PDA

View Full Version : Turn time for TA-183 misquoted in Hardball's and Compare



mortoma
05-16-2010, 08:35 PM
For those who care, the turn time at 1000 meters for the TA-183 should be around 21 seconds, not 31 or so seconds as shown in Hardball's Viewer or IL2 Compare. It was first a mistake in Compare and Hardball got most of his data from Compare.

Trying out about 10 corners speeds at 1000 meters altitude, I was able to achieve 21 seconds in a full circle through 360 degrees but I forget at what speed. So I think what happened was some sort of typo originally. I always thought something was up because it always felt more agile in turns than a Me-262, which has a turn time of 30 to 32 seconds. I turned out to be right in my suspicions.

AndyJWest
05-16-2010, 09:19 PM
The version of IL-2 Compare I have gives a best turn time of 39.15s in the data provided textually, and about the same on the graph: http://i958.photobucket.com/albums/ae65/ajv00987k/Ta-183-turn.jpg
I can't see how a 'typo' could account for the graph plot, so I'd suspect there is more to it than that. I've seen suggestions that the turn data in IL-2 compare is suspect anyway, but I'm not sure how to independently verify it - measuring sustained turn speed isn't easy. I may be able to come up with a rough figure using my prototype autopilot, but it wasn't designed to do this. Part of the problem may be that jet thrust is much more tied to speed than with a prop-driven plane, so a slowly-decaying turn-speed may look 'sustained' without careful measurement.

Kettenhunde
05-16-2010, 10:28 PM
Part of the problem may be that jet thrust is much more tied to speed

No, a jet's thrust is considered constant at a given altitude independent of velocity.

Power has a velocity component and a propeller aircraft power is considered constant at a given altitude. Therefore its thrust varies with velocity.

AndyJWest
05-16-2010, 10:49 PM
Originally posted by Kettenhunde:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Part of the problem may be that jet thrust is much more tied to speed

No, a jet's thrust is considered constant at a given altitude independent of velocity.

Power has a velocity component and a propeller aircraft power is considered constant at a given altitude. Therefore its thrust varies with velocity. </div></BLOCKQUOTE>

If I was being pedantic, I'd ask if "a jet's thrust is considered constant", then by whom? But why should I bother? I don't give a monkey's who considers what, when, or why when it comes to questions of aircraft performance any more, unless it is backed up with (a) evidence, and (b) an explanation of what the 'heck' this has to do with the point I was trying to make. I'm no longer the slightest bit interested in your nit-picking pedantry, Kettenhunde. Does your 'insight' add the slightest to the question that Mortoma raised? No. Did you think it was actually going to help answer the question? I doubt it very much...

Kettenhunde
05-16-2010, 11:45 PM
AndyJWest you shouldn't say anything when you don't know what you are talking about! The guy would walk off with the wrong concept.


http://img85.imageshack.us/img85/2152/thrustproducers.jpg (http://img85.imageshack.us/i/thrustproducers.jpg/)

http://img408.imageshack.us/img408/8890/thrustandpowerproducter.jpg (http://img408.imageshack.us/i/thrustandpowerproducter.jpg/)

Knowing this, it is easy to see that the shape of the curve is not correct for a thrust producer.

AndyJWest
05-16-2010, 11:59 PM
"overall, the thrust of a turbojet remains relatively constant over a wide range of speeds"

Quite possibly. This isn't what you said though:
"a jet's thrust is considered constant at a given altitude independent of velocity."

'is' doesn't mean 'relatively', and 'independent of' doesn't mean 'over a wide range'. Your sole purpose for posting in this thread seems to have been to score points for pedantry. If you are going to be a pedant, at least try to be correct. http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif

Kettenhunde
05-17-2010, 12:02 AM
Trying out about 10 corners speeds at 1000 meters altitude,


The other thing mortoma is the graph presented shows sustained turn performance at 1000 meters.

Corner speed is not sustainable performance for a Ta183.

It is not the same thing as what you tested.

