Echo38

Hmm. At max power? Without having the module, with which to confirm, I can't really determine whether or not my associate was doing something wrong. If you can trim the 109 to pitch down, hands-off, at max power & max level speed, then perhaps he was having a hardware issue.

A most interesting idea! Someone sent me a video of trimming the DCS 109 full-down and taking his hands off the stick, and the thing looped in short order. I've been following both issues (although I don't have the module, the Me 109 is my favorite aircraft of all the ones currently in DCS), but it never occurred to me that the two could have the same cause--some kind of gross excess of lift? Good question!

Heh--I meant that, with practice, things which were possible will become easy, and things that were impossible will become possible ("possible," not necessarily "advisable"). If you can handle the prop effects, however, you can become airborne quicker (with less runway used) if you shove it forward briskly. This is why I try to practice takeoffs with as rapid of a throttle-up as I can make without risking the engine.

Hmm. I always advance it as quickly as I can without risking the engine--so, smoothly over about two seconds. Lately I've been even holding it with the brakes, although if you don't let go when it reaches a certain amount of power, it'll overcome them and pivot left. I don't doubt that it's more difficult to do it this way, but it's still doable--with a full rear tank, even. (Both take-off assistance and auto-rudder off, of course.) As others said, practice!

Because of this, IIRC, it was more or less the standard round for the Browning .50 [nods]

Did you read post #3 in this thread? There he already addressed that point. Is he incorrect? If so, how?

I wish every flight simmer had at least a few hours flying a real aircraft--even just a Cessna. There's so much misunderstanding about trim amongst simmers; I've been seeing it for years. I can't blame people, though; without having felt it oneself, it's difficult to understand how it works IRL. Lemme try to put it simply: In an accurately-modelled P.C. flight sim, during flight, if you're holding your physical gaming joystick constant, and you alter the trim in the sim, your virtual stick moves. In a real aircraft, during flight, if you're holding the stick constant, and you alter the trim in the sim, your stick doesn't move. It just increases or decreases the pull on your arm(s). The stick will only move if you allow it to move (as in "letting go" or relaxing your arms). There's no way of accurately simulating this in a P.C. flight sim. Force-feedback joysticks are the closest thing, but they don't have strong enough motors to portray it accurately under all conditions. This phenomenon--the inherent discrepancy between RL (with one analogue stick) and a PC flightsim (with one physical stick digitally commanding a separate virtual stick)--is responsible for much (or most) of the pilot-induced oscillation, and other control issues, which we simmers experience.

Addendum: this is starting to pique, so I went and drew lines through that dubious comparison image, and superimposed a protractor over it. The difference is almost exactly one degree (seven-degree dihedral versus eight). So, yes, we're talking about a difference on a clock face of two minutes (as in time), barring errors in perspective, and assuming that the sketch from that book is accurate. I've updated my previous post with the actual numbers. Note: this one-degree apparent difference between the two superimposed images, is not necessarily a one-degree error, for reasons previously discussed (e.g. perspective). No error in the 3D model's dihedral has yet been ascertained--which, I repeat, doesn't necessarily mean that there is none. Also--for the record--I would consider such a one-degree error in dihedral to be a large one, given that it'd be ~1/7 or 1/8 of the total dihedral.

In honor of the love we all share for fighters & flightsims, I made a little comic strip to represent the discussion of aircraft selection.

As I've said before, we should have gotten a 109G instead of a K. However, continually trying to "one-up" the "other side" is only going to make the problem worse. Given that we have 109K and 190D, we should be hoping for good matches for those, and not trying to get something that outclasses them. Your objectivity is in doubt. I'm getting the sense that some have the attitude of "well the 109K is too much for the P-51D, so we need to 'get back at them' by bringing in a F4U-4 so we can have something that's too much for the 109K!" This is childish and solves nothing, but only creates more problems. "Two wrongs" and all that. I am, and long have been, in favor of the WEP ratings of the P-51D & P-47D being increased as appropriate, to historically-common boost levels, with the end result of trying make a fair match for the Me 109K & FW 190D, without decreasing historical accuracy. Quit trying to persuade ED to pick a variant that outclasses your opponent aircraft, people! This goes for both sides. German fighters, American fighters, whatever. Knock it off, for love of blue. It's going to kill multiplayer; for your own sakes, even if you lack objectivity, you should listen to me. Unless you like flying on empty servers ...

