PDA

View Full Version : FYI: FB gun performance tests



XyZspineZyX
10-12-2003, 10:34 AM
Hi folks!

I did some testing on gun performance of ShVAK, MG 151/20, UBS and Browning 0.50 cal in FB. My time is limited, therefore the following findings do not present a truth or something like that, however they are true and I did my best to find equal conditions for every test.

I had a look at:
ROF
muzzle velocity
dispersion
gun power

ROF test were extremely simple. I set unlimited ammo and pressed the trigger. I had the FB console count the rounds I fired within a minute. I?ll name the weapon, the plane I tested with and the rpm. In parenthesis ROF I found in the web. My numbers are quite accurate (I?d say no more than +/- 5), even though I find the numbers for the P-40 a little strange. Interestingly, there now is a difference for sync. There wasn?t in IL-2.

UBS (Yak-3): 900 rpm
UBK (Hurricane fieldmod): 950 rpm (1050)
Browning 0.50 (P-39 Q10): 650rpm
Browning 0.50 (P-40): 733 rpm
Browning 0.50 (P-47): 750 rpm (750)
ShVAK (Yak-3): 800 rpm (800)
ShVAK (La-7): 650 rpm
MG151/20 (109G-6): 720 rpm (740)
MG151/20 (190D-9): 690 rpm

Muzzle velocity:

Well, I didn?t actually measure muzzle velocity. Instead I measured the distance the covered within the first second after leaving the muzzle. These numbers are an approximation (That?s why I rounded to the nearest 50 meters). In parenthesis muzzle velocity in found in the web.

UBS/UBK: 750m (850m/s)
Browning 0.50: 700m (870ms)
ShVAK: 650m (800m/s)
MG 151/20: 550m (800m/s)

Dispersion:

I switched into arcade mode an attacked an smokestack. I set gun convergence to 200 meters and was 200 meters away from the smokestack. I fired a very short burst. The dots represent the hits.

http://mitglied.lycos.de/jaytdee/dispersion.jpg


Gun power:

I set up a small dogfight with me attacking a novice Yak-9u and had FB console count the hits until the Yak was history. I started right behind the enemy and therefore most of the hits came from 6 o?clock. I never fired from more than 200 meters distance, and basically 90% of the hits came from within 100 meters distance. Sometimes I was lucky, snapping of a wing almost instantly, sometimes I had to work hard. For every weapon I made 20 tries, so the average should be about average. The numbers are:

Weapon (number, plane): min hits---average hits---maximum hits
ShVAK (1, Yak.3): 4---16.0---30
MG151/20 (1, Bf 109G-6late): 4---11.6---19
UBS (2, Yak-3): 18---34.9---53
Browning 0.50 (2, P-39 Q-10): 6---30.7---52

I hope you find this helpful to some extend. If you don?t understand how my tests worked, just ask.

XyZspineZyX
10-12-2003, 10:34 AM
Hi folks!

I did some testing on gun performance of ShVAK, MG 151/20, UBS and Browning 0.50 cal in FB. My time is limited, therefore the following findings do not present a truth or something like that, however they are true and I did my best to find equal conditions for every test.

I had a look at:
ROF
muzzle velocity
dispersion
gun power

ROF test were extremely simple. I set unlimited ammo and pressed the trigger. I had the FB console count the rounds I fired within a minute. I?ll name the weapon, the plane I tested with and the rpm. In parenthesis ROF I found in the web. My numbers are quite accurate (I?d say no more than +/- 5), even though I find the numbers for the P-40 a little strange. Interestingly, there now is a difference for sync. There wasn?t in IL-2.

UBS (Yak-3): 900 rpm
UBK (Hurricane fieldmod): 950 rpm (1050)
Browning 0.50 (P-39 Q10): 650rpm
Browning 0.50 (P-40): 733 rpm
Browning 0.50 (P-47): 750 rpm (750)
ShVAK (Yak-3): 800 rpm (800)
ShVAK (La-7): 650 rpm
MG151/20 (109G-6): 720 rpm (740)
MG151/20 (190D-9): 690 rpm

Muzzle velocity:

Well, I didn?t actually measure muzzle velocity. Instead I measured the distance the covered within the first second after leaving the muzzle. These numbers are an approximation (That?s why I rounded to the nearest 50 meters). In parenthesis muzzle velocity in found in the web.