Kettenhunde
05-17-2010, 12:12 AM
"a jet's thrust is considered constant at a given altitude independent of velocity."

http://img714.imageshack.us/img714/2834/jetplanethrust.jpg (http://img714.imageshack.us/i/jetplanethrust.jpg/)

For the purposes of determining Jet performance, we consider thrust constant at a given altitude. It only varies if the throttle is moved. At 100% throttle the fuel consumption is the same whether the aircraft is sitting still or moving.

Rpm and altitude are the only things that vary a Jets thrust production at a given fuel consumption. Therefore, at a constant rpm and altitude, thrust is considered constant.

Thrust = Fuel Flow / Thrust Specific Fuel Consumption

Look, what you put out was totally wrong and not even close in principle.

Jet thrust is NOT tied to speed...

That means it will produce a different shape to the turn performance curve posted above.

Propeller aircraft thrust is tied to speed.

AndyJWest
05-17-2010, 12:30 AM
Rpm and altitude are the only things that vary a Jets thrust production at a given fuel consumption.
Assuming this is true, can you explain how the engine RPM will be independent of velocity?

And once again, you are referring to texts that say 'we consider' this.

I'm quite prepared to admit I was inexact, or even just plain wrong, in my reply to Mortoma. I still fail to see why you chose to hijack the thread with pointless nit-picking, rather than addressing his original question - unless you aren't interested in anything except boosting your own ego.

Kettenhunde
05-17-2010, 12:37 AM
Assuming this is true, can you explain how the engine RPM will be independent of velocity?

Can you rev your car engine to 4000 rpm in neutral in the driveway? Rpm is controlled by the throttle.

Well you can do the same to the Jet engine by using the throttle. In fact for several methods of determining the performance of a jet, rpm is expressed as a percentage range not an actual rpm.

http://img193.imageshack.us/img193/766/jetthrustvsrpm.jpg (http://img193.imageshack.us/i/jetthrustvsrpm.jpg/)

At very very low speeds and very very fast speeds you will see some variation in thrust in a jet. The low speed variance is insignificant and your jets are not going fast enough that the high speed variance is a factor.

AndyJWest
05-17-2010, 01:02 AM
Ok, Kettenhunde, you've now admitted that a jet engine's thrust is affected by velocity. will you now do the decent thing and admit that the whole question was irrelevant to Mortoma's question anyway?

We can both agree that his measurement of 'corner speed' is misleading if one is trying to measure sustained turn. We've also concluded that thrust and/or power is velocity related, which must make measuring sustained turn rates more difficult, though no doubt if one had the appropriate mathematical formulae, one could allow for this. The point is that IL-2 Compare data was supposedly based on such formulae, and may not match what the sim actually does. As with our discussions about ground effect, one cannot say that IL-2 is wrong without also providing evidence for what is right, and bald assertions that formula X or formula Y are correct don't constitute evidence, as far as I can see.

M_Gunz
05-17-2010, 02:34 AM
Mortoma, the AP used to get IL2Compare data can't handle flying near stall. It's good for comparisons, not turn data.

They use aileron to maintain wings level or bank angle, if there's slip correction it reads 'the ball' and both of
those are mistakes close to stall. Maybe some day someone will code in the necessary shift, rudder instead of aileron
and bank instead of slip.
It does leave me wondering how the AI manages to land though. Perhaps it cheats.

TheGrunch
05-17-2010, 02:37 AM
Yeah, it does. You can see it from the utterly unreal lineup they make. They just pivot in place until they're in line.

AndyJWest
05-17-2010, 08:35 AM
Originally posted by M_Gunz:
Mortoma, the AP used to get IL2Compare data can't handle flying near stall. It's good for comparisons, not turn data.

They use aileron to maintain wings level or bank angle, if there's slip correction it reads 'the ball' and both of those are mistakes close to stall. Maybe some day someone will code in the necessary shift, rudder instead of aileron and bank instead of slip.

It does leave me wondering how the AI manages to land though. Perhaps it cheats.

If IL-2 Compare is deriving data from an autopilot, it is presumably using DeviceLink, as mine does, and not the IL-2 AI code - they are two different things, with differing performance characteristics - the AI uses a simplified model.

And yes, the AI cheats on landing (and on takeoff for that matter).

