of the model texture ingame.
The scars I pasted in your test model appeared behind the ship at all times, even when they where close to the eyepoint.

I cannot test this yet, as I´m still learning on how to work with the Creator, but it may be worth giving a try.

Edit:
Just looked for the switches. You can create a switch, then an object. You may attach the object to the "SWn" switch. You may attach faces to a switch via the object on it.

Could I get some model FLT´s and/or 3Ds files to experiment with ?
The Creator tutorial FLT´s seem to be unusable for HOB conversion.

[thunderchero]

Just looked after this.
The first 24 bytes of scar data seem to position them, the second 24 bytes are for shape and size.

that statement does not make sense to me. why editing the first 12 bytes did not effect the other 11 scars.

(intermixed them from 2 vanilla models into your test model)

I have not tried this. but plan on looking at more vanilla scar data.

Did you add the slots in the Creator before calculating the BSP nodes ?
I think this is needed to get them displayed on top of the model texture ingame.

yes and no slot were there with model then model separated by "face" then all slot objects separated by object.

The scars I pasted in your test model appeared behind the ship at all times, even when they where close to the eyepoint.

I have never experienced this they are always in vanilla position or wrong position/expanded. lol

Edit:
Just looked for the switches. You can create a switch, then an object. You may attach the object to the "SWn" switch. You may attach faces to a switch via the object on it.

My last attempt used this but made no difference how data was added in phaser array.

Could I get some model FLT´s and/or 3Ds files to experiment with ?
The Creator tutorial FLT´s seem to be unusable for HOB conversion.

without you knowing more on how to convert this may only confuse you but I will send you my latest attempt folder. I will PM you a link in a few min

thunderchero

[Callahan]

Thanks for the materials !

Confused ?!?? Yep. I think I am confused, but I´m not quite sure - maybe.

Just added 31 slots and 12 switches to see the converter stores them in 2 arrays - not 1 as with the vanillas.

So I think scars are actually using slots, but with data of other format.

Just looked after this.
The first 24 bytes of scar data seem to position them, the second 24 bytes are for shape and size.

that statement does not make sense to me. why editing the first 12 bytes did not effect the other 11 scars.

Each scar is using 48 bytes. I replaced the first 24 of each entry with HM3 data and the last 24 of each with RM1 data. Plugged it into a model to see it generates RM1 scars on HM3 positions.

I may figure out placement, but with having scars (visually) placed behind a ship even if the ship is 100 miles behind the scar location, this is of no use right now.

If I could just make them appear ON the hull instead of behind I would even place them by hand, editing all the HOBs.

As you wrote, my slots got diced up and renamed, so I will try to add new slots to the finished BSP nodes next time.

[thunderchero]

this is the closest I have gotten.



when rotated it shows on top too (2 sided) this should not happen and some angles it disappers

thunderchero

[Dr_Breen]

i guess because your damage texture is not part of the models bsp tree, so botf cannot tell wether it is on top or bottom of the miranda as it lacks a z-buffer.
I'd like to try it myself.
can you post those 48 bytes in here? or where in rm1 did you find them?

[thunderchero]

i guess because your damage texture is not part of the models bsp tree, so botf cannot tell wether it is on top or bottom of the miranda as it lacks a z-buffer.
I'd like to try it myself.
can you post those 48 bytes in here? or where in rm1 did you find them?

The scar data is 576 bytes (48x12) if model has 12 scars (most do) on vanilla models.

the data starts at 0x00668 (1640) in rm2_a.hob

also yes I did try to name slot to dam01 ect they are ignored since only (slot, sw, lod) are supported by converter.

thunderchero

[Callahan]

Excellent job, Thunderchero !

Your H3 A3.hob seems to be perfect for my future tests.

Pasted vanilla HU scars on it and they displayed just fine - top, below, side view.

As I can see them now, I can continue my efforts at positioning.
I already altered Vanilla scars to be seen from 2 sides. ( by accident) I think this is mainly a matter of scar data from now on.

I´ll start with positioning right away.

[thunderchero]

On test model I am using I used the vanilla hl2_a.hob full scar data (576 bytes) added to test model.

I have tested several ways now and if I only replace the first 12 bytes with the X Y Z data. it is seen it on top of model but not on bottom. (same as vanilla scar)

I also tested replacing first 144 bytes (first 12 bytes repeated 12 times) no scars could be seen moved on to model but 3 vanilla scars were no longer shown. (like repeated values cancels out each other)

So we need to analyze the first 48 bytes and figure out what each section does.

first 12 bytes moves position from my tests

so much more testing is needed.

thunderchero

[Callahan]

I think I got em all.

