Gunnery Question

[The Fugitive]

Question: What does vsync have to do with anything?

If you have it off your video card will be throwing more "pictures" at your monitor than it can display. With the extra pictures your monitor won't be showing you the right pictures all the time and you'll be shooting at things that aren't where you think they are so you will miss and not score "hits" and so you get rubber bullet syndrome. Having your monitor synced to your card shows you what is there and you can hit what your aiming at with out the video misses causing problems.

[mtnman]

So our tracers are not giving us the false image warned about to 8th airforce gunners that the flare circle you are visualy tracking with your eye is delayed from where the actual round is?

I've looked at slow motion film of my offline drone chasing at a few feet from the drones parts being impacted by my rounds. The tracers looked and acted alot more like plazma bolts from games like Quake and Halo. If the leading tip of the streak touched the airframe a hit flash resulted. Then there was the film that surprised me into realising not all the rounds in your belt in AH are tracers....especialy with the Mk108.

That's a good question...

I can't say as I've ever seen anything in my testing that would lead me to believe the tracers fly a different trajectory than the rest of the rounds, but I've never tested it specifically either.  Then again, most of the testing I've done probably wouldn't highlight a difference (if there is one).

The one test that really comes to mind though (where I really used tracers to land hits) would lead me to believe that the tracers are an accurate (or at least an "accurate enough") representation of where the "real" rounds go.  In that test I was really testing the trajectory model while firing inverted and on knife-edge.  I was using the lead-computing gunsight in the TA, and found that the tracers DID go where the bullets were going, and could be used to hit my target.  However, the lead-computing gunsight DID NOT give an accurate representation of the required aiming point needed to score hits.

It would be interesting to configure a solid method to test the tracers vs. bullets impact points.  I wonder if the film viewer tells a precisely accurate story?

BTW, I'm really enjoying those dot-sights you made for me!  I haven't had nearly enough time in the game to stay proficient, but I believe I've settled on the "8 pixel" as my favorite.  I actually think I like the 6P better in some cases though, so I'll have to play around with it some more.  It gives me all the info I want, but it doesn't distract me from concentrating on my target.  I've always preferred "fine" sights over "course" and these fit the bill!

Thanks again for making those up for me!  

[bustr]

The lead comp gunsight seems to set your lead to impact either the tip of the inside wing in a turn or if you are high or low 6 the tip of the aircrafts nose. I have knocked off the wing in a turn, stopped firing, then resumed firing and the impact point was for the now missing wing tip I just shot off. Maybe it's mapping a disk that represents the location of the aircraft based on the center of it's x-y-z and diameter based on wingspan. I have noticed the impact point moves closer to the fuslage the closer you get.

Wonder if that has anything to do with offline you can dismantle a bomber down to a flying fuslage. When you finaly kill said fuslage the resulting explosion throws out a pair of wings in the debre. Was on full zoom once offline at about 100 feet flying slow trying to single shot with a MK108 at a dismantled fuslage only bomber's rear turrent  and watch the results. Guess I killed the pilot on an over shoot and a pair of wings materialised out of nowhere in the explosion right in my face on full zoom.

Glad you liked the dots.

[mtnman]

What i found with the lc gunsight was that it seemed to get comfused when i wasn't wings level.  It wasn't real noticeable in a shallow bank, and was difficult to detect if i was pulling more than 1g.

It wasn't something I was looking for, it was just something that made me say "Oh...  Look at that...".  I didn't explore it more than just enough to understand what I was seeing...  I was actually doing some trajectory tests as a result of a long drawn-out BBS conversation that some of the trainers and I were having.

Rolex and I got together to run some tests in the TA as a result.

What I found was that if I flipped inverted behind Rolex (who was on auto-level) the LC gunsight didn't take into effect the fact that I was upside down, and as a result the LCG was calculating the elevation of the aim-point "backwards".  The elevation of the green "+" was wrong, and if I aimed precisely at it I could not hit Rolex at all (who appeared to be a stationary target in front of me).

I had to aim AWAY from the green + to hit Rolex.

We tried it both ways (I was an auto-level target for Rolex) and had identical results. 

This was on a target that was on auto-level, and I was manually-trimmed to fly inverted behind him.  Essentially, I was about as close as I could get to being on auto-level inverted while firing on a target that was on auto-level in front of me.  The LCG was "right" while I was right-side up, but "wrong" while I was inverted.

We then wondered at what point the LCG started to display "problems" with calculating the trajectory changes that result from banking out of wings-level flight.  Our next test was with a target on auto-level again, but the firing plane was flying in "knife-edge" behind him.  We believed we were already seeing the issue there too (but it's really difficult to maintain knife-edge and fire more than brief little snapshots which makes it tough to draw conclusive results), so I believe that the "problem" shows up any time you fire with the wings banked.

