THE CROSSHAIR & CONVERGENCE SECTION IS CURRENTLY INCORRECT FOLLOW UP VIDEO IS UP WHERE I CORRECT THE MISTAKE: ruclips.net/video/wYiQxYbBKHs/видео.html Also, for the part about the .50cals catching up to the hispanos, this is wrong and misleading, the .50cal belts I was using on the British Spit LF Mk9 has similar but not better velocity than the hispanos when I was testing (there are like 5 different .50cal variants and I forgot to include a few)
@@flynntaggart8549 I said it in the video but when convergence is set to "no", the projected distance of the crosshair and mouse cursor is 1000m or 1km. You can test this yourself in replay mode, where your mouse cursor shows up at the distance you've set in free camera mode
If you make up a follow-up video you should mention the ballistic coefficient. Noting that m is proportional to d^3 for a given geometry, the ballistic coefficient is proportional to d, and hence large-caliber shells maintain their velocity better at range. For example, on the graph at 6:22 the low-velocity 37mm catches up to the faster shells. This is very apparent in ground and especially naval battles.
While this trend is generally present, i dont think the trend is strong enough to really necessitate pointing out. like a lot of 7mm MGS go very slow and a lot of 20mms go far but the .50cal browning goes very far for its caliber. I could list the ballistic coefficients but i think its more visually easier to understand if I just show the graphs like i did in the video where you can see how far it travels in a given amount of time
@@CatWerfer It gives you an idea how much penetration a round will maintain at range, which is hard to tell from the graphs. It's more relevant for tanks/ships; a 122mm shell fired at with 800m/s will be effective much farther than a 30mm shell fired at 800m/s. None of this is intuitive for new players.
4:48 If you want a really good, immediately visible representation of this, test fly the XA-38 Grizzly. Take some test shots on the runway with the 75mm HE belt (600m/s), then equip the mixed belt which starts with AP (800 m/s), and look at how much closer to your cursor the HE shells come in the mixed belt.
I saw this argument but I'm not really one to take things to be true just by looking at them. The rounds have different masses and drag parameters and I wanted to be 100% sure without resorting to a qualitative assessment
I was pretty bad at it but it was a lot of fun back then, I think I stopped playing when they changed the way the techtree worked (God that was ages ago)
Smol little funfact i noticed today: In the sunrise time setting in Air RB, the sun is not as expected in the east, but in the west (according to map and to cockpit compass)
@@CatWerfer ayo thanks for answering :)) And also i really like content that goes into mathematical and physical explanations, so keep up the good work :)
Same in SIM EC, noticed it in a match on Zhengzou recently, where I saw the sun peeking over the horizon my left while heading north and thought: "oh boy another dusk/night mission, which kills energy fighting and CAS...." I was flying british against the Japanese, which means having to switch to energy and BNZ tactics.
Dude, i'm absolutely loving this new wave of technical videos! You're probably already pretty annoyed that i keep saying this but i'm loving this vibe of AdamTheEnginerd's spiritual successor, with the little paint demonstrations and all the theory behind with the equations, along with the effort put into figuring out ways to test and demonstrate your theories. I really can't wait for more. I had heard of the belt bug just a few weeks ago, so it really was good to see it being confirmed like that, one must wonder if Gaijin will ever solve a bug that's been in the game since it's dawn. Unlikely, but with Gaijoob you never know. I've seen in this comment section that you're studying mechanical engineering, i've also been starting out in mechanical engineering this year in college, and seeing content like this makes me all the more excited that i'm already understanding some of the equations and logic, lol
@magicmissile7533 hell yeah brother. Studying mechanical engineering in college has been kinda interesting in that while the classes have mainly been theory, they expect you to know a lot of hands on stuff by the time you graduate that they don't really put in effort to teach you. They expect you to have learnt those things thru extracurriculars like robotics or some sort of engineering design competition clubs, and one regret I have with my time in college I'd that I didn't participate in those as much as I feel like I should've. I also feel like I should've taken more computer science courses too, those would've helped... Anyways, glad you're liking the technical stuff! I know most people don't really care since it's just a fun pew pew game but it's good to know that there's still a somewhat small group of people who are interested in what happens under the hood like I am so it gives me motivation to know I'm sharing something interesting not just to me but to you as well!
@@CatWerferIt's funny, that sounds very similar to my experience with mechanical engineering. Design teams just didn't work for me, I couldn't focus on school and keep up with the extra work. In the end I preferred to just build my own projects, but of course you do get more support and equipment working with a team. I collected a lot of gear over time to be able to do some work at home.
A lot of guns now have more realistic drag coefficient but .50cals are still one of the most ballistics efficient rounds :) Another bug to talk about is self-destroying shells, which don’t explode based on a timer but a set distance. So MG 151/20 and MK 108s can explode much earlier then they should during high speed tail chases.
It always annoyed me that there isn't a "crosshair distance" setting. I like to change the gun convergence depending on the aircraft but the differing crosshair position messes with my aim. The belt bug is an embarrassing bit of spaghetti code and I'm impressed they haven't fixed it all these years. It makes the MG151/20 "air targets" belt significantly worse than it should be
I'm glad that you cleared up if the gun bud is unique to aircraft or also effect ground vehicles, I've been trying to figure that out but no one seemed to have concrete answer. Specifically I've wondered about how the Falcons belts act, because as someone who has played a lot of it, I really thought the APDS rounds had less drop than the rest of the shells, but I must have been mistaked and just thought so on "vibes".
