[CLOSED] M61 Dispersion

[Yo-Yo]

So, having official data for "80% circle diameter" of 0.008 it's possible to convert it to median. A bit of common sense told me (and NineLine was a witness, as I told him :) ) that median for this case is something average between 1 and 2 mils. After exact mathematical conversion I have got 1.50.

 

So, 2.2 mils gives 50% more dispersion and M197 has 50% less... I can say that nobody would notice 50% changes if it is done without notification or changing stable version number... I mean that after each new version a lot of people notice changes where there are no changes at all. :)

 

Why the dispersions were made as they are now? I have no idea, probably because that time we have no exact references and could only guess - was it 50%, 80% or 100% circle, diameter or radius, etc.

 

Anyway, it is not a bad estimation, I think.

[backspace340]

So, having official data for "80% circle diameter" of 0.008 it's possible to convert it to median. A bit of common sense told me (and NineLine was a witness, as I told him :) ) that median for this case is something average between 1 and 2 mils. After exact mathematical conversion I have got 1.50.

 

So, 2.2 mils gives 50% more dispersion and M197 has 50% less... I can say that nobody would notice 50% changes if it is done without notification or changing stable version number... I mean that after each new version a lot of people notice changes where there are no changes at all. :)

 

Why the dispersions were made as they are now? I have no idea, probably because that time we have no exact references and could only guess - was it 50%, 80% or 100% circle, diameter or radius, etc.

 

Anyway, it is not a bad estimation, I think.

 

So is it going to be changed?

[NineLine]

So is it going to be changed?

 

Yes, he is just prepping you for the possibility it won't be some huge astounding difference.

[WindyTX]

Yes, he is just prepping you for the possibility it won't be some huge astounding difference.

Thanks look forward to it.

 

Sent from my GM1915 using Tapatalk

[NineLine]

Good example. For comparison, a BMP-2 is 6.7 meters long.

 

 

Can't work out a precise value without the range, but given how gently the pilots pull up after the attack, it's further than I normally shoot (G override every time to avoid the ground).

 

So the pattern width is within a vehicle length, at a non-suicidal distance. This is not the case in DCS.

 

Edit: In-game test. I started shooting one second after the break X started flashing. Pulling away from the hill caused greyout and I just missed the trees. I think it's fair to say the Axalpa shooting distance is further, yet the in-game pattern is twice the size.

 

 

One more thought; even if, in theory, clamps can increase the dispersion, the sim should model them without. Equipment in the sim is usually portrayed in an ideal state, not a degraded state. Like how the Ka-50 gun has practically no dispersion despite it being a flexible mounted helicopter gun, optics always track perfectly, gun computers always have perfect information from the sensors, and muzzle velocities are always to manufacturer spec, with no account for wear.

 

Couple things about in-game tests, especially reference with pictures, there is no way of telling how smooth you are on the stick, especially when you pull the trigger. How much of a 'wiggle' would it take to through the dispersion into double what it should be? Not much I would think. Even with changes, Yo-Yo is looking at, there are factors out of our control, I would be interested to see if practice make perfect, such as from Lex, on how his first squeeze of the trigger was compared to where he ended up in real-world training.

 

Now this isn't to say what you are showing is what is actually going on, I just want to say that pictures don't always tell the whole story.

[WindyTX]

While not in an F18 I have lots of A2G strafe time in the F3 and I have very good hotas setup , that was why I made the original post. I know what my Pipper control was like IRL and in game. I just couldn't believe the F18s gun was that inaccurate compared to the F3.

 

Sent from my GM1915 using Tapatalk

[Preendog]

And in-game you can use time compression to get an unrealistically accurate burst. It's basically a rigid test stand.

[Eldur]

Couple things about in-game tests, especially reference with pictures, there is no way of telling how smooth you are on the stick, especially when you pull the trigger. How much of a 'wiggle' would it take to through the dispersion into double what it should be? Not much I would think. Even with changes, Yo-Yo is looking at, there are factors out of our control, I would be interested to see if practice make perfect, such as from Lex, on how his first squeeze of the trigger was compared to where he ended up in real-world training.

 

Now this isn't to say what you are showing is what is actually going on, I just want to say that pictures don't always tell the whole story.

