Acceleration comparisons

[sniperton]

Originally Posted by MaxGunz

If during a run the plane goes up and down even 1 or 2 meters that will change the acceleration and trim state. By the time the plane has come back down it's going to look like a sudden change. Sound familiar? Guess why I quit trying to make test runs? I'm not good enough!

You're not good enough, OK. I'm not good enough, OK. My point is different: under the given circumstances mentioned in my post, is there any living human (could there be any living human) who could be good enough to make the trick? Don't misunderstand me, I don't have problems with aircraft performances, and I don't complain for this or that plane being 'porked'. I simply try to draw attention to an important factor of the game beyond maths: there's a gap between 'objective' values (which I don't dispute) and how far they are practically relevant for the player. They are 'reference values', that's OK. But we can move one step further and ask why the player (you and me) is not good enough to achieve them. My problem relates to the way we communicate with the game engine and not to the parameters it is programmed to work with.

[MaxGunz]

So get a Volvo universal joint, about 5 meters of square steel tubing, a seat, various other hardware bits and some electronics and wire then do the cutting, bending and welding and you can join the elites who have.

IRL test pilots did and do fly the tests.You and I are not test pilots and probably never could have been. We're not theoretical physicists either, or gold medal downhill skiers or heavyweight boxing champions. I for one am totally unsuited to be a playboy bunny for at matters.

It's more important for online combat to have good SA than to be able to squeeze the most from your plane. It's more important to have a good wingmate that you act in unison with. It's even more important to be a hotshot marksman than to close the last few percent in pilot skills.

IL-2 is not a test pilot game, it's a fighter pilot game.

[sniperton]

Originally Posted by MaxGunz

So get a Volvo universal joint, about 5 meters of square steel tubing, a seat, various other hardware bits and some electronics and wire then do the cutting, bending and welding and you can join the elites who have.

It's more important for online combat to have good SA than to be able to squeeze the most from your plane. It's more important to have a good wingmate that you act in unison with. It's even more important to be a hotshot marksman than to close the last few percent in pilot skills.

I basically agree, but I put the emphasis elsewhere. Our personal performance and 'feel of flight' is affected by three factors: 1) the potential flight performance of our plane as best exploited by the AI and represented in test charts; 2) our personal skills as listed by you; and 3) the ways and means we apply our skills to the game via various peripherals. This latter is very much compromised and gives a huge 'deadband' as to personal performances: no wonder that some people build custom cockpits to narrow this deadband (exactly what you suggested ironically), and remember how much advantage 6DoF gives over the hat switch. My point is that this No. 3 affects our experience much more than a few kph's change in No. 1, so that it would make much more sense to discuss the 'realism' of peripherals and displays than the 'realism' of the charts.

[JtD]

If the potential flight performance is best exploited by AI, how come I can beat them 9 out of 10? They don't even know how to properly use radiator and pitch.

[sniperton]

Originally Posted by JtD

If the potential flight performance is best exploited by AI, how come I can beat them 9 out of 10? They don't even know how to properly use radiator and pitch.

I, too, can beat the AI thank to my superior intelligence.
What I can't do is to fly and land my fav plane as smoothly as the AI does.

[MaxGunz]

Let me start with the end which is:
I'm not complaining but I am pointing out why some of the conclusions I've seen in this thread are wrong.

Originally Posted by sniperton

My point is that this No. 3 affects our experience much more than a few kph's change in No. 1, so that it would make much more sense to discuss the 'realism' of peripherals and displays than the 'realism' of the charts.

As far as the game goes that is true. But when we try and determine the performance of the game's models and start interpreting results of flaws in our procedures as flaws in the model then it's something else.

IMO this stems from fantasies about Ace pilots and becoming one in-game. At the heart of most FM-whining posts I've seen there is the assumption that the player is test-pilot good and has an absolute understanding of everything that happened in whatever event set them off, point of view and what they didn't see having no effect on their omniscience. It must have happened as, how and why they think it did.
Accounts from WWII pilots are taken as absolute truth to the tiniest detail. If the pilot said it was a Tiger tank then it was regardless of the times when USAAF pilots strafed and bombed Shermans, all the data that says NO is ignored while data that says YES or even MAYBE is taken as absolute support.