Bremspropeller
05-17-2010, 09:50 AM
Ok, Kettenhunde, you've now admitted that a jet engine's thrust is affected by velocity.

It's not, unless you're almost standing still or going fast as hell.

A RPM setting will give a thrust-value, leading to a specific airspeed (at which drag and thrust eventually are equal).
Thrust production (i.e. how much thrust can be produced at which speed), however is constant over constant RPM.
That is outside the above stated boundaries.

M_Gunz
05-17-2010, 10:05 AM
Originally posted by AndyJWest:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by M_Gunz:
Mortoma, the AP used to get IL2Compare data can't handle flying near stall. It's good for comparisons, not turn data.

They use aileron to maintain wings level or bank angle, if there's slip correction it reads 'the ball' and both of those are mistakes close to stall. Maybe some day someone will code in the necessary shift, rudder instead of aileron and bank instead of slip.

It does leave me wondering how the AI manages to land though. Perhaps it cheats.

If IL-2 Compare is deriving data from an autopilot, it is presumably using DeviceLink, as mine does, and not the IL-2 AI code - they are two different things, with differing performance characteristics - the AI uses a simplified model.

And yes, the AI cheats on landing (and on takeoff for that matter). </div></BLOCKQUOTE>

Or.. Youss might have from the start made IL2C in cooperation with Oleg and used parts of the AI.
Of course just what code mechanism the AI uses to interface with the game engine is beyond debate, but...

mortoma
05-17-2010, 10:27 AM
Originally posted by AndyJWest:
The version of IL-2 Compare I have gives a best turn time of 39.15s in the data provided textually, and about the same on the graph: http://i958.photobucket.com/albums/ae65/ajv00987k/Ta-183-turn.jpg
I can't see how a 'typo' could account for the graph plot, so I'd suspect there is more to it than that. I've seen suggestions that the turn data in IL-2 compare is suspect anyway, but I'm not sure how to independently verify it - measuring sustained turn speed isn't easy. I may be able to come up with a rough figure using my prototype autopilot, but it wasn't designed to do this. Part of the problem may be that jet thrust is much more tied to speed than with a prop-driven plane, so a slowly-decaying turn-speed may look 'sustained' without careful measurement. Wow that's weird because my version says 31.28 seconds and the graph agrees with mine too!! So much for IL2 Compare being very useful. Oh well.. http://forums.ubi.com/images/smilies/16x16_smiley-sad.gif

AndyJWest
05-17-2010, 10:27 AM
MG, there is no way to use 'parts of tha AI' with the standard DeviceLink interface, as far as I'm aware. I'm not even sure how relevant AI code is to the question of measuring pilot-controlled sustained turn rates anyway - this is a FM issue.

AndyJWest
05-17-2010, 10:30 AM
Mortoma, from memory, the Ta-183 turn performance may have been altered in the 4.08 patch - perhaps the IL2C version you have has later data than mine.

Edit -- Yes, the FM may have been adjusted - from ReadMe_408m.rtf:


-Changed parameters for the Ta-183, Me-262HG-II, Lerche-III, Bf-109F-2, MiG-13/I-250, Yak-3R;

thefruitbat
05-17-2010, 10:38 AM
from day one when youss first made il2 compare, oleg said it was a guide, not as absolute truth.

nothing has changed since then...

the information is derived from infomod, from an FMdump.

Erkki_M
05-17-2010, 10:51 AM
Originally posted by thefruitbat:
from day one when youss first made il2 compare, oleg said it was a guide, not as absolute truth.

nothing has changed since then...

the information is derived from infomod, from an FMdump.

As did YOUSS aka GWYNBLEIDD say himself as well. He also mentioned that because the way the AI works, none of the data at speeds of 270kmph(I think) or less is any trustworthy.

M_Gunz
05-17-2010, 11:41 AM
Originally posted by AndyJWest:
MG, there is no way to use 'parts of tha AI' with the standard DeviceLink interface, as far as I'm aware. I'm not even sure how relevant AI code is to the question of measuring pilot-controlled sustained turn rates anyway - this is a FM issue.

No way? For you or me but when Oleg Maddox is helping do you think it is still so?
You seem to be very sure of a lot of things that you really should not be.

