Looking for information on nova spell bolt distribution
#1
From Jarulf's guide, nova releases a total of 92 bolts from the caster. I am trying to find out how many bolts are concentrated in each of the tiles adjacent to the caster.

Attika did some testing in the past which showed the spell's dead zones (http://thedsfcommunity.yuku.com/topic/23...ones-quot-). From this, I am guessing that the 92 bolts are not evenly concentrated among the 8 adjacent tiles to he caster, perhaps it is more concentrated on the tiles below the caster.

I remember there was an old site which explained how all the lightning based spells worked, though I cannot remember the link. If anyone can shed some light on this let me know.
Reply
#2
My guess is that it's a result of "pixelation." When you try to draw a straight line with squares, you're not always going to get something that looks pretty. The way Nova works is to send the lightning out in straight lines from the center of a circle, the caster, to some external point, the edge of the circle. But even in making 96 straight lines, you are going to have issues.

Here's a picture I made with MS-Paint, a prime example of drawing with squares https://lh3.googleusercontent.com/-Keha4...xample.jpg

Basically, I made a circle, with a center, and just drew lines from the center to the edge. I targeted every point on the edge of the circle, but you can see how gaps are still present. I'm fairly certain this is the same thing observed with the Nova spell.
Reply
#3
(08-29-2011, 11:08 PM)Brogden Wrote: My guess is that it's a result of "pixelation." When you try to draw a straight line with squares, you're not always going to get something that looks pretty. The way Nova works is to send the lightning out in straight lines from the center of a circle, the caster, to some external point, the edge of the circle. But even in making 96 straight lines, you are going to have issues.

Here's a picture I made with MS-Paint, a prime example of drawing with squares https://lh3.googleusercontent.com/-Keha4...example.jp

Basically, I made a circle, with a center, and just drew lines from the center to the edge. I targeted every point on the edge of the circle, but you can see how gaps are still present. I'm fairly certain this is the same thing observed with the Nova spell.

What you are describing is the result of trying to use integer values to represent real numbers, and yes that is essentially what happens with Nova. However, knowing that does not help answer the question asked.


"What contemptible scoundrel stole the cork from my lunch?"

-W.C. Fields
Reply
#4
(08-20-2011, 05:51 AM)maht1 Wrote: From Jarulf's guide, nova releases a total of 92 bolts from the caster. I am trying to find out how many bolts are concentrated in each of the tiles adjacent to the caster.

Attika did some testing in the past which showed the spell's dead zones (http://thedsfcommunity.yuku.com/topic/23...ones-quot-). From this, I am guessing that the 96 bolts are not evenly concentrated among the 8 adjacent tiles to the caster, perhaps it is more concentrated on the tiles below the caster.

I remember there was an old site which explained how all the lightning based spells worked, though I cannot remember the link. If anyone can shed some light on this let me know.

I think, that from great job that Attika did with his friend, we could try to answer to that question.

What we know from Attika research could be described like that:
[Image: nova1.gif]

When we try to draw straight line to fit into gaps that was researched by Attika we end with something like this:
[Image: nova2.gif]
From, what is drawn we clearly see, that there EXIST straight lines from center (caster) that could go with 'approximately' equal angle from each other, dividing quarter into smaller sectors AND NOT EXISTING IN DEAD TILES (!).

Those straight lines exactly (look at doubts section of this post) avoid dead tiles from Attika's research. That simply means: projectile doesn't exist in those locations.
Test didn't cover that, but I assume that we could spread this solution to all quarters, and get that:
[Image: nova3s.gif]
Quickly counting - there appears that there is not 92 directions, but 32.

Oh, and by answearing the question:
Direction spread like that (as you already saw at above pictures)
[Image: novadistributions.gif]
<click> to get hi res picture.

-----------------
Doubts section
-----------------

1. Why JG says 92? Who knows, maybe there is 96 bolts, but only 32 directions. That would mean that 3 bolts going along in 1 direction. That would be nice expain if there would be 96 not 92 (96/3=32).
It's only theory. Seems someone have to verify it (also could be tested by amount of damage received).

2. There are few tiles that I am not sure off.

a) tiles marked by Attika as dead, but, line is going through them, as you can see on pictures.
b) lines that I am not sure, how many projectiles go through it (tiles).