UBS/UBK: 750m (850m/s)
Browning 0.50: 700m (870ms)
ShVAK: 650m (800m/s)
MG 151/20: 550m (800m/s)

Dispersion:

I switched into arcade mode an attacked an smokestack. I set gun convergence to 200 meters and was 200 meters away from the smokestack. I fired a very short burst. The dots represent the hits.

http://mitglied.lycos.de/jaytdee/dispersion.jpg


Gun power:

I set up a small dogfight with me attacking a novice Yak-9u and had FB console count the hits until the Yak was history. I started right behind the enemy and therefore most of the hits came from 6 o?clock. I never fired from more than 200 meters distance, and basically 90% of the hits came from within 100 meters distance. Sometimes I was lucky, snapping of a wing almost instantly, sometimes I had to work hard. For every weapon I made 20 tries, so the average should be about average. The numbers are:

Weapon (number, plane): min hits---average hits---maximum hits
ShVAK (1, Yak.3): 4---16.0---30
MG151/20 (1, Bf 109G-6late): 4---11.6---19
UBS (2, Yak-3): 18---34.9---53
Browning 0.50 (2, P-39 Q-10): 6---30.7---52

I hope you find this helpful to some extend. If you don?t understand how my tests worked, just ask.

XyZspineZyX
10-12-2003, 12:19 PM
Just wanted to add something here. In the P-39 Q-10, the Browning MG's are synced. US synce system sucked and lowered the ROF by half. Oleg stated before that he coded the nose guns with TWICE the ammo at half the power to simulate this. This may help your calculation. The main problem is that the recent patch messed the P-39's nose MG's up. It now has half the ammo at half the dmg. If you want to know the true power of the .50 on a P-39, fly a Q-1 and drain the nose guns first. They will run out before the wing guns (in old IL2 and FB 1.0, the nose guns would last TWICE as long) and use the wing MG's. That will give you an idea of its true power. Unfortunatly you dont have much ammo left in the wing guns, but if your aim is good, it will be enough.

Gib

P.S. Yes, I told Oleg of this bug.

No fancy quote or cool photo.... YET

XyZspineZyX
10-12-2003, 12:23 PM
Some good analysis!

S! Simon.
<center>

'''''''''''''''''''''''''''''''''''''''''''''''''' '''''''''''''''''''''''''''''''''''''''''''''''''' '''''''''''''''''''''''''''''''''''''''''''''''''' '''''''''''''''''''''''''''''''''''''''''''''''''' '''''''''''''''''''''''''''''''''''''''''''''''''' '''''''''''''''''''''''''''''''''''''''''''''''''' ''''''''''''''''''''''''''''''''''''''''''''''''
Download the USAAF campaign folder here (http://mudmovers.com/Sims/FB/fb_essential_files.htm).

http://extremeone.4t.com/images/ex1_soon.jpg
<font color="#000000">It's my attitude not my aptitude that determines my altitude.</font></center>

XyZspineZyX
10-12-2003, 12:37 PM
Nice test you`ve done there, been looking for something like that...

XyZspineZyX
10-12-2003, 12:40 PM
Gibbage1 wrote:
- Just wanted to add something here....

I am not sure that this is still like this in current FB. Before we had the same ROF for synced and not synced guns and the nose guns only did half the damage. Now the ROF is different (650 vs. 750) and the nose guns in the P-39 appear to be as powerful as their wing mounted equivalents. Furthermore, reducing the ammo by 50% and reducing ROF by only 13.3% will shorten the total firing time almost by half (44%). So I THINK that everything is fine with the power and ammo of the Q-10 nose Brownings.

XyZspineZyX
10-12-2003, 12:40 PM
Hey great work JTD for someone with limited time.Your test clearly show that the P47 needs to be taken back and readjusted for synch since the dispersion is way off.

T_O_A_D
10-12-2003, 12:41 PM
Very nice but how did you know you were 200 meters from stack? And was you parked on the ground? I take it you put a static camera close to stack aswell,correct?

It would be nice to get a target range set up, just for the heck of it.

<Left>
131st_VFW (http://www.geocities.com/vfw_131st/index.htm)

<Left>
/i/smilies/16x16_smiley-surprised.gif MY Track IR Fix (http://forums.ubi.com/messages/message_view-topic.asp?name=us_il2sturmovik_ts&id=zwqtg)


<Center>http://home.mchsi.com/~131st_vfw/Mad_toad.jpg </a>

XyZspineZyX
10-12-2003, 09:17 PM
T_O_A_D wrote:
- Very nice but how did you know you were 200 meters
- from stack?