M_Gunz
05-17-2010, 11:45 AM
Originally posted by mortoma:
So much for IL2 Compare being very useful. Oh well.. http://forums.ubi.com/images/smilies/16x16_smiley-sad.gif

It's just not useful for what you want.

AndyJWest
05-17-2010, 11:56 AM
Originally posted by M_Gunz:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:
MG, there is no way to use 'parts of tha AI' with the standard DeviceLink interface, as far as I'm aware. I'm not even sure how relevant AI code is to the question of measuring pilot-controlled sustained turn rates anyway - this is a FM issue.

No way? For you or me but when Oleg Maddox is helping do you think it is still so?
You seem to be very sure of a lot of things that you really should not be. </div></BLOCKQUOTE>

If 'Oleg Maddox is helping' then does this help involve using the standard DeviceLink interface? As I understand it, IL2C used data derived from infomod (a mod, as its name suggests). Exactly how this interacts with the IL-2 FM and/or AI I don't know - I was talking about deriving turn data independently, since the question was whether IL2C was correct - Though if Oleg says it isn't, I'd be inclined to take his word for it.

What is there in any of this that I shouldn't be sure of?

M_Gunz
05-17-2010, 12:13 PM
Because there is a way and that's what was done but no it's not open to the public.


As did YOUSS aka GWYNBLEIDD say himself as well. He also mentioned that because the way the AI works, none of the data at speeds of 270kmph(I think) or less is any trustworthy.



I can't be sure of just what/how every release was done but some clues -were- given.
I don't have copies of the posts here and there made since IL2C first came out. More's the pity I guess.

AndyJWest
05-17-2010, 12:50 PM
Sorry, MG, I'm not sure what you are trying to say. There is 'a way' to do what?

BillSwagger
05-17-2010, 01:22 PM
I've been enjoying many more servers that lock out mods, which is unfortunate considering all the great things they can provide. Its just nice to be able to play a game with out all the performance issues that are sometimes brought about by mods.
This issue concerning the Ta-183 might have more to do with how the game engine interprets engine thrust.
Is Il2 Compare wrong sometimes? I actually don't think its too far off from what's capable in game, but i don't see how looking at sustained turns would tell you how tight a plane can turn. Nor looking at top speeds would tell you how fast a plane goes in a shallow dive. So its just a matter of how the information is used and what exactly you are comparing. Getting a sustained turn figure would probably take several turns at a constant speed which i don't think a human pilot is capable of accurately doing. And like someone mentioned before, the computer can make speed measurements in decimal points, where a human pilot might actually be measuring a sustained turn which has a minute amount of decay.
I could be wrong, and who's to say Il2 compare is the end all of evaluating aircraft performance in game. I usually only use the program for mission building, and even still, i have to do actual flight tests to be certain that the mission i've made is well balanced.

M_Gunz
05-17-2010, 01:57 PM
Originally posted by AndyJWest:
MG, there is no way to use 'parts of tha AI' with the standard DeviceLink interface, as far as I'm aware.

Yeah way.

AndyJWest
05-17-2010, 02:06 PM
Originally posted by M_Gunz:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by AndyJWest:
MG, there is no way to use 'parts of tha AI' with the standard DeviceLink interface, as far as I'm aware.

Yeah way. </div></BLOCKQUOTE>
Um, how?

M_Gunz
05-17-2010, 02:11 PM
If you're not going to bother reading my posts then don't ask me to repeat them.

AndyJWest
05-17-2010, 02:51 PM
MG, I wrote that there is "no way to use 'parts of tha AI' with the standard DeviceLink interface, as far as I'm aware.

This is a statement about what I know, not about what is possible. As for what you posted, I presume you mean this:

Youss might have from the start made IL2C in cooperation with Oleg and used parts of the AI.
Of course just what code mechanism the AI uses to interface with the game engine is beyond debate, but...

I think it has been established the the IL2C data is derived from infomod - a mod. This doesn't use the standard DeviceLink interface - from what I can see from the published info I have, it doesn't use DeviceLink at all. Of course, I may be wrong, but I can only work with what I know.

