As title says.
It needs to go. its utterly broken and does not work. It is frustrating to have random non pens on enemy weakspots because the game decides it feels that way.
6 Likes
Coming out with a straight incorrect statement; if you want to pixel shoot again, just be honest.
There are no active bug reports against volumetric, and volumetric works perfectly fine.
There is no randomness behind it either.
You should know the width of your shell.
I do not want to go back to the time when Conqueror could be fragged by hitting the turret ring with a Panther II.
10 Likes
It does no.
I explained that too you 3 years ago in well over 10k words you are just resistant to facts
go too youtube and you will find 10k videos showcasing the issues
Agree, volumetric is not the best but still way more better than the pixel shoot
14 Likes
volumetric is an upgrade to the previous pixel weakspot meta, I prefer some sketchy bounces occasionally to the weird metas that arose around hitting 1mm wide weakspots
3 Likes
t26e1-1 aphe (224mm pen) non-pen jagpanther side
That or shooting barrels and having a 50% chance to not do any damage on a flat hit.
I don’t hate volumetric but the randomness of it can lead to very infuriating stuff.
That and the fact that it makes some vehicles way stronger which as far as I can tell has not been taken into consideration balancing wise.
1 Like
shots 1-5 clearly missed
1 Like
You didn’t shoot the side, you shot roof armor.
Volumetric working as intended with no bug present in your video.
Aim lower or higher next time.
2 Likes
The thing is, the pixel sniping for weakspots represented reality better than volumetric does.
The example Alvis gave?
THAT is how it works in reality. the long 8.8 is fullycapable of going through the conquerors turret ring. The M103 had its ring armour incresed to resist the russian 85mm and the IS 3 had its turret adjusted so low that even a 7.7mm round could not fit.
Volumetric is neither working nor realistic.
You know what woudl work far better? Make it 4-8 pixels instead of 1. Removes all the super weird 1pixel weakspost caused by the old system without any of the issues of volumetric.
never knew that a “sideskirt” is the JPs roof… And even then all your screenshots proof is that the system is broken. The Jp has not point on its armour on the side exeeding 100mm thickness in reality. That is a fact born out of reality. So the game having such a spot is already a bug by definition. Man your copium reserves must run out at some point
cope harder buddy. That is the best documented bug with volumetric in existance.
And your argument about “no bug reports being present”? Go make a bug report regarding volumetric and I 100% gurantee that it is deleated in less than 24h.
Gaijin investigated its own system and found nothing wrong. And you are just coping hard
To actually give Alvis some credit here for once (He’s just bad at explaining it).
What you likely hit there is this armor piece sideways (which i wouldn’t call a “roof” in that sense):
VERY unlucky, but honestly reasonable that it wouldn’t penetrate that amount of metal that it would have had to go through.
Edit:
As people read one post and comment on it before reading the rest of the conversation i’m going to add my answer here as well:
My answer:
3 Likes
Yeah but it seems likely that “in reality” such a thin bit of metal would have deformed and/or the shell deflected slightly - the concept of a much larger shell passing directly through a 16mm thick plate from edge-to-edge is pretty dodgy.
Which is an implementation issue - not a problem with the philosophy of volumetric - volumetric is definitely the way to go, and hopefully it will continue to be improved.
5 Likes
Valid point actually, might be harder to code though, not sure. Though even deflections will decrease the speed a bit in addition to the equivalent mm of armor it passes through while being deflected, that in addition to the round having a risk of starting to tumble depending on a slew of factors might still be enough for the next plate to stop it.
That is however VERY circumstantial and i don’t think even feasible to try to calculate everything on the fly with any reasonable degree of accuracy in the short amount of time it needs to be calculated. But hey, systems improve over time and so does hardware, we can always hope for improvements as you say :)
2 Likes
Except the shell getting absorbed isn’t realistic.
It should ricochet from the edge into the lower side, where it would penetrate, as there’s no way a high velocity 90mm is stopped by flat 40mm of armor.
The implementation of volumetric is inherently flawed.
On one hand we have RNG bounces in scenarios where the shell would 100% of time would perforate the armor, on the other hand Gaijin doesn’t understand that shells always follow the path of least resistance and shells should always ricochet against sloped armor when they can’t penetrate it.
There are of course exceptions in reality. Like very soft armor might catch the shell instead of causing it to ricochet, simply because the armor can deform so much.
See the post above.
There’s not enough armor on the side to magically stop a shell that could penetrate two Jagdpanthers side by side.
That energy has to go somewhere and in WT it just vanishes.
1 Like
I answered this point two posts below the one you answered :)
Not really. Right now the code seems to take the highest value of armor line-of-vector thickness within the area under the projectile radius of the point of impact to calculate penetration.
When it should be finding the least value of it (that is not zero for holes in the mesh) and using that.
The way it is modelled in the game really is arealistic. Even discounting the above “edge case” (literal) issue. There is no reasoning that a 4mm plate should resist penetration just because it was edge on and gave a notional 1000mm of RHA.
Its just bad modelling.
Note the 500m M61 penetration struck right on where the horizontal plate is.
6 Likes
The shell getting absorbed is just a simplified simulation of it getting stuck.
All simulators do that cause the math involved requires too many CPU cycles.
The implementation of volumetric is among the best in the industry, and there are no bugs as of present.
Also a 90mm round wasn’t stopped by 40mm of armor, it was stopped by 315mm of armor.
@_Baum
My screenshots no bug.
Roof armor stopping shells is a well known real-life situation.
As I said, you just want an arcade game, but instead your posts contain made up things about a system.
No it’s simple a math problem. The game see’s 300mm of armor protectiona and says: You can’t pen that with 220mm of penetration.
But where that 300mm of armor protection comes from, makes no sense.
The issue here is the layout of the armor:
The shell passes through the 50mm side armor but then immediatly hits the 16mm sponson floor armor which is modeled as volumetric, so it’s actually 16mm in height.
Shells aren’t actually 3D but they are only 3D to check what they impacted.
The shell is still pixel sized when it penetrated the 50mm side armor but now the shell is stuck inside that 16mm heigh armor plate, except that the armor is volumetric and it has a width of like 40cm.
If the armor is not volumetric and like the old model of the Befehls Tiger P, it’s actually very difficult to have the round non pen.
It will even behave to the expected degree and can cause a ricochet from one plate into the other.
The problem is that the game isn’t very good in dealing with volumetric armor.
Unrealistic behaviour was always the result when the game had to deal with 3D armor instead of simple planes.
When tracks were modeled with armor on every side, the game would have armor protection sky rocket when a shell hit a track at an angle that made the game think the armor dimension was much higher than physically possible.
So you would always have tracks eat your shells when hitting enemies from the side, or sometimes when you hit a piece of track add-on armor.
It’s becaus the game doesn’t simulate penetration of 3D objects but uses math that only works for a simple 2D approach.
Everytime the armor becomes too complex, the result can be wildly different from realitiy.
2 Likes