I used the grid in the FMB as reference when I set it up.

- And was you parked on the ground?

Yes.

- I take it you put a static camera close to stack
- aswell,correct?

No, I just zoomed away from my plane.

XyZspineZyX
10-12-2003, 10:09 PM
Excellent tests indeed. The specs are interesting to be sure...thank you for spending the time to do it.

http://freespace.volitionwatch.com/icefire/icefire_tempest.jpg
"Never in the field of human conflict was so much owed by so many to so few." - Winston Churchill

XyZspineZyX
10-13-2003, 01:48 AM
It's always nice when someone does these kinds of tests. Does a lot more than a million "DURRR FREAKIN LASERS DURR" style posts ever will.

I think there is some inconsistency in gun damage in FB - not as some kind of conspiracy or bias - just in terms of the way the game code itself treats different weapons.

As far as my understand of it goes, there's no real 'global' hitpoint rating for any aircraft. Each individual section of an aircraft has its own individual armor rating, and effect on the aircraft when damaged or destroyed. There's definitely a big bonus for weapons that can deal combined AP/HE damage, and for weapons that can strike the same spot consistently.

The ShVak cannons, for instance, seem to fire two different coloured tracers at once. Having two on a nose mount in aircraft such as the La5 and La7 series means that both shells are usually going to strike the same location on an aircraft, or close enough. And when they do, it seems that one shell will deal out AP damage and the other will do HE explosive damage. This is just the bees knees in terms of how FB calculates damage.

The 151/20 cannon on the Bf109 differs from this - for a start there's only one cannon, with more or less the same ROF as one single Shvak. Each hit will do either AP or HE damage from one shell only. Hitting the same damage location more than once is a lot harder if you have to hold the crosshair over a target for twice as long. You can add gunpods to get more or less the same firepower as an La7-3xB20, but to get the same effects from hits you'll have to hit at the exact convergence range - a bit hard when you're trying to smack a target with two wing mount cannons and a nose mounted cannon shaking the plane around.

It's a similar story for the FW-190 - 4x20mm cannons, sure, but getting them all to hit in the same place is a bit more difficult. The differences in ballistics between the wing root 20mms, wing edge MG/FF in models that carry it, and the nose mounted MGs make hitting the same point quite hard. Quite often, even at 'perfect' convergence range, your MGs will strike, say, the top of the rudder surface, and the wing root cannons will strike the elevator. Those are two different locations entirely as far as FBs damage modelling is concerned, and damage dealt to one location doesn't carry over to another. The only exception is from blast/fragmentation damage, and not every shell is going to dish that out.

In my opinion what most people see as damage model or weapon model defects are more the result of a misunderstanding of the way damage is calculated globally in FB, as well as the difference in piloting style between aircraft. The damage values for various weapons have been posted here before, and they all seem to be reasomably calculated. The difference is getting those shells to hit the target and hit it often enough in the same place to matter. I think there is still some work to be done regarding the recoil/plane shake from certain weapons, but I don't think that makes anywhere near as much difference as some people would like to think it does.

I think JTD's done some excellent work here, I've done some experimentation with this sort of thing myself but couldn't come up with a satisfactory way to test it. I think that testing vs a ground object possibly isn't the best way to do it, which is why I didn't use it in my own tests, but it's certainly a much easier way to go about repeating the experiment. JTD finished his work and I didn't finish mine so it seems to have been the right choice :>


http://home.iprimus.com.au/djgwen/fb/worker_parasite.jpg

Need help with NewView? Read this thread. (http://forums.ubi.com/messages/message_view-topic.asp?name=us_il2sturmovik_gd&id=yzbcj)

adlabs6
10-13-2003, 03:05 AM
Gibbage1 wrote:
- Just wanted to add something here. In the P-39
- Q-10, the Browning MG's are synced. US synce system
- sucked and lowered the ROF by half. Oleg stated
- before that he coded the nose guns with TWICE the
- ammo at half the power to simulate this.

So... the nose MG's are firing twice as many shells as they should, but those shells are at half of the correct hit power?