Especially in more arcade flight sims ( pretty much all sims I had before 1998 ) it was -easy- to be a top pilot and shooter too. Table driven sims got you there almost automatically. Even IL-2 which is *not* perfect now added whole not-before-included factors as of 4.0 that didn't get complimented by a different control interface method until 4.07.

I consider it a benchmark when a sim includes factors that players have to learn and get used to to even begin to get near top performance.

IRL I spent hours trying to hold a plane +/- 50 ft in steady level flight. I did manage that and note that speed changed more than a couple knots the whole time. A real pilot with more time would hold the porpoising down closer to zero and might get there or really damned close for a short time but how many can stay within 1 meter long enough to pull reliable data out while changing speed?

Now imagine anyone trying to get it exact for a whole full power run from stall to top speed and being so confident in their flying that they use data derived from that to point out flaws in the plane? The difference in gear or whether the pilot is sitting in a moving plane or a chair behind a desk goes not cover the similarity of trying to do that IRL or in game, the only difference is in chutzpah.

We've seen much worse. We've had G... and T... and The Joke who went beyond honest I-didn't-know mistakes to full blown BS creation so:

I'm not complaining but I am pointing out why some of the conclusions I've seen in this thread are wrong.

[sniperton]

Originally Posted by MaxGunz

As far as the game goes that is true. But when we try and determine the performance of the game's models and start interpreting results of flaws in our procedures as flaws in the model then it's something else.
I'm not complaining but I am pointing out why some of the conclusions I've seen in this thread are wrong.

As you might have seen, I'm not really interested in minor performance issues, which I regard secondary in importance. What I'm interested in are the reasons behind the 'flaws in our procedures' -- yours and mines. Many of our flaws, I believe, have something to do with the game interface being 'realistic' in a questionable way. E.g. the 'realistic' image of a cockpit, as we have it on our monitor, is rather unrealistic and mutilated if compared to the full visual perception one might have in that cockpit. Imagine you have to drive your car relying on the image of a single board camera. The image you get is 'realistic', but the visual experience is not. Discussing and debating such issues would be probably more useful than adjusting the charts IMHO.

[MaxGunz]

Originally Posted by sniperton

As you might have seen, I'm not really interested in minor performance issues, which I regard secondary in importance. What I'm interested in are the reasons behind the 'flaws in our procedures' -- yours and mines. Many of our flaws, I believe, have something to do with the game interface being 'realistic' in a questionable way. E.g. the 'realistic' image of a cockpit, as we have it on our monitor, is rather unrealistic and mutilated if compared to the full visual perception one might have in that cockpit. Imagine you have to drive your car relying on the image of a single board camera. The image you get is 'realistic', but the visual experience is not. Discussing and debating such issues would be probably more useful than adjusting the charts IMHO.

I think we agree though I do feel that for really good reasons like there have been in the past, adjustments are warranted. I'd love to see a P-51D or two with empty or mostly empty fuselage tank!

The other issues I am used to seeing since I first got into the flight sim community back in 1998. They are definitely good topics and the discussions got lively at times even before the wow kind of stuff available now was ever known.

Yes, we as players are quite hampered, even with head tracking. That's one reason why I always stood for search keys, not everyone has head tracking (I never did).

There is the monocular view issue with canopy spars blocking view that would not be blocked with stereo vision. When IL-2 came out and during development average players didn't have enough video power to ghost the spars which while not perfect would help simulate stereo vision. This is something that IL-2 will probably never get.

We don't get any feel sitting in stationary chairs. IRL you can feel slip or skid as a pull to the side kind of thing. IRL you know when you're pulling G's, rising or descending, tilt, G-forces and turbulence.
In sims you have only visual cues and instruments. One major sim skill is integrating those into 'feel' where changes in speed or VSI or The Ball mean something just through practice.
And yes, this is also something that used to come up for discussion. What you brought up is kind of a minefield of topics that it's probably good to bring up now and again.

I can suggest some help with the regular joystick (I have an X52) which is to add a lot of FILTER to the pitch axis in the stick sensitivity screen. I mean like 40% or more. It doesn't add much delay to stick moves, a fraction of a second, but it will really smooth your flying out.

Another thing, and this takes loads of practice, is to not let yourself rest the weight of your hand or arm on the joystick. It's something you really have to work hard at catching yourself doing, especially when things get exciting. But when you do keep a light touch it pays off well in performance. Try flying no combat with just thumb and 1 or 2 fingers on the stick for a few minutes and see. What's worse than resting arm weight is the stick ham hand death grip that we all do at some time.
With a full length stick that doesn't matter as much but still a light touch is better just as IRL.