 

 

Suggesting:

 

 

 

Fly with ATTH and BALT (omit attacking something though), then fire the gun via keyboard press and take one pic there. No wiggle on the stick. Clearly visible how much spread there is and the mils of the pipper should be well documented in some legendary operational how-tos that might exist somewhere around the globe

 

 

And of couse, take a track of that as well

 

 

We also could try to take a peek at the miniguns on the Huey. Different gun, different calibre, but same principle. It's tech specs can be found somewhere, I guess. Compare the tech specs IRL and the results in DCS and see how much they differ to each other in both cases. Add to that that the UH-1 in flight is a very wiggly platform, unstable to the bone. If that minigun still is more accurate, we actually might have an issue here

[NineLine]

Suggesting:

 

So I should just fire the gun off into the great blue skies, and that will tell me if dispersion is off or on? I don't think that's how that works ;)

 

Pretty sure to see how well it works is firing at a target at the appropriate range, I mean if I fire the gun off into space, and am very sure the rounds will end up dispersed all over the DCS World :)

[Sergeant_Hamlet]

So I should just fire the gun off into the great blue skies, and that will tell me if dispersion is off or on? I don't think that's how that works ;)

 

The funnel sight has clear and stable range indicators. Firing from a completely stable aircraft you could absolutely measure the rate of dispersion over those distances. Maybe not the best way of doing things, sure, but it's not totally pointless. It's an idea, at least.

[Eldur]

So I should just fire the gun off into the great blue skies, and that will tell me if dispersion is off or on? I don't think that's how that works ;)

 

Pretty sure to see how well it works is firing at a target at the appropriate range, I mean if I fire the gun off into space, and am very sure the rounds will end up dispersed all over the DCS World :)

 

 

Since dispersion in DCS is modelled as "the round leaves the barrel at a given angle up to x and never ever changes it's deviation after that except for gravitational ballistics" (in short: it's simplified), this actually would not make a difference. The dispersion could just be measured by the tracers, no matter if they hit something at a given range or not. The angle is the same. IRL this wouldn't work actually - so maybe this is the part where changes have to be made?

 

And it could be done with another plane in front at a fixed distance as well... at the moment you hit, take a snapshot and see where all the tracers are.

 

 

Anyhow - I could do those tests as well, but having experience in A/G gunnery with the Hornet tells me things are off. When firing as soon as the IN RANGE cue comes up (which I would understand as appropriate), I need 200 rounds at least to get a simple truck down. If I want to kill it with a short 20-30 rounds burst, I need to hold fire until Betty wants me to pull up and then quickly squeeze the trigger just before doing so. I wouldn't imagine the gun is designed to be used in the very last moment before crashing into the ground...

[NineLine]

When firing as soon as the IN RANGE cue comes up (which I would understand as appropriate), I need 200 rounds at least to get a simple truck down. If I want to kill it with a short 20-30 rounds burst, I need to hold fire until Betty wants me to pull up and then quickly squeeze the trigger just before doing so. I wouldn't imagine the gun is designed to be used in the very last moment before crashing into the ground...

 

Not been my experience... excuse the last pass... was aggro'd by my daughter :)

 

[Aries144]

I know you like posting that image, and I am looking into it now, but don't expect the same results from an aircraft in flight to one strapped down and locked in place.

Nineline,

 

Respectfully, the only way to accurately examine dispersion is exactly that seen in the video: the gun's dispersion absent all outside factors. The gun in the video was even mounted to an aircraft during all testing which accounts for vibration. I'm suggesting committing to a similar test within DCS to make it comparable which I will lay out below.

 

I would suggest that attempting to define the gun's performance by shooting the ground while moving is not a good way to examine this. There's far too much user variable. Any accurate modeling of the gun's fixed performance should play out in-sim just the way a real pilot, with real human error in-sim, should be familiar with.

 

The F-5 has a gunsight with a 50 mil ring. This gunsight ring is projected out to infinity, which makes it useful for measuring dispersion as they are both measurable using the same angular units (mils) regardless of projectile distance. In other words, it's as good as shooting a target of known size at a known distance and examining the bullet holes.

 

Try the following by editing the shell_table.lua:

 

 

• Input the desired value for whatever aircraft gun to the F-5's M39 gun.
  