Would this not degrade the effectiveness of the MG's in a raking deflecion shot angle? What happens in a high deflection shot, if only one shell strikes the target critically, or strikes the pilot? The damage level would not be correct! Wouldn't I have to land TWO shells on the engine to get the correct amount of damage? This is not likely in a highly deflected burst.

Am i right on this?

<html>
<body>
<table cellpadding="2" cellspacing="0" border="0" width="600" align="center">
<tbody>
<tr>
<td valign="top" bgcolor="#ffffff">
<font face="Helvetica, Arial, sans-serif"><font color="000000">adlabs<font color="#ff9900">6</font></font>
</td>
</tr>
<tr>
<td valign="top" bgcolor="#42524e">
<div align="center"><font color="#999999">
http://www.geocities.com/adlabs6/B/bin/sigtemp.JPG (http://mudmovers.com/Sims/FB/fb_skins_historical_adlabs6.htm)
<small><font color="#ff6600">NEW</font> at mudmovers! Click the pic to download my skins from mudmovers.com!</small>
</font>
Skinner's Guide at mudmovers (http://mudmovers.com/Sims/FB/fb_skinnersguide.htm) | Skinner's heaven (http://www.1java.org/sh) | IL2skins (http://www.il2skins.com)
<font color="#999999">
My Forgotten Battles Webpage (http://www.geocities.com/adlabs6/B/index.html) Current Wallpaper: <font color="#999999">Bf-109 Morning Run</font></font>

<A HREF="http://forums.ubi.com/messages/message_view-topic.asp?name=us_il2sturmovik_gd&id=zhiwg" TARGET=_blank>"Whirlwind Whiner"
The first of the few</A>
</div>
</td>
</tr>
</tbody>
</table>
</body>
</html>

XyZspineZyX
10-13-2003, 07:18 AM
Someone mention lazers?

XyZspineZyX
10-13-2003, 08:34 AM
FB Console.... is that a 3rd party FB applicationlike the old IL-2 loader. If so.... where can I get it from? Thanks

FuryFighter

XyZspineZyX
10-13-2003, 10:03 AM
FB Console is the thing that appears when you press shift+tab.

XyZspineZyX
10-13-2003, 10:25 AM
adlabs6 wrote:
- So... the nose MG's are firing twice as many shells
- as they should, but those shells are at half of the
- correct hit power?
-
- Would this not degrade the effectiveness of the MG's
- in a raking deflecion shot angle?

It would likely be the case. The figures JtD posted
seemed to show equal effectiveness of the .50s in the
P39Q as VVS 12.7mm guns. What the shots that JtD took,
I don't know.

With regard to the dispersion on the P47 - did you
try the P40 as well? Also were equal numbers of rounds
fired (the more rounds that are fired, the more likely
you are to get an outlier). Basically some sort of
statistical analysis needs to be done on the dispersion
to get an equal playing field.

XyZspineZyX
10-13-2003, 11:17 AM
AaronGT wrote:
- It would likely be the case. The figures JtD posted
- seemed to show equal effectiveness of the .50s in
- the P39Q as VVS 12.7mm guns. What the shots that JtD
- took, I don't know.

I was aiming for the fuselage mainly. However, Brownings went everywhere.

- With regard to the dispersion on the P47 - did you
- try the P40 as well?

Just did, looks more or less the same like P-47.

- Also were equal numbers of
- rounds fired (the more rounds that are fired, the more
- likely you are to get an outlier).

I fired five round per gun with UBS, ShVAK, MG151/20, four rpg with P-39 Brownings and only three rpg with P-47. In total it's five for ShVAK, MG151/20, eight for 0.50(P-39), ten for UBS and 24 for 0.50(P-47). The pattern just gets more dense if you fire repeated short burst. Longer burst will make the aircraft turn and will trow off aim.

- Basically some sort of statistical analysis needs to be
- done on the dispersion to get an equal playing field.

I agree, but I think the impression you get from the screenshots is correct (just not scientific).

XyZspineZyX
10-13-2003, 02:38 PM
JtD wrote:
- I was aiming for the fuselage mainly. However,
- Brownings went everywhere.

I was just wondering if you were doing mainly shots
from the 6, or ones at maximum deflection (i.e.
did any additional factors from synchronised guns
creep in?)

- Just did, looks more or less the same like P-47.

Is it possible to get a pic. (or a source for the
mission set up so I can try it myself when/if I get
some time - work and home are very hectic this week).