Also the velocity bug only affects initial velocity. If the APDS has less drag and more mass then it will still travel farther than the rest of the rounds
Great video! I like that tou calculuted time to travel based on drag, useful! When you did Spitfire 50cal vs hispano test, keep in mind Spit M2 browning have much lower muzzle velocity (793m/s or 832m/s depening on belt ) than US 50 M2 browning (899m/s on tracer). I hope it's not braggy but i made a similar vid that with the same conclusions about first round determining all velocity. Btw. Right now NR-30 has by far the lowest drag coefficient in the game.
Awesome video, you should cover how projectile penetration is calculated next! I’m assuming gaijin fudges some values for balance but I’m curious how they programmed it under the hood
that would be a lot harder to test without custom mission setups and I'm no good with those... There are people on the datamine discord that are much more into the way they're calculated, but penetration power does have to do with impact velocity. Idk, i'd like to try and find out but without a method to do so I'm kinda stuck on square one. Glad you like the video!
The belt bug doesn't seem to work for everything. It's especially apparent when using the Bradley or BMP as they have 1 HE shell with their APDS and it is noticeably slower and as a result drops to the ground at closer ranges.
nice video, appreciate the detail on it do you know if gaijin use the horizontal differential equations(which have the analytical solution you showed) for all drag calculations, even if you nose is at, say, a 45 degree angle(or any nonzero angle above flat), or does it use the non-analytical set for quadratic drag in 2 directions? ive written a calculator for the latter and how precise you want the code to be can significantly increase compute times so i can see why they'd just go with the former
I'm not entirely sure if they do analytical calculations at all, i mean for all we know it could be a per-frame calculation with a certain amount of directional deceleration applied at each frame to oppose velocity. I don't know how computationally expensive that would be though, I don't know much computerscience so tbh you probably have a much better idea on how they calculate it in game lol.
My hypothesis with the first round determining the speed for all rounds is that gaijin don want to deal with (write code for) rounds colliding in the air.
I think in itself it isn't that crazy (its a bit worse than the .50cals) but the fact that it's a high explosive fragmentation round with that amount of velocity means that it's incredibly punchy and can potentially oneshot a wing, so in that regard it is kinda absurd lol. Not to mention that it gets paired with one of the best 20mms in game on the a21s
@@CatWerfer It's absolutely comical if you compare it to the Italian/Japanese 12.7mm HEF round. Swedish props could be at BRs reasonable for their flight performance if their weapons weren't so much better than their competition.
yeah the japanese HEF 12.7 used to be really good but i think it got nerfed during realshatter. Not sure if the 13.2 got nerfed with it... I agree that the J22 has such a terrible flightmodel for 3.7 and the only reason its so high is those damn guns
I just recently found this channel you do good work but I never realized how much of my college math classes I forgot lol do you have any tests or info on damage between Ap and He shells? I've been just goofing off with ground target belts on m2s and the 190 with aphe and it seems like sometimes they do way more damage but I haven't done any serious testing
mmmm unfortunately no I don't really have that much info on how shell damage works. I feel like after realshatter, a lot of stuff became pretty inconsistent in damage output but overall I feel like for most belts, the high explosive and fragmentation (and sometimes APHE though they usually have a fuse and need to hit something thick) bullets usually do the most damage. If they don't have that, then armor piercing incindiary (not armor piercing or incindiary by themselves) are good for damage or at least setting fires. These are more vibes though, I can't be certain. You can always test using the protection analysis in the hangar but I also feel like that's inconsistent too. Maybe if I learn how fragmentation and damage is actually calculated i could do a more in depth explanation on how damage to components is calculated ingame... I know that each component in an aircrafts damage model has its own HP value though but not much beyond that. If i were to try and investigate this tho, I'd probably go talk to people in the war thunder datamine server. They seem to know a lot more about how the game engine works so if you want a super technical explanation, that's probably where I'd go to see if you can piece together an explanation.
@@CatWerfer thank you for the reply, I started because I noticed the air target belt on I believe ki61 seemed to give weird results but of course 12.7 he is not very powerful
I remember that they used to be really good before the realshatter update, probably thanks to their HEF-I rounds, though after realshatter the fragmentation probably got nerfed (like it did on a lot of planes)
.50s are going to feel a lot weaker than any cannon armament at range despite technically better long-range performance because .50 really relies on kinetic energy and multiple hits on a single component to deal real damage, at range the velocity and dispersion makes hits weaker and spreads them out across the fuselage, anything 20mm or bigger with explosive filler is going to do a lot better with just a handful of hits spread out across an aircraft.
Did you consider whether bullets lose their damage output, assuming we're talking about an inert project, aka one that doesn't explode such as an armour piercing round. They might have tied math to the velocity of the bullet to the damage to the target
I think there are damage loss multipliers based on distance in arcade, but judging by the fact that you can oneshot someone from long range, I have my doubts that this applies in realistic. Thought i can't be sure and it would be difficult to test
Thank you so much for testing the belt velocity thing! I heard about it a few days ago for the first time and am still baffled. Also, I have a question, maybe you know the answer: Are the Belt Rounds reset after earch burst or not? For example, when shooting with the HE->HE->AP Belt on the P-63, and shooting 2 projectiles with a single press, then waiting, and shooting another shot, will it be AP or reset to HE?
@@CatWerfer Wait, my question wasnt actually related to the velocity thing, but instead to wether a HE or AP round will actually leave the barrel and do the damage.
oh sorry. The game remembers what shots have been fired so the belts do not reset to start over again. So in your example, the 3rd shot fired after the first two will be an AP
Ok i have one question, does firing your gun at sea level will have the same velocity curve as firing it lets say at 8k meters altitude, because of change in density of air, does WT account for this?