• Make every round a tracer with a very long tracer burn time
  
• Create a mission with an F-5 at high altitude
  
• dive at or near 90 straight down, then activate Active Pause
  
• deactivate one gun breaker so only one gun fires
  
• fire the magazine until empty while recording video
  
• in the video, note the outside-most position of each tracer (this will occur as the tracers pass beyond the points of vertical zero and parallax error for the sight)
  
• Assemble a composite image of all the tracers. Timing is not critical so long as each is measured at the point where it has traveled as far from the center as it is going to.
  
• Using the 50 mil gunsight ring for Mil measurement comparison along with pixel measurements in Photoshop should allow a fairly precise measure of both Extreme Spread (the overall largest diameter of the shot group) and Standard Deviation (the average measurement that all shots dispersed from the center of the shot group).
  

 

This should allow an accurate comparison between the sim and our video evidence.

Edited June 30, 2019 by Aries144

[NineLine]

Nineline,

 

Respectfully, the only way to accurately examine dispersion is exactly that seen in the video: the gun's dispersion absent all outside factors. The gun in the video was even mounted to an aircraft during all testing which accounts for vibration. I'm suggesting committing to a similar test within DCS to make it comparable which I will lay out below..

 

There is no way to really test the gun in a similar way to how it's tested on the ground in those videos, the most important element you are removing is the pilot and his skills. Vibration is only a small part of it, but if you nudge the stick, even the tiniest bit, that translates to huge movements on the ground.

 

Anyways, Yo-Yo said there is some room for adjustment, not in the way stated here as you are not seeing the math behind it, but also the gun works, I have tested it if you do things the way you should, the gun works quite well.

[Tom Kazansky]

What about firing gun while active pause turned on?

 

 

Hmm... have to check if it's even possible when I'm at my rig....

 

 

EDIT: it works perfectly, imho. so it should be very easy to setup a mission where you can test different AC's guns from the same point on the same target, and see if the Hornet's is less accurate than other/same/similar guns.

 

 

EDIT2:

Set the ammo to unlimited and you might get nice circles with max dispersion. Is it possible to setup very small concentric target zones where you can count impacts depending on distance from the centre?

Edited June 30, 2019 by Tom Kazansky

[NineLine]

 

[Eldur]

BTW in order to conduct proper testing:

 

 

What does the winding down bar on the pipper translate to? The DCS manual doesn't inform us there and I've yet to stumble across legend books (you know, those N-word thingies) to tell me what range it is scaled to.

 

 

In short: I don't know the slant range. How far are we shooting at if we pull the trigger as soon as IN RANGE comes up?

 

 

 

If I'm not mistaken, in the F-14B the pipper goes to "in range mode" (diamond goes away) at 4000 ft which is roughly 0.66nm or 1.22km in proper units. Shooting a quick burst at that range can be pretty devastating there.

[Preendog]

"In range" shows up at 1nm last I checked ingame.

 

 

 

e. Composite of screenshots from a 2 second burst. The building just happened to be exactly 100ft wide.

 

 

It's hard to see with the smoke, but a 100% circle is at least 30 mils wide.

Edited June 30, 2019 by Preendog

[Aries144]

There is no way to really test the gun in a similar way to how it's tested on the ground in those videos, the most important element you are removing is the pilot and his skills. Vibration is only a small part of it, but if you nudge the stick, even the tiniest bit, that translates to huge movements on the ground.

 

Anyways, Yo-Yo said there is some room for adjustment, not in the way stated here as you are not seeing the math behind it, but also the gun works, I have tested it if you do things the way you should, the gun works quite well.

 

Nineline,

 

It's clear that I am being dismissed without the information I have provided being examined.

 

If I have somehow insulted someone, then I apologize. That was not my intention.

 

Please encourage the team to seek out someone they trust who is experienced with ballistics. The way you are describing things is not correct.

 

Best of luck to you and the ED team.

Edited June 30, 2019 by Aries144

[NineLine]

Nineline,

 

It's clear that I am being dismissed without the information I have provided being examined.

 

If I have somehow insulted someone, then I apologize. That was not my intention.

 

Please encourage the team to seek out someone who is experienced with ballistics. The way you are describing things is not correct.