Back on topic, I've had a bit of an experiment to see whether my AP will be any good for measuring best-turn rates. As it stands, it has problems in the high AoA/high bank situation necessary. I may be able to adjust it a bit to get round this, though it is unlikely to ever be particularly stable or accurate - it may be able to do better than a human pilot though, so this might be worth pursuing. I'm not going to bother with this if there are better ways of making the same measurements though. Does anyone have any constructive suggestions?

na85
05-17-2010, 03:33 PM
I think it has been established the the IL2C data is derived from infomod - a mod.

Clearly that has not always been the case. IL2c has been around long before mods.

AndyJWest
05-18-2010, 06:58 PM
Well, regardless, it doesn't derive data from the standard DeviceLink now, so I can't access the methods it uses to extract data - and if I did, it still wouldn't necessarily help me verify turn rates independently, which is what I'd like to be able to do.

I've had a bit of a tinker with my AP, and as I suspected, simply increasing the maximum AP control movement range doesn't help with high-bank turn tests, at least with the Ta-183. I'll probably add an option to toggle the range limit on and off so I can try it on other aircraft, but it looks as if the problems are more fundamental and I'd have to do a major rewrite of the relevant code to correct it - if I ever could. I think I'll leave this for now, and get back to ground effect testing - that really needs more work on the AP's low-speed handling characteristics, but that is of more general use anyway.

M_Gunz
05-18-2010, 11:35 PM
Plane in high bank has a higher stall speed. Using aileron to keep bank near stall is the same problem as using
aileron to keep wings level near 1 G stall, except that the ball is quicker under Gs. Solve one and you pretty
much got the other. Try hand-flying near stall by slip and aileron alone then see what you do when you want to
hold course.

270kph at 3 Gs is as close to stall as 156kph at 1 G.

AndyJWest
05-19-2010, 12:07 AM
Originally posted by M_Gunz:
Plane in high bank has a higher stall speed. Using aileron to keep bank near stall is the same problem as using
aileron to keep wings level near 1 G stall, except that the ball is quicker under Gs. Solve one and you pretty
much got the other. Try hand-flying near stall by slip and aileron alone then see what you do when you want to
hold course.

270kph at 3 Gs is as close to stall as 156kph at 1 G.
Possibly, MG, but I'm fairly sure that most planes in a sustained best-rate turn aren't that close to a stalling AoA. In any case, if you start using rudder to control turn n such a context, you will find it difficult to maintain a consistent altitude. The 'bank angle' method of controlling turns I've been using clearly doesn't work beyond 60-70 degrees or so, and isn't that stable at lesser bank angles either. I suspect (without a great deal of evidence) that I'm running into the same problems of non-linear response that I've encountered in the low airspeed/high AoA context - the simplistic model I've used for AP control just can't handle the dynamic situation involved. When I first designed it (or more accurately, built it, then amended it so it looked designed), I wasn't expecting it to handle anything but normal cruise conditions. Controlling an aircraft at the margins of its flight envelope requires a 'fly-by-wire' approach, rather than the simplistic algorithms I'm using. In any case, I'm not sure that DeviceLink is even capable of delivering the data required at an acceptable rate.

Erkki_M
05-19-2010, 12:49 AM
Originally posted by na85:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">I think it has been established the the IL2C data is derived from infomod - a mod.

Clearly that has not always been the case. IL2c has been around long before mods. </div></BLOCKQUOTE>

Not exactly the case. IL2C appeared the same time with AFWI beta campaign, where YOUSS himself flew as GWYNBLEIDD. Near the end of it, he had over 200 sorties, during which, he had 0 disconnects, but still, he had been shot down by a fighter 0 times. All his shot-downs were due to flak, gunners, and a few collisions. He had more than 2 kills and over 5 ground kills/mission. I had the pleasure to fly together with him, us both in voice comms as well, it would take him, looking at the mission raport later, 19 seconds to shoot down 4 FW190s of the Finnish FSNS squad, this after spotting them from 7000m above, through an overcast, which through he dived.

It was somewhere around those days when Kegetys was experimenting himself as well, with stuff like gun recoil, views, engine power etc. You can still find some his videos in the youtube.