I haven't tested it but i feel like if they bother using a coefficient of drag and they already do altitude drag considerations for planes and missiles i think its possible, though i dont think it will make a noticeable difference
@@CatWerfer interesting, sometimes the snail makes me wonder they make/simulate all these little things yet fail to fix some glaring issues in the game lol
yeah i think they had an excellent foundation at the start with regards to the physics engine for prop dogfights but they stopped caring about keeping it clean a while ago
I feel like every american on my tail has not gotten the memo, they all spray at over a kilometer at me. I guess if you got that much ammo you might as well :P
Well, regarding gunnery, there was this bug ehen WT started, that was only noticed in Sim or Full Real Battles as it was called then. Namely that all aircraft seemed to be double the size they should be at their ranges, and the gunsights were the wrong size, which messed up lead and convergence calculations. I first noticed it, when I got on the tail of a P-47D in my 109 E-3 (Yes MM was that borked back then) at 100-120m if you go via size and gunsight, fired with my guns set at 150m* convergence only to see the sehells cross paths behind him and then extend outwards completely missing him. I then asked some people on the forums, the old forums about it and they told me that they knew about it and had experienced the same. And that every gunsight was off by a different ammount. We knew this, because we've flown the same aircraft in IL-2 1946 and knew the gunsight dimensions in Miliradian. If you know that size and the target's Wingspan you can calculate the range. And un direct comparison to IL-2 1946 they were way off , sometimes 30-40%. Additionally we found out that the aircraft were too large, compared to hangars, airfueld runways and aircraft carriers, which lead to many a takeoff or landing accidents on a carrier. We found out the physical distances were right, but the appearant distances were off by a factor of 2. I wonder if that bug still exists. *As per my experience on IL-2 1946 where 150-200m convergence gave me the best hitrates and bullet/shell damage efficiency.
huh i'm not a big sim guy so I didn't really know about that. that's really weird tho. i wonder if that was also the case in realistic battles and it just didn't come up as a problem cuz of the spotting system
@@CatWerfer Yeah it did exist, that's how we confirmed it. As it became rather appearant then and we could figure out the actual ranges. Now the size seems to be broadly in shape with the other assets now. I still wonder however, as ingame you seem to need quite a bit more lead and holdover to hit your target, compared to other games/sims. So I wonder if everything is fixed and the correct size, or if the other assets just were made bigger, which would explain a lot of the weirdness in the ballistics, as the guns would need to fly further that it appears. Especially since the .50 cal had a possible max range of close to 2,000 meters, ballistically, but was usually limited by the gunsight technology, which is why the K-14 and later the A-4 Gunsight on the later Sabres were such a big deal.
@@LupusAries you see enemy farther when you fill his wing in your e3 gunsight circle then, that just means the circle is smaller calibrate to a longer distance, i don't think actual model dimension and hud distance is not to scale
1) select a plane in the hangar. if you own the plane, you can either view it in your lineup or hit "customization" from the techtree. If you don't own it, you can hit "preview" from the techtree. 2) select "armor" to turn on armor viewing mode. 3) under "armor", select protection analysis mode. this will bring up the ability to test fire rounds at the selected plane and see how much damage it takes 4) under "shell" it tells you the currently selected ammo type. you open the dropdown menu and look through all the different ammo types. 5) if you hover over each option, it will tell you different stats about the round, like the muzzle velocity, mass, penetration, and type of damage (but not coefficient of drag). If a belt only has one type of ammo in it, you can also just view these stats in the modifications tab dropdown where you select what kind of ammo you want to use, though most belts have mixed ammo types so you can't do that for all belts. You can also go to the datamine page, find the cannon you're looking for, then find the stats of the round you're looking for in the file for that cannon, but that can be a pain in the butt...
Huh, that music sounds familiar. I wonder wher- wait a second. IS THAT THE ROBOCRAFT SOUND TRACK RAAAAAAAAAAAAAAAAAAAHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
Is it confirmed that shell velocity affects penetration of ground vehicles? I.E. diving faster while shooting means better penetration. If there was a way to calculate this, it would be interesting to compare with the statcard penetration numbers
I haven't tested it (idk how I would) but it probably would only make a difference if you're going really fast, where the speed of your aircraft would have to be a significant fraction of your muzzle velocity
@@CatWerfer You could maybe calculate how long it takes for rounds to decrease back to normal muzzle velocity (plot speed, instead of distance, against time) that would be an interesting idea on its own
@@puff7145 I did do those speed vs time plots on my own befor emaking this video to see if they were easier to understand but they didnt look as neat or practical as the distance vs time one. i dont really wanna make new videos every time i decide to plot something new on an old topic so maybe i'll make some kind of platform where i can share that kind of data
as with all of these things, I just present a bunch of data but there isn't really a "best" convergence and you should use what's most comfortable to use. I just end up using 600m with no vertical targeting on everything I fly because I'm just used to it. (i would avoid 50m convergence tho lol)
Can you plz test the tracer arc in different roll angle at 50m vertical convergence setting (ep J5N1)? It seems like the arc doesnt roll with the aircraft, It always went above the crosshair after passing it even if the planes up side down
Are you asking why it seems like for 50m vertical convergence it seems like turning the plane upside down does not change the fact that the bullets intersect the crosshair at 50m, when if you had a farther vertical targeting (say 800m), turning the plane upside down would make you miss? If that's your question, the reason why you end up missing when upside down with 800m vertical targeting is that vertical targeting is calibrated so that the bullet crosses the crosshair at 800m when the plane is horizontal. If the bullet didn't experience gravity, this would be just a straight line, but because it does, the guns are angled more upwards to account for the amount of drop the bullet will experience over that distance. When the plane turns upside down, this additional upward angling point the guns downwards relative to the earth and so instead of counteracting bullet drop, you actively make it worse. Now at 50m, because the range is so close, with bullets that travel that distance very quickly, the effective trajectory of the bullet from gun to crosshair is basically just a straight line. The guns won't be angled very much higher because across that short distance, there's not much time for gravity to affect it and so any bullet drop will be very negligable. Because of the fact that the gun-to-crosshair trajectory is basically a straight line with very little added upwards angling to the guns, flipping the plane upside down will not change the trajectory from 0m to 50m very much, it will still basically just be a straight line. If this wasn't what you were asking about, let me know.