AI takeoffs and landings are scripted. The AI can't fly the models down near stall without rolling over and spinning. What I find amazing is how well the AI does fly the models at all. More than once I had urged that the AI's should fly table-driven models as IMO that would make the AI code simpler not just flying but for AI tactical planning and would nail down what the AI could do. It wouldn't take much to have a different table for each level AI.

[horseback]

Notes on USN/USMC Fighters tested @ 5K ft.

1. The early Wildcats seem to compare with what I know of their performance, not spectacular but stable; however, the FM-2, which was 500 lbs lighter & somewhat aerodynamically cleaner than the F4F-4, as well as enjoying a 160 hp power advantage at altitudes up to nearly 20,000 ft, is portrayed in-game as even more sluggish than its predecessors. I was so surprised by this, I ended up re-running the FM-2 to ensure that I hadn't left the landing gear hanging down or something, but it was just as sluggish and 'meh' as the first time. Historically, this was simply not the case. The FM-2 was widely acknowledged as the 'wilder' Wildcat; being lighter, cleaner and more powerful at low and medium altitudes, it had superior climb and acceleration, and a somewhat better top speed at low and medium altitudes. It was a much better match for the Zero, even the later models. Someone's got some 'splaining to do.

2. There is very little if any difference between the F4U-1/Corsair Mk I and the F4U-1A, which makes little sense, given the -1A's water injection and the fact that the runs made in the -1A were all much more level overall. The Dash 1A is ultimately faster once you reach 480 kph, but it should be no contest from the start. This just doesn't seem right. I also added the Dash 1D, and it has a noticeably better jump, but the same general top end.

3. Both models of Hellcat continue to be a huge letdown. With or without water injection, it is portrayed as a slug, and much slower than the official numbers I have found. A standard F6F-3 should be capable of 290 kts/330 mph true airspeed, or about 530 kph at this altitude. If the Wonder Woman 'speedometer' is correct, that would mean an IAS of about 470/480 kph in-game. The best level TAS I got from the F6F-3 was just over 510 kph, and the best level TAS on the F6F-5 was around 515 kph, or about 460 kph indicated for both. Again, there was very little difference in acceleration between the two, in spite of the extra 200 hp or so that the water injection of the Dash 5 is supposed to have. Again, I re-ran the Dash 5 to make sure that I hadn't done something wrong. The main difference is in top speed, but from 270 to 380 IAS indicated, there is no difference.

cheers

horseback

[sniperton]

Originally Posted by MaxGunz

Yes, we as players are quite hampered, even with head tracking.

There is the monocular view issue with canopy spars blocking view that would not be blocked with stereo vision.

We don't get any feel sitting in stationary chairs. IRL you can feel slip or skid as a pull to the side kind of thing. IRL you know when you're pulling G's, rising or descending, tilt, G-forces and turbulence.

In sims you have only visual cues and instruments. One major sim skill is integrating those into 'feel' where changes in speed or VSI or The Ball mean something just through practice.

And yes, this is also something that used to come up for discussion. What you brought up is kind of a minefield of topics that it's probably good to bring up now and again.

I think it's clear that G-forces and the rest you mentioned are simply outside the scope of a sim, at least under armchair conditions. Still, they too can be transformed or translated into visual cues, something like as Wonder Woman View does (surely a bit too generously). WWV is a constructed image which makes a wide range of sensual information available in visual form, and in this sense it's more realistic than a faithful camera image which filters out all non-visual sensual input.

Now turning to the purely visual cues one might have in a RL cockpit, I feel that they too are rather compromised in the 'faithful' camera image we have. Even with a TIR you need several huge monitors to have everything relevant appear in a meaningful distance and a meaningful size (and with the use of huge monitors we have already left the realm of 'airmchair conditions', btw -- ask the wife ). The instruments are too close and the skies are too far, and the camera view with its zoom function can hardly replicate the ease the human eye and mind adjusts itself to the task IRL. My solution would be a third type of view with a split screen with fixed FOV for both the static lower instrument part and the dynamic upper combat part (with some zoom allowed when using gunsight view). Just an idea, don't kill me