 

Best of luck to you and the ED team.

 

No one has been dismissed, and we have a number of team members that are quite qualified to be able to design these systems, Yo-Yo stated that he can make a slight tweak, Wags has shown vs real-world footage that the M-61 in the sim is quite devastating as expected. Even my crappy little quick vid shows that the gun functions quite well.

 

Being reasonable and asking players to also look at their own situations, skill and set up is not unreasonable here as well.

 

So not sure where you are feeling dismissed, you are not understanding how the guns are set up in DCS and are wanting a test environment result from a combat run environment.

 

The best thing here is to wait and see what Yo-Yo does, if anything, because again, the guns shoot stuff up real nice already, and then we will go from there.

 

As a side note, when there is an issue, we ask for tracks. Where are the tracks of the gun not functioning correctly?

Edited June 30, 2019 by NineLine

[Aries144]

Nineline,

 

Respectfully, the point for comparison is all wrong. You can't compare strafing videos to results from DCS, there are too many uncontrollable variables, like pilot error that can't possibly be accounted for. To my knowledge, no other guns' dispersion in DCS were modeled to take into account some average of pilot error, but based on data about the guns' mechanical attributes, is this correct?

 

I'm arguing that any DCS testing should be made to compare with the real world test video evidence we have access to, so the goal should be to recreate that setup in DCS, not to attempt to replicate strafe pass results.

 

If the mechanical performance of the gun is correct with a stationary aircraft, then it's performance when the aircraft is subject to aerodynamic forces and pilot error should be correct as well, so long as the aerodynamic effects are also correct.

 

I think still images might be more helpful here than tracks, but I can provide those as well. However, I want to be certain that I'm not going to spend several hours collecting data only to have it rejected out of hand because it isn't a strafing pass, but made to compare with the linked test video. Can you confirm that an attempt to provide data for comparison with the linked test video would be acceptable?

[WindyTX]

I promise this will be my last 2 cents worth to this debate.

 

Courtesy of Wags video I decided to take another look as flat screen video just doesn't work for me anymore. I did a quick flight in an F18c and in an F14B with a couple of strafe runs in each. In VR the sim has the correct feel for being in the jet, tgts and everything look right. I guess in a way you can blame yourselves for making 99% of it look so right that the 1% now appears as a glaring problem ( Like people complaining that the Hydraulics don't start up right :) )

 

On a standard Acoustic scored Strafe panel I could consistently score over 50% in an F3 and even scored over 70% in the Hawk with an Iron sight ( Once on a nice day ).

 

When I fire the F14B gun the dispersion look dead on, if I mess up I would miss and if I am on target I would get a good score.

 

When I fire the F18c gun its a mess I am not saying I can't destroy a target with it I am just saying that if I was 100% on target on an acoustic panel I would be lucky to score 40% whereas I would be looking at an 80% score in an F3. Maybe there is an F18 pilot would care to comment as I know there are a couple around here.

 

For all I know the F18c gun really is that bad and significantly worse than the F3 ( which I will admit had a very accurate gun just not enough rounds or fire rate ) but as of right now the F18c gun in game is way worse than the F14 in game and worse than an Aden cannon bolted on to the bottom of a Hawk with an Iron sight IRL, which is why I posted this in 1st place.

[NineLine]

I see lots of "its a mess" "looks dead on" etc, terms that are really not worth much to enact changes. But as we have shown the gun does what it is supposed to, also we are not modelling the F/A-18C to match the F-14, we are modelling it to match the F/A-18C, we have seen nothing besides the tweak that Yo-Yo might make in the math to suggest there is an issue.

 

If you want to submit a track that shows the gun is not capable of hitting and killing targets, I will look at it, but from my experience, and what Wags has shown, chances are you will be hard pressed to show anything wrong.

 

Most RL Hornet pilots probably aren't willing to come on here and say its exactly right or it would be better this way for obvious reasons.

 

And on top of that, you want to dismiss real-world strafing videos, but that is what we are going for, not for an aircraft strapped down and locked in place not moving at all firing at a target at 90 degrees of it. You guys suggest we are dismissing you, then turn around and dismiss us.

 

So beyond what Yo-Yo has suggested, I don't see any reason to change how the gun functions right now.