@@CatWerfer man this is getting rly complicated here😭 For those guns mounted below the nose like J5N1, a 50m vc setting basically turns them into schrage musik and aim a bit towards the sky when the plane is level. Therefore when I flip the plane up side down ,the guns shoudve aim slightly towards the ground right? But in third person view the arc looks the same in any roll angle, this is where I'm confused.
@@CatWerfer also Im not sure if the "enble vertical targeting" option is working on my j5n1, it seems like the vertical convergence is always forced enbled when using a convergence setting
Ok so from the brief testing i did using the j5n1, f15a, f5e, mig21pfm, and av8c: It seems I was wrong in my video about the mouse cursor and gun crosshair being in line with the line of sight made by the pilot and the in cockpit gunsight. Instead, while it seems like it could be parallel to the pilot-cockpit gunsight line, it is definitely shifted below it, which you can find when you use 50m convergence in cockpit mode using mouse aim, where your cursor actually falls below the center of the in cockpit crosshair. Now I'm not 100% certain how far down the line for the mouse cursor and 3rd person gun crosshair are. My guesses are it is either a preset axis that is used when modelling the plane, or a line that comes starting from the Center of Mass or Center of Lift (aerodynamic center), though these are also probably precalculated since it would be weird for these to change if the flight model were to change mid flight (different Center of Mass due to lower fuel or different aerodynamic center due to using flaps or wingsweep). This is also likely the position at which the virtual cockpit is located (the view that you can switch to to see nose angle and the velocity vector direction). At the very least, its a line that starts somewhere inside of the plane and it is not a line that starts from the barrel of the gun because you can see that on planes with low mounted guns like the mig21 and av8c, the gun is still angled up. Working with this potential explanation, the reason why the j5 doesn't seem to be very affected by short range vt targeting would be because the guns are mounted not too far off of this fake line of sight that exists for the mouse cursor and third person gun crosshair. This means that the guns basically aren't angled up that much from the horizon because the vertical distance between the gun barrels and this fake line is very little. You can see that this also happens on other planes with relatively high mounted guns, especially the La7 and its much easier to see happening in virtual cockpit. Again I'm not 100% certain this works but it seems to be the case for most planes. Thanks for bringing this to my attention, this is really interesting and that's some really good attention to detail on your part. Now I wonder if there is a plane with a gun mounted ABOVE this fake cursor line, which would mean that vertical targeting would actually aim the gun DOWN... I'll do a bit more testing, but I think this means that I'll have to issue a major correction to the video...
@steve_i400 mmmm I don't think so. Usually in the datamine files, gunner based guns are more prone to dispersion. You maybe died because remember that since gunners fire their guns backwards, their bullets aren't flying against the wind and have a lot less drag than if they were forward, meaning they fly a lot more in terms of relative distance
i love firing my gun backwards while running and seeing it fly away from me, but the guy standing still recording it sees the bullet sitting stationary in the air
I study mechanical engineering, which means i know just enough about physics to explain the kind of stuff you see in this video. Stuff about quantum mechanics or relativity or space stuff or nuclear stuff that the hardcore physicists study is way beyond my expertise.
THE CROSSHAIR & CONVERGENCE SECTION IS CURRENTLY INCORRECT
FOLLOW UP VIDEO IS UP WHERE I CORRECT THE MISTAKE:
ruclips.net/video/wYiQxYbBKHs/видео.html
Also, for the part about the .50cals catching up to the hispanos, this is wrong and misleading, the .50cal belts I was using on the British Spit LF Mk9 has similar but not better velocity than the hispanos when I was testing (there are like 5 different .50cal variants and I forgot to include a few)
i was gonna ask what happens to the crosshair distance if you don't set a convergence range
@@flynntaggart8549 I said it in the video but when convergence is set to "no", the projected distance of the crosshair and mouse cursor is 1000m or 1km. You can test this yourself in replay mode, where your mouse cursor shows up at the distance you've set in free camera mode
If you make up a follow-up video you should mention the ballistic coefficient. Noting that m is proportional to d^3 for a given geometry, the ballistic coefficient is proportional to d, and hence large-caliber shells maintain their velocity better at range. For example, on the graph at 6:22 the low-velocity 37mm catches up to the faster shells. This is very apparent in ground and especially naval battles.
While this trend is generally present, i dont think the trend is strong enough to really necessitate pointing out. like a lot of 7mm MGS go very slow and a lot of 20mms go far but the .50cal browning goes very far for its caliber. I could list the ballistic coefficients but i think its more visually easier to understand if I just show the graphs like i did in the video where you can see how far it travels in a given amount of time
@@CatWerfer It gives you an idea how much penetration a round will maintain at range, which is hard to tell from the graphs. It's more relevant for tanks/ships; a 122mm shell fired at with 800m/s will be effective much farther than a 30mm shell fired at 800m/s. None of this is intuitive for new players.
bro's playing math thunder
Bro you ain't seen nothin yet, wait til I get to philosophy thunder
4:48 If you want a really good, immediately visible representation of this, test fly the XA-38 Grizzly. Take some test shots on the runway with the 75mm HE belt (600m/s), then equip the mixed belt which starts with AP (800 m/s), and look at how much closer to your cursor the HE shells come in the mixed belt.
