[Resource | Code] Conical bullet spread! *UPDATED*

[Snarboo]

Here's a quick and dirty conical bullet spread example! Now it's been updated to correct the fixed point issue, so bullet patterns should no longer form bullseye patterns. Be sure to redownload the file!

Code: Select all • Expand view

/*
** Conical Bullet Spread Example
** Formula works like this:
**      X = (Radius * Cos(Angle)) + Offset
**      Y = (Radius * Sin(Angle)) + Offset
**
**      Where Radius is the size of the circle you want to draw (can be
**      randomized), Angle is an angle between 0 and 360 (can be fixed or
**      random), and Offset is how offset the circle is from your crosshair.
**
**   NOTE: Two of these values must be identical between X and Y! Radius must be
**        the same and angle must be the same! However, if you want elliptical
**        bullet patterns, simply change the X or Y value to be larger or
**        or smaller than the other.
**
**        Weapons that make use of this MUST use the EXPLICITANGLE flag! Same
**        goes for the equivalent monster function A_CustomBulletAttack.
**
**        Conical spreads are a bit limited with weapons, since weapons do not
**        have acces to user variables, but they do have access to arguments.
**        There are problems relating to fixed point numbers and arguments, but
**        a workaround exists by using the following formula:
**        X = (((Radius * 10) * Cos(Angle)) * 0.1) + Offset
**        Y = (((Radius * 10) * Sin(Angle)) * 0.1) + Offset
**
*/

---

Is there any advantage to this over the classic spread?
Conical spreads are tighter than square shaped ones, so your weapons should be slightly more accurate using this method. This method can also be used to make fixed, circular patterns. Otherwise, this is mostly for those that prefer "realistic" bullet spreads.

Does this work with missiles?
Theoretically yes, but I haven't tinkered with that yet and I'm not sure which values this method applies to.

Do you have an example file?
It's attached to this post!

Did you come up with this on your own?
No, I consulted this article for help!

---

• Cone gets more skewed as you look up or down.
• Args[] do not seem to store float values, which creates "bullseye" patterns. - Fixed!

[Nash]

Looks like one of those mysterious alien pattern thingies you find in obscure places... =D

[Snarboo]

I MAY have done the math slightly wrong! :p But yeah, I recommend using floats for this so that doesn't happen.

Edit:
Actually wait, can arguments store floats?!

[Nash]

Also, and you may already know this but, this doesn't seem to respect player pitch... in OpenGL with an extreme view pitch, the cone doesn't work...

[Snarboo]

Yeah I've definitely screwed up the math somewhere! I'll look into it.

[Enjay]

Also, and you may already know this but, this doesn't seem to respect player pitch... in OpenGL with an extreme view pitch, the cone doesn't work...

You're not running a build with this bug are you?

http://forum.drdteam.org/viewtopic.php?f=25&t=6492

That has now been fixed if you are.

[Nash]

Enjay: no, my screenshot is ZDoom and the software renderer. :D And the behaviour of this conical spread not working correctly with extreme view pitch is also on an old GZDoom build without pixelratio.

(I haven't updated to any of the new pixelratio builds yet)

[Snarboo]

The cone does get very skewed at greater pitches, but I'm not sure how to fix that if it can be fixed.

[Enjay]

Enjay: no, my screenshot is ZDoom and the software renderer.

...but I was responding to

in OpenGL with an extreme view pitch, the cone doesn't work...

[Snarboo]

I just tried this with a fixed square spread, and the same thing happens: It gets very skewed as you look up and down. Wonder if this is a limitation of the hitscan code, or a limitation of FBF_EXPLICITANGLE?

Edit: Offset doesn't have to be the same with both values, so I updated the first post! I'm also looking into some way of correcting the cone at extreme pitches.

Edit 2:
I did some simple tests in other ZDoom games with different weapons, and it seems that at extreme angles, there is no horizontal spread at all! I'm not sure I can fix this issue, but given other weapons behave the same way, I guess I don't need to.

[Snarboo]

I've fixed the "bullseye" issue, so be sure to redownload the example file!

Edit:
Fixed a math related error.

[TiberiumSoul]

sorry for bumping this but... did anyone ever figure out how to get missiles to work with this?