- I fired five round per gun with UBS, ShVAK,
- MG151/20, four rpg with P-39 Brownings and only
- three rpg with P-47. In total it's five for ShVAK,
- MG151/20, eight for 0.50(P-39), ten for UBS and 24
- for 0.50(P-47).

It would be nice to get an equal number of rounds per
plane fired, as that makes for a better statistical
analysis. With 5 times as many rounds fired for the P47
then even with the same dispersion, then there is a good
chance of getting more outliers, even with the same
level of dispersion. For a good analysis we really need
equal number of rounds fired.

- I agree, but I think the impression you get from the
- screenshots is correct (just not scientific).

Difficult to know about the spread, though, without
a fair number of rounds fired per plane (100 or more -
Ogre uses 5000 simulated rounds) and with the same number
of rounds per plane. It might help tease out if there
is any difference between P40 and P47 then. The number
of rounds fired so far aren't likely to allow analysis
with a large degree of statistical confidence.

By the way - I am not knocking your work - it is great
work - but we can build on it. If you could PM me a place
where I can find the mission files that would be great.
I'm curious about what the difference between the B239
fuselage and wing guns would be (ditto P39Q1), and UBS/K
in wing and fuselage mounts.

XyZspineZyX
10-13-2003, 04:37 PM
AaronGT wrote:
- I was just wondering if you were doing mainly shots
- from the 6, or ones at maximum deflection (i.e.
- did any additional factors from synchronised guns
- creep in?)

From 6 o'clock.

- It would be nice to get an equal number of rounds
- per plane fired, as that makes for a better statistical
- analysis. With 5 times as many rounds fired for the
- P47 then even with the same dispersion, then there is a
- good chance of getting more outliers, even with the same
- level of dispersion. For a good analysis we really
- need equal number of rounds fired.

This isn't really possible with my setup, since the plane twists sideways. You can see the result esp. with the ShVAK guns of the Hurricane below. Because of this you can only take the height difference as an indicator, but not the width.

- Difficult to know about the spread, though, without
- a fair number of rounds fired per plane (100 or more

I did fire short burst repeatedly and they all looked very similar to the pictures in my first post. They were just set off a little to one or the other side.

- Ogre uses 5000 simulated rounds) and with the same
- number of rounds per plane. It might help tease out if
- there is any difference between P40 and P47 then. The
- number of rounds fired so far aren't likely to allow
- analysis with a large degree of statistical confidence.

No, it's only good for a show. What's up with Ogre and his 5000 simulated rounds? Sadly, I don't happen to know what you are talking about. Is there a link somewhere?

Difference between nose and wing mount of UB and ShVAK in upper half. I compared Yak-3 and Hurricane Fieldmod and this time fired 24 Shots for each picture. I think there is no real difference with the ShVAK and a small one with the UB. UBK is less precise.
Pic of P-40 and B-239 short burst in lower half of picture.
B-239 wing and nose looked alike. If there's a difference than it's smaller than the individual dispersion and my test is not good enough for this.

http://mitglied.lycos.de/jaytdee/dispersion2.jpg


The mission Setup is this one. (Not even zipped.)
http://mitglied.lycos.de/jaytdee/dispersion.mis

XyZspineZyX
10-13-2003, 06:06 PM
JtD wrote:
- No, it's only good for a show. What's up with Ogre
- and his 5000 simulated rounds? Sadly, I don't happen
- to know what you are talking about. Is there a link
- somewhere?

http://www.cris.com/~reaper/gunnery/gunnery.html

XyZspineZyX
10-13-2003, 06:12 PM
Looking back at the top pics it looks like the
.50 dispersion is much greater than the UBS/K
but not really any different between any of the
.50 armed aircraft, at least in the vertical.

Unless there is more horizontal dispersion
on the P47 (hard to control for in tests)
then who knows what the issue some find with the
P47 is due to? Could still be horizontal dispersion,
or something else entirely.

XyZspineZyX
10-13-2003, 07:02 PM
I want to add a chapter to the dispersion issue. I know my method has some shortcomings. So again, what I show is true, and certainly shows a correct tendency. But it's not supposed to be a scientific truth. If you want that, you'll have to do your own research.