If you were firing banked while pulling G's though (like in a turn, firing on a turning aircraft) you'd never be able to detect what we saw.  What you might notice instead is what you found, where it looks like the aim-point shifts "aim-point locations" on the target plane.

And again, this isn't what we were testing, and we didn't spend much time on it.  Just enough to figure out the "patterns" and understand what we were seeing.

The interesting point for us was that the modeled trajectory of the bullets in AH appear to behave like they should in RL, even in a bank, and even inverted or while flying at extreme nose-up or nose-down attitude.  Along with that, the tracers were "right in there" with the rest of the bullets (or at least close enough so that we didn't detect any discrepancy between them and the rest of the bullets).

What I took away from the tests was that the bullets fly "right", it's just the LCG that has some quirks.  It's one of the reasons I always recommended my students used it as minimally as possible.  I didn't want them to start paying too much attention to it if it wasn't quite right...  It's also why I recommend they pay more attention to the tracers than the gunsight.  The gunsight also starts to be an "inaccurate" representation of where your bullets are headed once you begin banking, and gets to be very inaccurate when you pull more (or less) than 1G. 

The tracers show where the bullets are going.  The sight and LCG don't necessarily do that...

And yes, I agree that the info the tracers give you can be imperfect due to limitations in 3D rendering and their overall size and lack of transparency.  I still think they're the best tool we have when it comes to learning gunnery.

[bustr]

I've added a ghost line ballistic drop calculator to a gunsight of 12Mil below reticule center. Between 200-600 if I hold it on the drones the wing mounted bullet streams land 98 percent of the time into the target destroying it with a 1 second burst. This is in horizontal chase with low G load and speed.

I've found as I increase my speed which then increases the G in my turn with the drone. Where at slower speeds a 600d lead was by 40-50Mil, I'm now leading with the right hand side of the windscreen. This becomes quite a problem at 400 and 200 where I will have to snap my nose ahead into a lead to place a few rounds on the drone as the G increases.

On zoom the tracer streams curve away and out along the line of the drones rudder and outside wingtip unless I manhandle the nose of my fighter inside of the turn. You can really see this from a P38 or 262 on full zoom at high speed and G when you think you have the drone lined up in the turn and your tracers stream off and out to the right of the drone like a fire hose.

If instead I G unload and fly straight to intercept the turning drones flight path. My original hold offs along with the shadow line drop compensator again place a 1sec burst into the turning con 200-600 destroying it.

I've watched high speed turning shots offline in a K4 on zoom and it looks like I'm tossing flaming grapefruits down and under my own wing. But, as soon as I unload G and fly straight to intercept the drone my drop compensation and hold off's work as calculated. Any kind of nose gun mounting seems to be drasticly effected by G loading in turning shots. Guess you shouldnt think like a spit pilot while flying a 38, yak, la or 109F-G-K-(non gondola) in a furball.

[mtnman]

Very good points, and well explained.

You've pointed out the secret to higher hit% and quick kills, regardless of gun type or mount location.

[NikonGuy]

If you have it off your video card will be throwing more "pictures" at your monitor than it can display. With the extra pictures your monitor won't be showing you the right pictures all the time and you'll be shooting at things that aren't where you think they are so you will miss and not score "hits" and so you get rubber bullet syndrome. Having your monitor synced to your card shows you what is there and you can hit what your aiming at with out the video misses causing problems.

Well that makes perfect sense Cheers <S>

[nrshida]

Have you got any films of your gunnery bustr? You seem to have gone into it quite deeply.

[bustr]

Ive deleted all of my drone chasing films that I took to SloMo watch the MK108 and to help develop the 12Mil average drop compensation for wing guns at 200-600.

I started testing and researching this 2 years ago becasue it seemed the MK108's lateral dispersion was too narrow compaired to a spiral dispersion chart I found from Rheinmetall-Borsig. It's drop was correct but, it's dispersion way to kind out to 600. Past 100M the only thing you might be able to hit is a 100ft wingspan bomber with a single MK108 based on Rheinmetall-Borsig data. The spiral dispersion radius from the test was 9.4M or roughly 30 feet. The document was not specific as to this being random fliers or the average spiral due to the 1:16 grooves in the 23 inch, 1.8 inch diameter bore to induce high speed spin to unlock the detonator.

There is nothing unique in the films except from the drones perspective at times zoomed into the impact areas looking back to the attacker. What constitutes an impact may surprise you. I've watched a tracer round touch the rear edge of a rudder and effectivly have an even half of the tracer pass sliding along the surface of each side of the rudder on past the aircraft and not register a hit sprite. I think most of the time we watch films from the attackers perspective and spend less time on the target's.