M_Gunz
05-19-2010, 01:47 AM
Erkki, wasn't IL2C first out during FB:Aces (2.x)?

Andy... near stall you use aileron to lift a wing, it does the opposite in reality and in IL2.
The aileron that goes down increases the AOA of that part of that wing which is already on the edge.
Using the rudder to maintain level or bank through slip-roll coupling is what works.

How slow can you fly in level or climb? What do you have to do? What do you have to not do?
And of course what plane, but many are similar for me at 140kph - 150kph by the speedbar.

AndyJWest
05-19-2010, 07:05 AM
MG, three points:

Firstly, as I wrote earlier, if you are trying to measure sustained turn rates, the AoA isn't likely to be that high, so adverse yaw etc shouldn't be particularly significant.

Secondly, in a perfect sustained turn, you should not have to make any control inputs at all - think about it...

Thirdly, I've said several times already that my AP was never intended to work in extreme conditions - to achieve this would probably involve more work than I'm prepared to put in (and a better PC), and might not be feasible anyway.

M_Gunz
05-19-2010, 09:05 AM
Your best sustained turn will be very close to if not right on the stall line.
It will also be your tightest sustained turn.

If IL2C shows you different then remember the caveats given with IL2C.

AndyJWest
05-19-2010, 09:21 AM
If the best sustained turn AoA is is almost at the stall, how is it possible to turn tighter in a non-sustained turn?

M_Gunz
05-19-2010, 11:00 AM
By turning at the stall at a higher speed than you can maintain that turn. You don't even have to go to stall
with enough extra speed if you don't mind a wider radius with higher turn rate but you need more than slightly
higher speed to do it.

This is assuming you are already at full power, full revs, trimmed for speed, not in slip, etc.

Riding the stall puts you at an edge of your performance envelope. You're trading all your power for angle.
If you have extra speed then you can trade that for angle. You can also trade height for angle, an option
that not everyone thinks of that pays off very well.

In a level turn if you are at your best maintained bank and you bank higher yet hold level, what do you expect?
Turning harder while slowing down? Maybe a little since "at or near stall" has the possibility of margin but in
much less than 90 degrees you will lose turn rate and speed.

Just doing the first approximation turns for a P-47 I found that best sustained is a bit over the stall. With
thrust vectoring added it pushes closer to the stall. Of course this depends on the plane so I would keep the
scope down to WWII monoplane prop fighters. For JETS you have that whole power increases with speed thing going
that I can't say just how that affects best sustained turns, where with props power stays constant while thrust
increases with lower speed at least in flight.
Getting above stall speed for any bank angle, the turn radius widens pretty quickly with circumference lengthening
at over 3x as fast. That's not just hypothesis either.

OTOH you can ask aerobatics or military pilots.

M_Gunz
05-19-2010, 11:04 AM
Rate.

IL2C probably gets the jets' turns closer than the props' turns.
Regarding all the planes in IL2C, I'm over 90% correct.

Ooops, where'd your post go K?

Kettenhunde
05-19-2010, 11:09 AM
Rate.

IL2C probably gets the jets' turns closer than the props' turns.
Regarding all the planes in IL2C, I'm over 90% correct.

That is good you defined it and much clearer what condition you are discussing.

I deleted my post as it occurred to me, I don't care and have no interest in a discussion on it.

I would say to examine the USN FTM for further insight and continue with the online educational sources.

M_Gunz
05-19-2010, 11:23 AM
Well the IL2 jets suck at low speeds but to make up for it the swept wing jets have higher stall as well.

ADD: http://forums.ubi.com/images/smilies/halo.gif http://forums.ubi.com/images/smilies/blush.gif http://forums.ubi.com/images/smilies/winky.gif http://forums.ubi.com/images/smilies/16x16_smiley-very-happy.gif http://forums.ubi.com/images/smilies/partyhat.gif that's a joke!

jermin122
05-19-2010, 09:56 PM
IL2 Compare is only credible when it comes to low altitude performance. Above 6000m its data is a mess.

In IL2C Bf-109K-4 can both fly and climb faster then P-47D Late at all altitude. But If you have fought with a P-47D Late at high altitude, you may have found that you couldn't outclimb or outrun it in almost any contemporary German fighters, same with P-51D20NT.