I think you are overly confident in your ability (and that of others) to judge by visual estimation. I can not "see it clearly," and it isn't my eyesight that's the problem. It does seem that there might be more dihedral in the photos, after you mentioned it, but this feeling could be fully accounted for by optical illusions & the power of suggestion. If there is an error, it isn't large enough that we are able to clearly see it without some of the "hard evidence" I specified. If there is an error, we are talking about approximately one degree off, which is ~1% of a 90-degree angle. That isn't something that most people can "clearly see" without a side-by-side comparison (and, of course, I mean at the same perspective et al.). To put it into perspective (so to speak): we aren't talking about being able to tell between the difference between 2 o' clock and 3 o' clock (or even between 2:30 and 2:45), but rather the difference between ~2:44 and ~2:46--with the two clocks at different distances and angles, and without any markings on the clock faces. ("Minutes" as in "minutes of time, on a clock face," not minutes of arc.) Those are the actual numbers, using the comparison image you posted, and a superimposed protractor. There's no way that can sanely be described as a clear difference. This, exactly. I'm not being an Eagle Dynamics apologist; I acknowledge that the accuracy of the model is in some doubt. But, as I've said repeatedly, there isn't yet enough evidence to jump to a conclusion, the way some have done (in both directions, I might add!). Assumption was the cause of any mistake there may be; making another assumption isn't the way to correct it. I hope that someone can come up with some reliably-sourced diagrams, or at least some exactly front-angle photographs, so that the issue can be settled, because, currently, the question has not been decided one way or the other. If the model is wrong, I would like to know, and to see it corrected someday.

Yes, K4 & D9 are both what I'd call end-of-war tech, but not quite as much of a technological leap from their predecessors as the -4 was. I don't think* they'd be an even match for the -4 under usual circumstances. Hence my objection. * I'm not an expert on the -4 by any stretch, but my understanding is that the thing was a monster, in a way that none of the other typical late-war birds (109K, 190D, P-38L, late-block P-51D & P-47D, etc.) quite were. ~2500 hp.? Weighs about the same as the P-51D, but with a better wing for dogfighting and an extra ~700 hp? I don't think there'd be any contest between that and our 190D, or even our 109K. I don't believe it would be a good model to put in, when there were other models which were both more commonly used, and also a more fair match for the opponent fighters. We shouldn't be asking for "superplanes" to be thrown into the sim, especially when they were less common than more down-to-earth (and more even with the opposition) models/blocks/etc. It's bad enough that we have a 109K instead of a 109G; we don't need to "up the ante" even further by asking for a F4U-4 instead of one of the classic F4U-1 variants, see? If a -1 was the more commonly-used fighter, and also the more even match for the opponents it'll be facing in the sim, then it seems a no-brainer to pick that instead of the -4, which was both less common and less of an even match. Unless one is trying to rig the matches ... We should be asking for classic matchups, close matchups, not asking for the most crazy-powerful variants with which to make lopsided online battles. This isn't the actual 1945 arms race; there's no call for trying to rig the game, in the sim, by urging the developers to choose the most over-powered variants of favored fighters (to be pitted against less powerful ones). That isn't good for the health of the sim & its community. It makes for a planeset which is neither historically-representative, nor competitively-balanced.

Ah, they said that about the real one, too. ; )

Isn't the -4 a bit overkill for even the 109K-4 and FW 190D-9? As I understand it, the F4U-4 was faster and also more maneuverable, in general, being end-of-war tech. Not as good of a match as some of the earlier--and more usual--Corsairs, unless I am mistaken about the performance.

Mm, yes, that's sort of related to the other issue I was talking about earlier, of trying to match up all four variables (X, Y, distance, FoV). Here's a pair of screenshots I took in a different flight sim/game a long time ago; the two shots were taken at the same time (game paused), from the same deflection angle, but the distance & FoV were different from the first shot to the second. This effect is easily reproducible in DCS and any other sim that has variable zoom/FoV and camera distance, as well as in real life (if you have a variable-zoom camera with a wide enough max FoV). It gets exponentially more difficult to line up a pair of shots when you add the other two dimensions (X & Y camera deflection) to these two (FoV & distance), as the effect is not merely additive, but rather multiplicative. In the case of my screenshot below, it was fairly easy to match up the wingspans, but when you move the camera to the left or right, it makes the wingspan look smaller or larger (in addition to both FoV and distance being independently able to make the wingspan appear to change size on the image). All these reasons (those posted by me and others in this thread) are why I am--although not convinced that the model is free of errors--skeptical of an amateur superimposition between a screenshot and a lone sketch from an unverified source. It isn't as simple as some are making it out to be, and "eyeballing it" is not okay. Any errors (either on the part of the 3D modeller, the 2D sketcher, or even those making the claim that the model is incorrect--or on the part of several of these, or others) are a result of someone trying to eyeball it!

Technically, this isn't an accurate statement; inability to prove the model incorrect doesn't necessarily mean it's correct. More generally: { unable to disprove A } /= { A proven true } and also: { unable to prove A } /= { A proven false } Hypothetical example: I claim the moon is made of rock instead of cheese, but fail to make a convincing argument or offer evidence. While those I'm trying to convince (assuming they don't know the moon is rock) are right to doubt me, my failed argument & lack of provided evidence doesn't actually mean that the moon isn't made of rock, or that no such evidence exists. It merely means, in this case, that I suck at proving it, and/or do not have a rational reason to believe my own claim. But it doesn't actually mean that the claim is wrong; in this case, it happens to be correct, as the moon is, in fact, made of rock and not cheese. (Right? Right, guys? [looks around] ...) Silly example, serious point; see? Now, as to the FW 190D-9 dihedral: there has actually been more evidence provided in this thread indicating that there is an error, than evidence indicating that there is no error. In fact, no actual evidence has been provided in this thread to indicate that there is no error. However, burden of proof does appear to be more on those claiming that there is an error, and what little evidence has been provided so far is not convincing. One hand-sketched picture from a book I don't recognize, and a few random photographs from inconsistent angles etc., isn't going to cut the idiomatic mustard. I remain open to being convinced of either possibility. (Not that I'm the one whom anyone should be particularly keen on convincing, given that I'm not one of the testers et al.) It is a legitimate question, and the lack of conclusive proof so far is not grounds to assume either way.