If you have the time you can see everything I've described offline chasing drones and shooting at full zoom. The ideal shot solution starts with very littel induced G. This way you keep all of your lead computation as percentages of the radius of a 100Mil circle. Other than with the MK108 due to it's slow velocity(500M/sec). And generaly 12Mil of elevation to which you adjust in your stream to hold sprites on the target. But as a general starting point with wing guns 200-600 12Mil elevation and apropriate lead based on G in a turn is an initial aim point. Nose mounted is a bit less at 10Mil. You still need to watch your tracers and hit sprites to tweek it. But, at least now when I miss I miss very small. I get alot more assists these days.

Both the British and the Germans taught deflection shooting as a percentage of the radius of the primary graticule ring which was 100Mil in diameter. Both taught that you needed to know when the deflection percentage was outside of the ring radus. At that point G forces would cause you to miss behind the con and attempting to compensate ment snapping your lead possibley into a blackout condition while shooting blind under your nose.

If I'm not mistaken a description of Joacim Marsallas famous snap shot was to chop throttel and snap his nose ahead of the con he had forced to make a break turn and open fire. He had a wingman to keep preassure on the target so it would not reverse him if he missed and dumped his E.

Ever try to fly an La7 like a spit5 in a furball and make the same high G lead shots with the La7 you did with the spit5? And did you notice in the spit5 just befor you got into a lufbry circle with your con, you had been missing constantly behind the con just as the circle started as the G's increased in your turning manuvers?

The WW2 fighter Pilots manual "Bag the Hun" is a very good reference to what I'm blatherskiting about. It works in this game if you use zoom a bit more often to see your con's real angle OFF from your line of travel.

http://bs.beckament.net/files_pub/FlightSim/Pilot%20Guides/Bag_the_Hun.pdf

[nrshida]

Thanks. I can see you've done a lot of work. I am familiar with that manual and also the German equivalent which has some funny content:

http://bbs.hitechcreations.com/smf/index.php/topic,319501.0.html

I also do a bit of practice offline and made a sort of primitive deflection gunsight / system, but I can't think in Mils or anything else, I just use the position of the vertical stabiliser relative to the wing and adjust the lead according to my ring sight. I tried to make a ballistics chart for the Ho-5 too.

I just fire tiny bursts at 600 yards. Aces High is the only opportunity I have to practice gunnery as I'm not allowed a real gun where I live 

[bustr]

Make a new gunsight in 512x512 mode. That way 1Mil = 2Pixel. Remember what ever you name the gunsight file, (abc.bmp) you will need a corrisponding text file with the number 256 in it saved as (abc.mil). Place that in your sights folder along with the bmp file. I've found if you save it as (abc.mil.txt) first, then rename it after saving it by removing the (.txt) you avoid Aces High not seeing your (.mil) file.

1. Create a new bmp file 512x512 and save it in 8bit format.
2. Fill it first with black.
3. Set your grid to pixels and at (256, 256) as your center create a dot 8-14pixel dia.
4. Choose your Elipse/circle tool at 3pixel width antialiasing enabled and pull a circle from (155, 155) or 100Mil diameter ending at (355, 355).
5. Choose your line tool and set it to 1pixel width and dotted format. Pull a horizontal line across the circle along 284 which will give you a drop compensation line about 14Mil below gunsight center. As you test offline you can move it up if needed or move it down.

512x512, 1Mil = 2Pixel

284pix-256pix=28pix  28pix/2=14Mil

100Mil dia = 200Pix dia.

Use what ever color works best for your eyes. I've found White, Cyan, Yellow and Orange work best in this game for gunsights. If the gunsight shows up very tiny then your (.mil) file was forgotten or you did not save it and rename it the way I mentioned.

The drop at 14Mil works best with wing mounted guns. A 10-12Mil works for nose mounted. Try a spit16 convergence MG/Can300. Shoot using 3/4 zoom to make use of the drop line. Between 200-600 you pretty much hold the line of the cons wings in the line of the drop line and pull the trigger. Being on zoom you will see the tracer stream arch up and curve out and down into the drone. You will have to work on the lead and make sight corrections to hold the stream on the con.

A rule of thumb for lead at distance if you are within G tollerances which following the offline drones does for you.

600-50Mil lead
400-25Mil lead
200-10-12Mil lead
Note: As you increase G you start increasing these lead numbers in response.

Remember that as the drone's angle of travel begins to exceed 20 degrees from yours towards 90 degrees between 200-600 you begin talking lead in diameters of the 100Mil diameter reticule. The 100Mil reticle is about 1/10 of the distance to the target. 20 yards at 200 and 60 yards at 600. Another rule of thumb is all of the rounds in your 1sec burst stream will travel through a 600yd point in about 1 second except for the MK108. The drone at 90 degrees to your line of travel will travel about 100yd in 1 second. Lead is about running that stream through a point that the drone will travel into in that second.