AndyJWest
05-19-2010, 10:19 PM
IL2 Compare is only credible when it comes to low altitude performance
Nothing I've seen in this thread suggests that IL-2 Compare ever offers anything other than an approximate guide to performance in any context. This is what Oleg Maddox seems to have stated right at the start, and nobody seems to have presented evidence to the contrary. Actually, I'd rephrase that and suggest that nobody seems to have presented much real evidence at all, one way or another. http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif

jermin122
05-19-2010, 11:17 PM
Don't overreact. What I did is just pointing that out. Nothing else. I was not saying your beloved P51 and P47 are over modeled.

AndyJWest
05-19-2010, 11:27 PM
Originally posted by jermin122:
Don't overreact. What I did is just pointing that out. Nothing else. I was not saying your beloved P51 and P47 are over modeled.

Err, what? I'm more of a Spit pilot by choice, though I'll fly more or less anything with a noisy go-forward contraption fitted. I'm equally capable of missing sitting-duck targets in anything. I'll admit to a certain fondness to the P-47, but only because I managed to bounce one off a stationary Zero and still belly-land with nothing other than a bent prop. http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif

jermin122
05-19-2010, 11:49 PM
I tend to believe the IL2C aircraft performance data is based on the performance data of their real-life counterparts. At least the performance of Fw-190As are quite identical.

Yoss once said that he collected all the data from the FM of AI aircrafts. Since it is all known that the performance of the AI aircafts is way superior than players flown aircrafts. It is very likely that the FM of the AI aircraft is more accurate than the ones we fly. This is just my guess though.

AndyJWest
05-20-2010, 12:26 AM
Originally posted by jermin122:
I tend to believe the IL2C aircraft performance data is based on the performance data of their real-life counterparts. At least the performance of Fw-190As are quite identical.

Yoss once said that he collected all the data from the FM of AI aircrafts. Since it is all known that the performance of the AI aircafts is way superior than players flown aircrafts. It is very likely that the FM of the AI aircraft is more accurate than the ones we fly. This is just my guess though.

IL2C data is supposedly derived from the sim, not 'real world' data. It may well reflect the simplified FM used by AI aircraft. I'm not sure you can call the AI FM 'more accurate' though, given its simplistic approach to engine modelling. Probably both the AI FM and the pilot-controlled FM are flawed, but in different ways. I suspect the only reason there are two different FMs in the first place is that the pilot-controlled version is so processor-intensive. Hopefully in SoW:BoB this will be less of an issue.

Erkki_M
05-20-2010, 01:08 AM
Originally posted by AndyJWest:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">IL2 Compare is only credible when it comes to low altitude performance
Nothing I've seen in this thread suggests that IL-2 Compare ever offers anything other than an approximate guide to performance in any context. This is what Oleg Maddox seems to have stated right at the start, and nobody seems to have presented evidence to the contrary. Actually, I'd rephrase that and suggest that nobody seems to have presented much real evidence at all, one way or another. http://forums.ubi.com/groupee_common/emoticons/icon_rolleyes.gif </div></BLOCKQUOTE>

And this is what YOUSS suggested himself - only use it as a tool to get general idea of relative performance.

M_Gunz, yep, but note that a few other "hardball's programs" to il2 that were before the IL2C were also done by him, for example the one with data on guns, bombs and ammo dug up - that right after AEP came out, and way, way before the mods...

JtD
05-20-2010, 02:41 AM
I've done a lot of in game testing lately and il2-compare got relative performance of similar planes right, but not the absolute performance. Also planes that differ a lot have different kinds and margins of error in relation to il2-compare data. So if you want to argue about 10 km/h or 1 m/s climb rate, you better do in game testing. That's also true for turn times.

A huge difference between il2-compare and real performance seems to be propellor efficiency, which is particularly notable with fixed pitch props, like on the early I-16's. Il2-compare seems to calculate with an optimum efficiency, while the game does not. I'd guess that some il2-compare programs have been adopted to account for it, which is why the same planes have different performances depending on who'se il2-compare you check.