I saw this argument but I'm not really one to take things to be true just by looking at them. The rounds have different masses and drag parameters and I wanted to be 100% sure without resorting to a qualitative assessment
the robocraft build music caught me off guard. Love this video
the RoboCraft music brings back memories and deserves a follow
I was pretty bad at it but it was a lot of fun back then, I think I stopped playing when they changed the way the techtree worked (God that was ages ago)
@@CatWerfer I miss when I was tier 8. Was good times
@@CatWerfer psst, check out RC15 project if you feel nostalgic for the old robocraft.
Smol little funfact i noticed today:
In the sunrise time setting in Air RB, the sun is not as expected in the east, but in the west (according to map and to cockpit compass)
Yeah recently ive noticed a lot of issues with time of day settings, I've even had them change from dawn to dusk mid-game it was very jarring
@@CatWerfer ayo thanks for answering :))
And also i really like content that goes into mathematical and physical explanations, so keep up the good work :)
Same in SIM EC, noticed it in a match on Zhengzou recently, where I saw the sun peeking over the horizon my left while heading north and thought: "oh boy another dusk/night mission, which kills energy fighting and CAS...."
I was flying british against the Japanese, which means having to switch to energy and BNZ tactics.
somebody'll probably do a bugreport on it but tbh i don't mind too much right now since it doesn't really affect air rb gameplay :P
Dude, i'm absolutely loving this new wave of technical videos! You're probably already pretty annoyed that i keep saying this but i'm loving this vibe of AdamTheEnginerd's spiritual successor, with the little paint demonstrations and all the theory behind with the equations, along with the effort put into figuring out ways to test and demonstrate your theories. I really can't wait for more.
I had heard of the belt bug just a few weeks ago, so it really was good to see it being confirmed like that, one must wonder if Gaijin will ever solve a bug that's been in the game since it's dawn. Unlikely, but with Gaijoob you never know.
I've seen in this comment section that you're studying mechanical engineering, i've also been starting out in mechanical engineering this year in college, and seeing content like this makes me all the more excited that i'm already understanding some of the equations and logic, lol
@magicmissile7533 hell yeah brother. Studying mechanical engineering in college has been kinda interesting in that while the classes have mainly been theory, they expect you to know a lot of hands on stuff by the time you graduate that they don't really put in effort to teach you. They expect you to have learnt those things thru extracurriculars like robotics or some sort of engineering design competition clubs, and one regret I have with my time in college I'd that I didn't participate in those as much as I feel like I should've. I also feel like I should've taken more computer science courses too, those would've helped...
Anyways, glad you're liking the technical stuff! I know most people don't really care since it's just a fun pew pew game but it's good to know that there's still a somewhat small group of people who are interested in what happens under the hood like I am so it gives me motivation to know I'm sharing something interesting not just to me but to you as well!
@@CatWerferIt's funny, that sounds very similar to my experience with mechanical engineering.
Design teams just didn't work for me, I couldn't focus on school and keep up with the extra work. In the end I preferred to just build my own projects, but of course you do get more support and equipment working with a team. I collected a lot of gear over time to be able to do some work at home.
A lot of guns now have more realistic drag coefficient but .50cals are still one of the most ballistics efficient rounds :)
Another bug to talk about is self-destroying shells, which don’t explode based on a timer but a set distance. So MG 151/20 and MK 108s can explode much earlier then they should during high speed tail chases.
It always annoyed me that there isn't a "crosshair distance" setting. I like to change the gun convergence depending on the aircraft but the differing crosshair position messes with my aim.
The belt bug is an embarrassing bit of spaghetti code and I'm impressed they haven't fixed it all these years. It makes the MG151/20 "air targets" belt significantly worse than it should be
yeah... gaijin is too busy adding new features like realshatter that noone asked for ig
Robocraft music is absolutly wild
"Smg cruiser spotted" 🗣🗣🔥🔥
thanks for the jumpscare at the end
Miau
I'm glad that you cleared up if the gun bud is unique to aircraft or also effect ground vehicles, I've been trying to figure that out but no one seemed to have concrete answer. Specifically I've wondered about how the Falcons belts act, because as someone who has played a lot of it, I really thought the APDS rounds had less drop than the rest of the shells, but I must have been mistaked and just thought so on "vibes".
i haven't personally tested it with ground vehicles but it does seem to be the case according to other ppl who have tested it
Also the velocity bug only affects initial velocity. If the APDS has less drag and more mass then it will still travel farther than the rest of the rounds
Great video! I like that tou calculuted time to travel based on drag, useful!
When you did Spitfire 50cal vs hispano test, keep in mind Spit M2 browning have much lower muzzle velocity (793m/s or 832m/s depening on belt ) than US 50 M2 browning (899m/s on tracer).
I hope it's not braggy but i made a similar vid that with the same conclusions about first round determining all velocity.
Btw. Right now NR-30 has by far the lowest drag coefficient in the game.