[Image: novadoubts.gif]

COMPARE IT IN HI RES !!!
<click>

This doubt need more research with true trigonometry calculations (sorry, was too lazy to do that for you...).
Considering Chapter 6.1.10 in JG:

JG says Wrote:6.1.10 Spell and arrow speeds

In the rest of the Guide, the speed of spells and arrows are given with a number in the 16 - 63 range (rarely going below 8). Here, a brief description is made in an attempt to convert those speeds to real speeds in the dungeon measured in reference to the tiles in the dungeon. This is also useful for estimating the distance penalty of arrows.

Any missile in the game (be it an arrow or a spell) will use true trigonometrical calculations for its movement, both for location and speed. Thus, it takes a missile longer to move through a location diagonally than straight along a side (roughly 40% longer) as opposed to character and monster movement which takes the same time regardless of it being diagonally or not. In addition, missiles actually move in small jumps each frame. That is, depending on their speed, they will move a certain distance each frame. For the purpose of keeping track of missiles, the game actually uses 22 extra bits of precision for its location, think of it as each location actually being divided in roughly 92 680 times 92 680 smaller locations. This higher precision is used for smooth movement only, for actual To Hit purposes the missile is simply considered to be in one location, regardless of where in the location it is. This have a few implications though. If the distance a missile is moved each frame is smaller than a location, it may end up staying in a location for more than one frame. In addition, it may, when not moving in straight angles along the locations, only enter the corner of some locations and may thus never really exist in a location that it in fact should pass through (for movement straight along the locations, this should never happen as it requires a speed value above 64 which does not exists in the game).

Some missile types prevents trying to hit the same location more than once (should it not manage to move into a new location when updated each frame) while most do not and may thus try to hit a target in a location more than once. This will of course make the total chance of hitting a target somewhat higher than it should be. This phenomenon is not that uncommon but extremely hard to predict or give exact numbers on as it depends both on the exact speed of the missile and in what direction you fire it. As you can virtually fire a missile in any direction, the possibilities are numerous. One should be aware of this phenomenon though as it might affect the result if anyone would actually measure the actual hit percentages of missiles.

To calculate the distance a missile of a specific speed travels, only examples of movement straight along the locations in the dungeon will be considered. It should be easy to apply to movement in any direction with true trigonometry. A missile with a speed value of 64, will travel exactly diagonally through one location each frame. This corresponds to traveling 1.4 locations in a straight line along the locations side. Alternatively you may say that a speed of 45.3 travels a location in one frame straight along the side of a location. The distance traveled per second can thus easily be calculated as:

Locations a missile travel per second: 25ˇspeed/45.3
As the distance penalty of an arrow is increased by 1 each frame, the total distance penalty per location (see chapter 2.1.4 and 5.1 for more information) can be calculated as:

For characters: distanceˇdistance/2 = (45.3/speed)ˇ(45.3/speed)/2 = 1 048/(speedˇspeed)
For monsters: distanceˇdistance = (45.3/speed)ˇ(45.3/speed) = 2 048/(speedˇspeed)
As noted, for some locations, the fact that the missile will try to hit more than once may make the actual To Hit be slightly different. Also note that a missile will actually start traveling in the location where the player or monster it originates from is currently standing in. It may at times take more than one frame to reach the next location but in no case will a missile make a To Hit check in its location of origin.

we clearly see that those spots could be under law related with speed of projectile and the way it appear in tiles moving along vector.
In short it means that projectile is appearing and disappearing every frame (20 per second) moving along vector. Sometimes (like in
example a) projectile could not appear in tile, because it appeared in previous tile, move a bit and appear in another, passing over "doubt tiles".

More information about disappearing (not existing projectile on its way) in discussion here <click>

----------------
Further analysis
----------------

It could be possible (following picture) that Nova IS targeting some tiles (that would mean center of that tiles) around, but not necessary in circle, but in square.

[Image: novatarget.gif]

also - note, that you, as player won't be ever able to achieve exactly same angles for some of vectors (not straight and diagonal), because those targeted tiles are far above your vision radius (screen size).

--------------------------------------------------------------------
Material presented above is only model of how Nova could work to fit its behavior presented by Attika. I am not sure that it is 100% correct. I totally ignored fact, there are no true lines in reality, and also in diablo (pixels). Also division of tile to 92 680 times 92 680 smaller locations was not counted (could be used for more calculations for 'doubt' tiles).
Reply
#5
Thanks for the analysis Galbraith. I should be able to figure out the bolt concentrations from here. I'll see what results I get from damage tests.

The reason I am looking for the nova bolt concentrations is that I want to know how effective the spell is at point blank range for dueling and PvM purposes.
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)