In the picture below I superimposed a La-5F silhouette over a dispersion pattern. I found no other way to see hits and misses against a plane this good. The La-5F is 200 meters away and so are the hit dots. Convergence for the guns was 200 meters again. I studied the dispersion pattern to examine which hit would actually have hit the aircraft. So "hits the aircraft" means "shots that would probably have hit the aircraft if it had been there". I'm fairly convinced that I made no mistake with this.
I show two tries with the P-47 and one with the Yak-3.

http://mitglied.lycos.de/jaytdee/hits.jpg


In the P-47 test try one I fired 72 rounds. I counted 10 hits into the fuselage, 2 went into the left wing. Altogether 12 hits, 60 misses.
In the P-47 test try two I fired 32 rounds. I counted 3 hits into the fuselage, 2 into the left wing and three into the right. Altogether 6 hits, 26 misses.
In the Yak-3 test I fired 12 rounds. I counted 11 hits into the fuselage. Altogether 11 hits, 1 miss.

The P-47 has 8 Brownings, with a ROF of 750 rpm each.
Try one shows a 17% to hit chance,
Try two shows a 19% to hit chance.
Given these figures, you'd put about 1050 rpm into a fighter sized target under these conditions.
The Yak-3 has 2 UBS, with a ROF of 900 rpm each.
The test shows a 92% to hit chance.
Given this figure, you'd put about 1650 rpm into a fighter sized target under these conditions.

Taking a more appropiate 2 seconds burst, you'd score 35 hits with a P-47 and 55 with a Yak-3.

As far as I can see there are two pros for the P-47:
1st: You really don't need to be a good gunner.
2nd: Your punch will increase considerably when reducing firing distance and convergence.

On the other hand the Yaks nose mounted guns are totally uneffected by convergence and will hit from zero to (at least) 200 meters equally well, no matter what convergence you set.

Additionally, both planes shake when guns are fired. It might (but likely won't)throw off the aim in a Yak-3, but doesn't matter in a P-47.

XyZspineZyX
10-13-2003, 07:04 PM
Thank you for the link, AaronGT.

XyZspineZyX
10-13-2003, 07:20 PM
Nice job JtD! Thanks for sharing it with us all.

XyZspineZyX
10-14-2003, 03:35 AM
Bump for JTD's great work.


http://home.iprimus.com.au/djgwen/fb/worker_parasite.jpg

Need help with NewView? Read this thread. (http://forums.ubi.com/messages/message_view-topic.asp?name=us_il2sturmovik_gd&id=yzbcj)

XyZspineZyX
10-14-2003, 04:41 AM
See what kind of accuracy you get when firing the 2 mg on any 109 in same test.

I bet it doesn't even come close to the super-lightweight-flying-sniper-rifle (yak3).

I'm kinda suprised the p47 can't score more hits at 200m with an effective ROF of 6000rpm than a yak3 with less than 2000rpm effective. I agree with anyone that thinks the p47 dispersion is way off.



Message Edited on 10/14/03 03:45AM by CrackFerret

Message Edited on 10/14/0303:48AM by CrackFerret

XyZspineZyX
10-14-2003, 10:36 AM
CrackFerret wrote:
- I'm kinda suprised the p47 can't score more hits at
- 200m with an effective ROF of 6000rpm than a yak3
- with less than 2000rpm effective. I agree with
- anyone that thinks the p47 dispersion is way off.

On the face of it (and I noted my concerns with
regard to a statistically significant sample) the P47
and P40 dispersion don't look any different from each
other.

Looking at Ogre's site, for an F6F (6 .50s) firing
at an A6M at 200 yards with 200 yard convergence,
with zero aiming error (as per JtD's tests) the
maximal hit rate is 55%. But in this test Ogre doesn't
take into account any effect of aerodynamic effects
(including wing flex), so if all factors are accounted
for in FB, we'd expect less than 55% in FB. As to
whether we'd expect 20%, that's another matter! It could
be that FB does account for wing flex, etc., but does
so with just an additional bit of dispersion, that is
constant no matter whether you are in the air or not,
or all sorts of things.

Ogre does do an analysis including aerodynamic effects,
but I am not sure how to interpret the graphs as there
are no labels on the axes on these plots.

The out of convergence plots are worth looking at too.

Sadly he doesn't do an analysis of cowl guns as well
as wing mounted guns, so it is hard to use Ogre's
analysis to compare to the UBS case.

I suppose the next thing is to ask JtD if he can
create an image with the bullet hits as before, but
with a vertical scale at the appropriate magnification
for 200 yards distance. We can then use that to calculate
the mils of dispersion.