I didn't know you had a youtube channel, ill check it out in a bit
I'm going thru the drag coefs right now for other planes and the a10's gau8 has a lower drag coef than the nr30
O! Good to know
OG Robocraft music 💯
"SMG cruiser spotted" 🗣🗣🔥🔥
Awesome video, you should cover how projectile penetration is calculated next! I’m assuming gaijin fudges some values for balance but I’m curious how they programmed it under the hood
that would be a lot harder to test without custom mission setups and I'm no good with those... There are people on the datamine discord that are much more into the way they're calculated, but penetration power does have to do with impact velocity.
Idk, i'd like to try and find out but without a method to do so I'm kinda stuck on square one.
Glad you like the video!
The belt bug doesn't seem to work for everything. It's especially apparent when using the Bradley or BMP as they have 1 HE shell with their APDS and it is noticeably slower and as a result drops to the ground at closer ranges.
Huh that is quite interesting, I should check it out
nice video, appreciate the detail on it
do you know if gaijin use the horizontal differential equations(which have the analytical solution you showed) for all drag calculations, even if you nose is at, say, a 45 degree angle(or any nonzero angle above flat), or does it use the non-analytical set for quadratic drag in 2 directions?
ive written a calculator for the latter and how precise you want the code to be can significantly increase compute times so i can see why they'd just go with the former
I'm not entirely sure if they do analytical calculations at all, i mean for all we know it could be a per-frame calculation with a certain amount of directional deceleration applied at each frame to oppose velocity. I don't know how computationally expensive that would be though, I don't know much computerscience so tbh you probably have a much better idea on how they calculate it in game lol.
Holy shit, robocraft garage music
@Hippiskus I wanted to use the mars theme for the music but was worried it would go too hard lol
absolutely insane and underrated video
My hypothesis with the first round determining the speed for all rounds is that gaijin don want to deal with (write code for) rounds colliding in the air.
Huh.... maybe...
This video better mention the hilariously absurd ballistics of the Swedish 13.2mm.
I think in itself it isn't that crazy (its a bit worse than the .50cals) but the fact that it's a high explosive fragmentation round with that amount of velocity means that it's incredibly punchy and can potentially oneshot a wing, so in that regard it is kinda absurd lol.
Not to mention that it gets paired with one of the best 20mms in game on the a21s
@@CatWerfer It's absolutely comical if you compare it to the Italian/Japanese 12.7mm HEF round. Swedish props could be at BRs reasonable for their flight performance if their weapons weren't so much better than their competition.
yeah the japanese HEF 12.7 used to be really good but i think it got nerfed during realshatter. Not sure if the 13.2 got nerfed with it...
I agree that the J22 has such a terrible flightmodel for 3.7 and the only reason its so high is those damn guns
I don’t wanna do math
you can skip to gun comparison section if you dont wanna math
this video popped up on my feed and i forgot i am subbed 😭
Forgor
hearing robocraft on a wt video is wild
Lol I'm tryna draw on as many different media sources for my music as possible., at least ones I enjoyed
I just recently found this channel you do good work but I never realized how much of my college math classes I forgot lol do you have any tests or info on damage between Ap and He shells? I've been just goofing off with ground target belts on m2s and the 190 with aphe and it seems like sometimes they do way more damage but I haven't done any serious testing
mmmm unfortunately no I don't really have that much info on how shell damage works. I feel like after realshatter, a lot of stuff became pretty inconsistent in damage output but overall I feel like for most belts, the high explosive and fragmentation (and sometimes APHE though they usually have a fuse and need to hit something thick) bullets usually do the most damage. If they don't have that, then armor piercing incindiary (not armor piercing or incindiary by themselves) are good for damage or at least setting fires. These are more vibes though, I can't be certain. You can always test using the protection analysis in the hangar but I also feel like that's inconsistent too.
Maybe if I learn how fragmentation and damage is actually calculated i could do a more in depth explanation on how damage to components is calculated ingame... I know that each component in an aircrafts damage model has its own HP value though but not much beyond that.
If i were to try and investigate this tho, I'd probably go talk to people in the war thunder datamine server. They seem to know a lot more about how the game engine works so if you want a super technical explanation, that's probably where I'd go to see if you can piece together an explanation.
@@CatWerfer thank you for the reply, I started because I noticed the air target belt on I believe ki61 seemed to give weird results but of course 12.7 he is not very powerful
I remember that they used to be really good before the realshatter update, probably thanks to their HEF-I rounds, though after realshatter the fragmentation probably got nerfed (like it did on a lot of planes)
.50s are going to feel a lot weaker than any cannon armament at range despite technically better long-range performance because .50 really relies on kinetic energy and multiple hits on a single component to deal real damage, at range the velocity and dispersion makes hits weaker and spreads them out across the fuselage, anything 20mm or bigger with explosive filler is going to do a lot better with just a handful of hits spread out across an aircraft.
@brosefmalkovitch3121 yeah I figured it had more to do with damage output (and skill issue)
Did you consider whether bullets lose their damage output, assuming we're talking about an inert project, aka one that doesn't explode such as an armour piercing round. They might have tied math to the velocity of the bullet to the damage to the target
I think there are damage loss multipliers based on distance in arcade, but judging by the fact that you can oneshot someone from long range, I have my doubts that this applies in realistic.
Thought i can't be sure and it would be difficult to test
Thank you so much for testing the belt velocity thing! I heard about it a few days ago for the first time and am still baffled.
Also, I have a question, maybe you know the answer:
Are the Belt Rounds reset after earch burst or not? For example, when shooting with the HE->HE->AP Belt on the P-63, and shooting 2 projectiles with a single press, then waiting, and shooting another shot, will it be AP or reset to HE?
It's not dependent on how you fire your ammo bursts, it only depends on the first listed round when you look at the belt in the modifications tab.