I wonder how much wing flex comes into play. Given the location of the landing gear, on the ground probably isn't much of a difference compared to 1G flight, but ...

The discrepancy in dihedral angles between the two superimposed images posted earlier is somewhere between 1% and 2% of a 90-degree angle, as best as I can tell using Windows Paint pixel-counting instead of a compass. This is indeed a significant (I'd say "large") discrepancy, given that the difference is something like 25% of the dihedral angle itself; however, when we're talking about trying to eyeball it, ~1.5% of 90 degrees is a small enough difference that the reference picture does need to be "perfect," or close to it, because most humans can't accurately eyeball a ~1.5-degree change. With my screenshot attempts, my own X&Y&dist&FoV errors are larger than 2% (cumulatively and even individually), and that alone rules out my screenshot attempts. You can't use a reference image that has larger errors than the margin of discrepancy between the other two image, see? This is the problem with eyeballing it. I wish 3D modellers and 2D sketchers alike would stop doing it. Accurate diagrams are needed, and must be superimposed for an exact match. In manufacturing, thousandths of an inch are used for fine measurements, not "finger's widths."

Damn; I can't get an angle anywhere close to right, in the sim, without knowing the FoV of the camera in the original photograph. I'm always at the wrong X angle and/or Y angle and/or distance and/or FoV; adjusting the camera deflection angles et al. to fix one thing looking out of place just makes something else fall out of place. Inconclusive--might be able to do it for a 100% head-on shot, because then only two degrees of variability (distance & FoV), but right now I'm trying to do it with four degrees of variability, and I just can't do it, not without a reference point (e.g. the same background in the sim-world as was in the original photograph, or distance & FoV of the camera in the original photo).

I wonder: Is the 3D model for the FW 190 in DCS the same for people who don't own the module, as it is for people to own the module? I'm going to fire up the sim and try to get a screenshot from the same angle as photographs, but I don't own the module, so I'll need someone who owns the module to try to match the same angle afterward, just to be sure that downloading & installing the module doesn't add a new model.

Same deal; sorry, mate. I'm not trying to be difficult; I'm not trying to blindly defend the sim. But that's a plastic scale model site. I used to own a plastic scale model P-40 that was hilariously wrong in its dimensions. P-38, too. I'm gonna need something more official to settle my mind one way or other; right now, it is in question whether the DCS model is in error.

Well, some of the P-51 textures in DCS do seem to have incorrect rivet lines (I haven't actually counted the rivets, ha-ha, but even the lines appear to be in the wrong places, assuming my reference model was using original lines--I assume all individual P-51D's originally had the same rivet lines--I don't know much about the manufacturing process), so I wouldn't be surprised if you're right. Still, I can't regard a hand-drawn third-party sketch as convincing, by itself.

Again, I've known eyeballed 3D models to be horrifically wrong before. Operation Flashpoint comes to mind; some of the dimensional errors were as high as ~20%. IL-2, also--the dihedral for the P-40 was horribly wrong, by a much larger margin than the discrepancy between the two superimposed pictures posted in this thread. Eyeballed drawings and eyeballed 3D models are both susceptible to this; the question is, which one was eyeballed in this case? As I said, I have no reason to assume that the sim is correct--if it's wrong, it needs fixing--however, I am reluctant to see a single drawing from a third-party source used as claim that it's "obviously wrong." Someone posts an original, official technical drawing, then I'm convinced. A hand-drawn sketch from a non-official book I don't know? Not so much. Not to disrespect the author of the book, mind you, or detract from the validity of third-party sources; Warren Bodie's book has a foreword from Kelly Johnson, with a high recommendation--but then, Johnson also mentioned in the same piece that Bodie was the only author who ever contacted him before writing a book about his airplane. Other books had factual errors which didn't match Bodie's better-sourced info, see.

Is it not possible that the sketch in the book is wrong, instead? It seems to me that seeking more sources might be wise, to confirm that it is the 3D model that is in error and not the sketch. Generally, I would expect a flight sim to be more prone to error than a well-researched book, yes; however, I have known books to be quite wrong regarding this sort of thing, and some artists just try to eyeball the thing they're drawing. Just an observation--I don't have any reason to assume that this is the case here, but I raise the possibility.

Does this sound like the same sort of problem? http://forums.eagle.ru/showthread.php?t=142856