I refer the values as SETs of 4 bytes each.
Y axis is bow-stern
X axis is port-starbord
Z axis is up/down

Data sets are for:
1 - X pos
2 - Z pos
3 - Y pos
4 - Y axis camera border for display/skip scar
5 - X axis camera border
6 - Z axis camera border
7 - stretch Y while slim down X
8 - stretch X
9 - roll around Y axis
10 - stretch Y
11 - rotation / roll Z axis
12 - roll X axis

Based on camera border, scars will be visible from both sides.

Can you confirm this ?

[thunderchero]

I think I got em all.

I refer the values as SETs of 4 bytes each.
Y axis is bow-stern
X axis is port-starbord
Z axis is up/down

Data sets are for:
1 - X pos
2 - Z pos
3 - Y pos
4 - Y axis camera border for display/skip scar
5 - X axis camera border
6 - Z axis camera border
7 - stretch Y while slim down X
8 - stretch X
9 - roll around Y axis
10 - stretch Y
11 - rotation / roll Z axis
12 - roll X axis

Based on camera border, scars will be visible from both sides.

Can you confirm this ?

I will try but will take some time

about 1,2,3 could this be start point (first vertex of polygon used)?
about 4,5,6 could this be the BSP plane used by polygon?
about 7 could this be distance from start point on Y axis
about 8 could this be distance from start point on X axis
I would think they would keep sets together so maybe reverse 9 & 10?

Right now statement above is my logical thought. (BOTF is not very logical) so I will test your finding first and post what I find.

If my 's is correct the hot file may give all info needed with some adding and subtracting then converting data to hex. (I guess not)
from data of 1,2,3 all scars will need to be added manually. this will make things much harder.

EDIT;

I can confirm 1,2,3 they are center point of scar (not corner as I thought)

I also tested 4 and 7 many times but either had no change or froze my computer (when testing 7 I had to sleep to end program control-alt-delete did nothing)

Note; for best viewing of scars I edited the scar1.gif to all white.

this also displayed scars and noticed no rotation of any scars (all scars were square to center line of model) I have seen roll but not tested/found those values.

A strange thing I noticed scar will cause shadows on other scars

I will continue testing tomorow

thunderchero

[Callahan]

about 4,5,6 could this be the BSP plane used by polygon?
about 7 could this be distance from start point on Y axis
about 8 could this be distance from start point on X axis
I would think they would keep sets together so maybe reverse 9 & 10?

Right now statement above is my logical thought. (BOTF is not very logical) so I will test your finding first and post what I find.

I also tested 4 and 7 many times but either had no change or froze my computer (when testing 7 I had to sleep to end program control-alt-delete did nothing)

thunderchero

For 7,8 & others. I think most of them are floats for scaling/stretching in %. Misuse of them can create scars to be a bit - oversized, causing a crash/freeze.
It´s most save to look at other models to figure out the valid values.

For 9 & 10. They appear to be right, but it´s a bit tricky. Turning scar 90 deg. seems to make Y stretch become X stretch effect.
Some kinda model of it´s own.

For 4,5,6 - I think this is %tual from the model length/beam and defines the relative border on which a scar is not shown anymore. But it could also be a flat -+ value from the center model for this use.
I´m not yet sure on this one.

[thunderchero]

I think you are correct about 4,5,6 Just have no clue yet best way to set this value.

8 - stretch X correct size port-starbord
9 - roll around Y axis my test showed roll on X higher port lower starboard
10 - stretch Y correct size bow-stern
12 - roll X axis my test showed roll on Y higher stern lower bow

7 & 11 I saw no change switching from known good values 00 00 00 00 may work for all new scars?

thunderchero

[Dr_Breen]

you guys sure did a great work figuring out the scar thing, however adding this way would require you about 1 hour o work for adding 8 scars to 1 hob only together with all the try and see how it looks ingame.

there has to be a way how it's done in creator, as the original botf creators sure did it that way as well

[thunderchero]

you guys sure did a great work figuring out the scar thing, however adding this way would require you about 1 hour o work for adding 8 scars to 1 hob only together with all the try and see how it looks ingame.

there has to be a way how it's done in creator, as the original botf creators sure did it that way as well

I hope a way can be found or you are right it is not worth it.

it took an hour to do just this 1 scar and it is not perfect



thunderchero

[Dr_Breen]

it does look good though. i remember jigallypuff figuring out how to do the phasers in creator after dcer wrote how it's done in a HOT file. perhaps he can help