@@CatWerfer Wait, my question wasnt actually related to the velocity thing, but instead to wether a HE or AP round will actually leave the barrel and do the damage.
oh sorry.
The game remembers what shots have been fired so the belts do not reset to start over again.
So in your example, the 3rd shot fired after the first two will be an AP
@@CatWerfer No problem, thank you. Alright, at least they made this aspect realistic.
Ok i have one question, does firing your gun at sea level will have the same velocity curve as firing it lets say at 8k meters altitude, because of change in density of air, does WT account for this?
I haven't tested it but i feel like if they bother using a coefficient of drag and they already do altitude drag considerations for planes and missiles i think its possible, though i dont think it will make a noticeable difference
@@CatWerfer interesting, sometimes the snail makes me wonder they make/simulate all these little things yet fail to fix some glaring issues in the game lol
yeah i think they had an excellent foundation at the start with regards to the physics engine for prop dogfights but they stopped caring about keeping it clean a while ago
Remember me, as the one, who the snail does not like
tbh the snail only likes whales
I feel like every american on my tail has not gotten the memo, they all spray at over a kilometer at me. I guess if you got that much ammo you might as well :P
@@hungryhedgehog4201 probably doesn't hurt to try
Well, regarding gunnery, there was this bug ehen WT started, that was only noticed in Sim or Full Real Battles as it was called then.
Namely that all aircraft seemed to be double the size they should be at their ranges, and the gunsights were the wrong size, which messed up lead and convergence calculations.
I first noticed it, when I got on the tail of a P-47D in my 109 E-3 (Yes MM was that borked back then) at 100-120m if you go via size and gunsight, fired with my guns set at 150m* convergence only to see the sehells cross paths behind him and then extend outwards completely missing him.
I then asked some people on the forums, the old forums about it and they told me that they knew about it and had experienced the same.
And that every gunsight was off by a different ammount.
We knew this, because we've flown the same aircraft in IL-2 1946 and knew the gunsight dimensions in Miliradian.
If you know that size and the target's Wingspan you can calculate the range. And un direct comparison to IL-2 1946 they were way off , sometimes 30-40%.
Additionally we found out that the aircraft were too large, compared to hangars, airfueld runways and aircraft carriers, which lead to many a takeoff or landing accidents on a carrier.
We found out the physical distances were right, but the appearant distances were off by a factor of 2.
I wonder if that bug still exists.
*As per my experience on IL-2 1946 where 150-200m convergence gave me the best hitrates and bullet/shell damage efficiency.
huh i'm not a big sim guy so I didn't really know about that.
that's really weird tho. i wonder if that was also the case in realistic battles and it just didn't come up as a problem cuz of the spotting system
@@CatWerfer Yeah it did exist, that's how we confirmed it. As it became rather appearant then and we could figure out the actual ranges.
Now the size seems to be broadly in shape with the other assets now. I still wonder however, as ingame you seem to need quite a bit more lead and holdover to hit your target, compared to other games/sims.
So I wonder if everything is fixed and the correct size, or if the other assets just were made bigger, which would explain a lot of the weirdness in the ballistics, as the guns would need to fly further that it appears.
Especially since the .50 cal had a possible max range of close to 2,000 meters, ballistically, but was usually limited by the gunsight technology, which is why the K-14 and later the A-4 Gunsight on the later Sabres were such a big deal.
@@LupusAries you see enemy farther when you fill his wing in your e3 gunsight circle then, that just means the circle is smaller calibrate to a longer distance, i don't think actual model dimension and hud distance is not to scale
Is that the robocraft soundtrack?
yeah lol
how do you find listed bullet velocity?
1) select a plane in the hangar. if you own the plane, you can either view it in your lineup or hit "customization" from the techtree. If you don't own it, you can hit "preview" from the techtree.
2) select "armor" to turn on armor viewing mode.
3) under "armor", select protection analysis mode. this will bring up the ability to test fire rounds at the selected plane and see how much damage it takes
4) under "shell" it tells you the currently selected ammo type. you open the dropdown menu and look through all the different ammo types.
5) if you hover over each option, it will tell you different stats about the round, like the muzzle velocity, mass, penetration, and type of damage (but not coefficient of drag).
If a belt only has one type of ammo in it, you can also just view these stats in the modifications tab dropdown where you select what kind of ammo you want to use, though most belts have mixed ammo types so you can't do that for all belts.
You can also go to the datamine page, find the cannon you're looking for, then find the stats of the round you're looking for in the file for that cannon, but that can be a pain in the butt...
@@CatWerfer thank you
Huh, that music sounds familiar. I wonder wher- wait a second. IS THAT THE ROBOCRAFT SOUND TRACK RAAAAAAAAAAAAAAAAAAAHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH
PLASMA COPTER SPOTTED RAHHHHHHH
Is it confirmed that shell velocity affects penetration of ground vehicles? I.E. diving faster while shooting means better penetration. If there was a way to calculate this, it would be interesting to compare with the statcard penetration numbers
I haven't tested it (idk how I would) but it probably would only make a difference if you're going really fast, where the speed of your aircraft would have to be a significant fraction of your muzzle velocity
@@CatWerfer You could maybe calculate how long it takes for rounds to decrease back to normal muzzle velocity (plot speed, instead of distance, against time) that would be an interesting idea on its own
@@puff7145 I did do those speed vs time plots on my own befor emaking this video to see if they were easier to understand but they didnt look as neat or practical as the distance vs time one.
i dont really wanna make new videos every time i decide to plot something new on an old topic so maybe i'll make some kind of platform where i can share that kind of data
Now make a video figuring out the best possible convergence lmao
as with all of these things, I just present a bunch of data but there isn't really a "best" convergence and you should use what's most comfortable to use. I just end up using 600m with no vertical targeting on everything I fly because I'm just used to it. (i would avoid 50m convergence tho lol)
Can you plz test the tracer arc in different roll angle at 50m vertical convergence setting (ep J5N1)?