XyZspineZyX
10-14-2003, 11:00 AM
This is a outstanding Test JtD! well done!

If you are bored please expand it!Special the compare of dispersion to Plane size is very intresting.

In my feeling the dispersion of the 50er cal. rounds is overmodelled,special the initial dispersion.At long bursts of 2 seconds i can understand it but at short bursts it should be more concentrated.But maybe this is a general fault of FB Bullet engine.

XyZspineZyX
10-14-2003, 12:10 PM
AaronGT wrote:
- I suppose the next thing is to ask JtD if he can
- create an image with the bullet hits as before, but
- with a vertical scale at the appropriate
- magnification for 200 yards distance. We can then use that to
- calculate the mils of dispersion.

Yes, this would be great. But actually I can't do much better than I did with the La-5F silhouette. Very roughly: The wingspan is 10 meters, the dispersion pattern has about one third the size. So at 200 meters you have a 3x3 (4m > x >2.5m) meters area. The surface area of a La-5F from this angle should be somewhere between 2 and 3 square meters (fuselage about 1.7). A dispersion pattern with a size of about 10 to 15 square meters would also match well with the to hit chance. Anyway, this is well into speculation as I know to little and can't get more out of my approach.

Oddly, the dispersion area does not seem to be a circle, but a square. You can see this better in this picture. I tested ShKAS and PWU machine guns with 100 meters distance and 100 meters convergence. (They are the best for long bursts because the little recoil doesn't move the plane).

http://mitglied.lycos.de/jaytdee/squares.jpg


CrackFerret, I also had a short glance at the MG 17 of an BF 109 E. It was the only weapon I've seen so far that came close to the UBS.

Col.Kurtz, I am not exactly bored. At the moment I have the opportunity to choose when to do what. Work will still be there tomorrow. I'll probably hate me in two months time for the choices I made now. ;-)

XyZspineZyX
10-14-2003, 02:02 PM
A circle is no doubt a lot more intensive to calculate when you're talking about the amount of rounds FB tries to keep active at one time and the various guns it simulates. And probably doesn't boil down to being that much different from a good ol' square, with some exceptions.

NB: first time anyone's ever criticised oleg for not cutting corners? ;>


http://home.iprimus.com.au/djgwen/fb/worker_parasite.jpg

Need help with NewView? Read this thread. (http://forums.ubi.com/messages/message_view-topic.asp?name=us_il2sturmovik_gd&id=yzbcj)

XyZspineZyX
10-14-2003, 03:55 PM
JtD wrote:
- Yes, this would be great. But actually I can't do
- much better than I did with the La-5F silhouette.
- Very roughly: The wingspan is 10 meters, the
- dispersion pattern has about one third the size. So
- at 200 meters you have a 3x3 (4m > x >2.5m) meters
- area. The surface area of a La-5F from this angle
- should be somewhere between 2 and 3 square meters
- (fuselage about 1.7). A dispersion pattern with a
- size of about 10 to 15 square meters would also
- match well with the to hit chance. Anyway, this is
- well into speculation as I know to little and can't
- get more out of my approach.

Ok working with those figures, the vertical spread
is about 8 feet, at a distance of 600 feet, for about
12 mils of dispersion, about 50% more than the figures
quoted by I can't remember who a few days ago. The
figures were 8 mils for a large collection of aircraft
for, some of which had nose mounted guns, so 12 mils
for wing mounted guns is 50% more than that figure,
but I don't know what the mils for a wing mounted
gun would be. If I knew what the plots on Ogre's
site were for various gun types (the plots with
aerodynamic factors included) we might have another
form of comparasion.

- Oddly, the dispersion area does not seem to be a
- circle, but a square.

Weird. Maybe an approximation used in the sim, or a
bug. The ratio of the density in a square as compared
to a circle would be PI:4, or about a 75% reduction.
Also if you measured the dispersion from the largest
extent of a square - i.e. across the diagonals (in
the mil computation above I just used top-bottom distance
to avoid this) you'd have a dispersion sqrt(2) greater
than the dispersion if everything was firing into a
circle.

XyZspineZyX
10-14-2003, 04:02 PM
Looking at Ogre's site (a reference for actual
dispersion for a wing mounted gun, other than
an elusive, and persumably averaged 8 mils would
be good) indicates a spread of 8 feet at 200 yards,
without taking into account aerodynamic factors.
This is about 13 mils.