Consider the British M2 Ring and Bar gunsight. 100Mil ring and the space in the bar was often set to a 30ft wingspan at 200. That space would also be a 50Mil circle like the US NAVY Mk8 gunsight 100Mil primary ring, 50Mil secondary. Or you can use the german method of percentage of radii of 100Mil for lead to the angle of travel of the con.

If you have a copy of Schiessfibel or (Horrido), you will see in the section on how to lead pg.14, in the picture of the reticule, drop compensation of about 10-12Mil. Remember drop compensation changes depending on your relative altitiude and angle of attack to your drone. The higher up you are you will aim ahead of the drone by reticle rings and the same from below. The 100Mil ring was a general standard in WW2 for lead hold off by fractions of the ring to a cons angle of travel as an eyeball computing device.

Sadly that means practice, practice, practice, sight picture, sight picture, sight picture. I've added a shadow drop line to my primary gunsights to make it easier for me to calculate {y} since it stays preaty constant to about 600 yards in this game through much of the average guns solutions we enter into. The majority of my kill shots are 100-300.

----------------------------------------------------------------------------------------------------------------------------------------
Note: The Mk108's drop comp of 10-12Mil works to roughly 250 then drops off drasticly past that.

I've gotten a better translation of Parabolischer Rechtsdrall from the Rheinmetall-Borsig MK108 testing document I posted here last year.

Parabolischer Rechtsdrall - Parabolic Drift or spin drift.

A.) Projectile or bullet length: longer projectiles experience more gyroscopic drift because they produce more lateral "lift" for a given yaw angle.

3cm M-Gesch Projectile length - 146mm-5.7inch

B.) Spin rate: faster spin rates will produce more gyroscopic drift because the nose ends up pointing farther to the side.

MK108 bore 1:16 to induce high spin rate to unlock the detonator.
 
C.) Range, time of flight and trajectory height: gyroscopic drift increases with all of these variables.

500M/sec = long time of flight to target.
----------------
MK108 30mm

100 - 1ft drop -1ft dispersion - right hand spin drift aprox 9.4m - 30 feet. 35degree 55' 41"
250 - 3ft drop - 2ft dispersion - right hand spin drift aprox 28m - 91 feet. 107degree 7' 55"
400 - 12ft drop - 3ft dispersion - right hand spin drift aprox 49.7m - 163 feet. 190degree 5' 40"

Looks like it shoots a right hand curve ball into the next door ballpark. As of yet I cannot find a document or pilot interview explaining how you aimed the Mk108 knowing it shoots a right hand curve ball. This could account for some of the gun problems experienced by combat groups working out firing problems with the MK108. Bet that is why 40mm auto morters today which are decendants of the MK108 have such short rounds or popup stabalisation fins.

[The Fugitive]

Bustr, while all of that is very interesting, and most likely can help with figuring your lead, reading it gives me a headache! Maybe if you added a couple of picture I could wrap my head around this stuff.

[bustr]

Thats ok. Download Bag the Hun for "angle OFF" lead.

Drop compensation is the small amount of elevation you need to place rounds ahead of your con rather than place the center dot on him and miss just short of his tail becasue forward speed is not compensated.

Take the AH default.bmp and place a horizontal line just below the center dot at the first point 1/3 down to the bottom of the circle. About 28 pixels down or on pixel line 158. Offline set the convergence in a spit16 to 300. Chase the drones. Between 200-600 use zoom, and account for lead while placing the line just under or on the drone. You now have drop compensation. On zoom you will watch your tracers arch up and onto your drone.

Some of my squad mates use this and hit everything they aim at in low/medium G turns 200-600. Some couldn't get used to it. Some won't use zoom to shoot.

[The Fugitive]

Thats ok. Download Bag the Hun for "angle OFF" lead.

Drop compensation is the small amount of elevation you need to place rounds ahead of your con rather than place the center dot on him and miss just short of his tail becasue forward speed is not compensated.

Take the AH default.bmp and place a horizontal line just below the center dot at the first point 1/3 down to the bottom of the circle. About 28 pixels down or on pixel line 158. Offline set the convergence in a spit16 to 300. Chase the drones. Between 200-600 use zoom, and account for lead while placing the line just under or on the drone. You now have drop compensation. On zoom you will watch your tracers arch up and onto your drone.

Some of my squad mates use this and hit everything they aim at in low/medium G turns 200-600. Some couldn't get used to it. Some won't use zoom to shoot.

Thanks, that helps get it set in my mind what your doing. I'll give it a go. I'm "for" anything that can help me with my aim!

[GradeyShane]

Maybe Im out of whack here but have you tried the easy stuff first?  I noticed a while back that I'd get right up in someone's butt and pull the trigger with my pip right on the target...and miss.  I had to adjust my convergence to keep my rounds from shooting over the top of my target, or if inside my convergence, aim lower.

Maybe nothing to do with what you are seeing and all the tech details here are very good info..but was wondering if it could be something like what I was getting with the rounds lobbing over.