It seems like the arc doesnt roll with the aircraft, It always went above the crosshair after passing it even if the planes up side down
Are you asking why it seems like for 50m vertical convergence it seems like turning the plane upside down does not change the fact that the bullets intersect the crosshair at 50m, when if you had a farther vertical targeting (say 800m), turning the plane upside down would make you miss?
If that's your question, the reason why you end up missing when upside down with 800m vertical targeting is that vertical targeting is calibrated so that the bullet crosses the crosshair at 800m when the plane is horizontal. If the bullet didn't experience gravity, this would be just a straight line, but because it does, the guns are angled more upwards to account for the amount of drop the bullet will experience over that distance. When the plane turns upside down, this additional upward angling point the guns downwards relative to the earth and so instead of counteracting bullet drop, you actively make it worse.
Now at 50m, because the range is so close, with bullets that travel that distance very quickly, the effective trajectory of the bullet from gun to crosshair is basically just a straight line. The guns won't be angled very much higher because across that short distance, there's not much time for gravity to affect it and so any bullet drop will be very negligable. Because of the fact that the gun-to-crosshair trajectory is basically a straight line with very little added upwards angling to the guns, flipping the plane upside down will not change the trajectory from 0m to 50m very much, it will still basically just be a straight line.
If this wasn't what you were asking about, let me know.
@@CatWerfer man this is getting rly complicated here😭
For those guns mounted below the nose like J5N1, a 50m vc setting basically turns them into schrage musik and aim a bit towards the sky when the plane is level.
Therefore when I flip the plane up side down ,the guns shoudve aim slightly towards the ground right? But in third person view the arc looks the same in any roll angle, this is where I'm confused.
@@CatWerfer also Im not sure if the "enble vertical targeting" option is working on my j5n1, it seems like the vertical convergence is always forced enbled when using a convergence setting
oh huh now that i test with this plane, yeah that is pretty weird, ill look into it a bit further and get back to you on it
Ok so from the brief testing i did using the j5n1, f15a, f5e, mig21pfm, and av8c:
It seems I was wrong in my video about the mouse cursor and gun crosshair being in line with the line of sight made by the pilot and the in cockpit gunsight.
Instead, while it seems like it could be parallel to the pilot-cockpit gunsight line, it is definitely shifted below it, which you can find when you use 50m convergence in cockpit mode using mouse aim, where your cursor actually falls below the center of the in cockpit crosshair.
Now I'm not 100% certain how far down the line for the mouse cursor and 3rd person gun crosshair are. My guesses are it is either a preset axis that is used when modelling the plane, or a line that comes starting from the Center of Mass or Center of Lift (aerodynamic center), though these are also probably precalculated since it would be weird for these to change if the flight model were to change mid flight (different Center of Mass due to lower fuel or different aerodynamic center due to using flaps or wingsweep). This is also likely the position at which the virtual cockpit is located (the view that you can switch to to see nose angle and the velocity vector direction). At the very least, its a line that starts somewhere inside of the plane and it is not a line that starts from the barrel of the gun because you can see that on planes with low mounted guns like the mig21 and av8c, the gun is still angled up.
Working with this potential explanation, the reason why the j5 doesn't seem to be very affected by short range vt targeting would be because the guns are mounted not too far off of this fake line of sight that exists for the mouse cursor and third person gun crosshair. This means that the guns basically aren't angled up that much from the horizon because the vertical distance between the gun barrels and this fake line is very little. You can see that this also happens on other planes with relatively high mounted guns, especially the La7 and its much easier to see happening in virtual cockpit.
Again I'm not 100% certain this works but it seems to be the case for most planes. Thanks for bringing this to my attention, this is really interesting and that's some really good attention to detail on your part.
Now I wonder if there is a plane with a gun mounted ABOVE this fake cursor line, which would mean that vertical targeting would actually aim the gun DOWN...
I'll do a bit more testing, but I think this means that I'll have to issue a major correction to the video...
Is there any evidence to suggest that gunners use different ballistics? I feel I've been killed at impossible distances/speeds by bomber gunners.
@steve_i400 mmmm I don't think so. Usually in the datamine files, gunner based guns are more prone to dispersion.
You maybe died because remember that since gunners fire their guns backwards, their bullets aren't flying against the wind and have a lot less drag than if they were forward, meaning they fly a lot more in terms of relative distance
i love firing my gun backwards while running and seeing it fly away from me, but the guy standing still recording it sees the bullet sitting stationary in the air
@@Numerose1 relative motion do be funny, just don't travel close to the speed of light 💀
@@CatWerfer 💀
3 stealth 1 tracer???
3 ball 1 sack
Do you study physics?
I study mechanical engineering, which means i know just enough about physics to explain the kind of stuff you see in this video.
Stuff about quantum mechanics or relativity or space stuff or nuclear stuff that the hardcore physicists study is way beyond my expertise.
@@CatWerfer Fair enough. I'm interested in engineering but thinking of studying physics.
3 minutes ago yeaaaaaaaaaaaaaa good videooo
bro the video is longer than 3min lmao how do u finish it that quick
@@CatWerfer I'm the fast shooter. Good video tho