my profile | register | search | faq | forum home
  next oldest topic   next newest topic
»  COMBATSIM.COM Forum Archive   » Game Discussions (Genre)   » Jets   » Reflections/lift lines for spatial awareness

   
Author Topic: Reflections/lift lines for spatial awareness
Attila
Member
Member # 853

posted 12-16-1999 05:17 PM     Profile for Attila     Send New Private Message   Edit/Delete Post   Reply With Quote
Hi all,

I have very little experience with reflections/lift lines from EF2000 and JSF and F2. Although somewhat helpful they still leave some confusion about head/aircraft attitude in padlock.

Leaving the discussion about reflections for visual effect etc. to the side, and focussing on reflections and canopy markings/lift lines for increased spatial awareness (understanding attitude of camera/pilot's head and aircraft) --- can someone answer/debate the following question: Does reflections/lift lines help your awareness of whether your're looking at 2 o'clock or 4 o'clock? At 30 degrees pitch or 60 degrees pitch (of camera/head)? And does it help your awareness of the tilt of the camera/pilot's head?

I suspect that innovatative canopy markings could help with these awareness problems. The pilots here clearly state that this loss of awareness is a sim problem --- perhaps we are justified in more awareness aids?

Eg. why not add a 3-9 line? Add some symbols/lines that will clarify tilt and pitch as well?

(See my F2 images on the Flanker 2 forum, "External view --- proposal", for an illustration of the problem, particularly tilt.)


Posts: 579 | From: Scotland | Registered: Nov 1999  |  IP: Logged
Andy Bush
Member
Member # 12

posted 12-16-1999 05:48 PM     Profile for Andy Bush   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
Attila

First...a shameless plug. My article on padlocks (with the focus on the USAF and Flanker 2 padlock views) should be posted on SimHQ's Air Combat Corner by Monday or so. Much of what you mention is discussed in this article.

On to your specifics. I like your interest in the subject. The result, I hope, will be a better grasp of the padlock concept by our readers...and maybe, a little attention from the developers too!!

I'll begin the discussion with an opinion. We all are familiar with the tendency of padlocks to 'flip-flop' at times. This often seems to happen when the target passes the pilot's 3/9 line. The result is the horizon line doing strange things in the view background...hand in hand with this is the 'lift line' behaving in a similar fashion. The problem is significant. In order for the pilot to avoid disorientation, he must ignore the background horizon and place his attention on the padlock mini-HUD attitude reference (assuming there is one).

It doesn't look this way in real life. The real world pilot uses the horizon to plan his BFM. The position of the horizon is a major input into out-of-plane maneuvering (meaning the beneficial effect that radial g has on turn performance).

Therefore, my opinion is that the padlock view should provide a stable horizon reference to the pilot, and the horizon should be displayed in the background as it would normally appear to a real world pilot in the normal seating position. This aspect (seating position versus head position) is what I think is so insightful about your comments).

Andy


Posts: 595 | From: St Louis, Mo | Registered: Sep 1999  |  IP: Logged
Kurt Plummer
Member
Member # 358

posted 12-16-1999 06:09 PM     Profile for Kurt Plummer   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
Hey Attila,

If you want to have a /useful/ differentiation between novice and 'topgun' modes, just ditch the funky arrows (another LOS blocker and general Exorcist spinninghead 'disorientation device') and the 3-9 and the airframe references (distorted) and the 'biomechanically accurate' (if you're an ostrich) head rotation too.

And simply put up, in the corners, or on the sides or top and bottom (you get to /choose/! ) a number indicating the clock value and another indicating the elevation value with NORMAL pan and track giving you _stable_ horizon rates into lead or lag.

Another big help would be finer representations of the tracked object. We're to the point where rez SHOULD be supporting a small-but-distinct silouhette size that would help remove the "why /is he/ moving so damn slow!" problem, that is 90% a function of baby-toyblock sized targets in a 'Leggo' (My Elbow) world...

The other things that deserve experimentation are the use of across-the-pilot views, lowered sills and global tracking 'SA memory'.

Christ I hope the remaining makers pull their heads out before we lose 'fidelity' Combat FSims altogether. I'd hate to gain currency on another Microsoft 747.


Kurt Plummer


Posts: 672 | From: | Registered: Sep 1999  |  IP: Logged
Attila
Member
Member # 853

posted 12-16-1999 09:24 PM     Profile for Attila     Send New Private Message   Edit/Delete Post   Reply With Quote
Kurt and Andy,

Let's simplify things. Lets assume that the image we see on the monitor is from a camera in the cockpit that swivels freely in pitch and yaw (no flipping), but is locked in roll (not tilted). Ignore aircraft and cockpit/canopy rendering as well for the moment. Now automatic padlock or manual panning points the camera in any direction, and the aircraft can be in any attitude (pitch and roll). How will things look on the monitor?

Imagine flying level in a 45 degree left bank. With no tilt of the camera you will have the the following pictures on your monitor depending on the clock bearing of the camera:

12 o'clock/forward: 45 degree tilted horizon from top left to bottom right on the monitor. Viewpoint at horizon.

3 o'clock: Flat horizon, viewpoint above horizon (towards the sky).

6 o'clock: 45 degree tilted horizon from bottom left to top right on the monitor (opposite to forward view). Viewpoint at horizon.

9 o'clock: Flat horizon, viewpoint below horizon (towards the ground).

My guess is that it is impossible to deduce the attitude of the camera and aircraft just by analyzing the horizon/picture --- even with no camera tilt. You have to know one to deduce the other (it's like one of those equations with two unknowns :-). Add camera tilt and you have three variables.

I'll put up the image I used in another thread to illustrate:

OK, the image illustrate the "equation" that we need to solve instantly in flight. What do we know? By processing the mini-HUD we understand the aircraft attitude (not the visible aircraft but the one we are looking from). By training we now instantly know we are looking in the rear left quarter around 8 o'clock assuming no or very little camera tilt. Why? The tilt of the horizon is opposite what we read on the mini-HUD, hence the camera is pointing to the rear hemisphere. The mini-HUD position, emphasised by LOS towards the ground, tells us the camera is pointing in the left hemisphere.

Remove the mini-HUD and/or add severe tilt to the camera, and there's no way to know.

How can canopy markings help? As an indicator of the camera attitude. Add a lift line, and you have a line of reference --- though only with high camera pitch angle (center the lift line and I'm looking up and out of the canopy). Add a 3-9 line and you would have clear indication of clock bearing, differentiating between the forward hemisphere and rear hemisphere. Add horizontal lines (circles around the canopy) and you would instantly see the tilt of the camera.

Now all variables are solvable --- even without a mini-HUD and with severe camera tilt! BUT --- assuming you can distinguish clock bearings. Forward and rear are simple to distinguish in most sims (HUD vs seat), but 3 o'clock and 9 o'clock are worse. I picked up on Kurt's suggestion for a bearing indicator --- why not just print the digits on the canopy? Probably only 3 and 9 are needed.

Add interior cockpit & HUD rendering, thus fully clarifying forwards line-of-sight (LOS), downwards LOS and over-the-shoulder LOS (part of wing might be visible as well for the latter). The canopy markings covers the rest of the viewing field upwards, and around 3 and 9. All these are implicit tools implemented by environmental modelling --- in my view very important for intuitive spatial awareness.

Then add explicit aids like the mini-HUD that (although mechanically) further helps orientation and maneuvering. Kurt's camera bearing & pitch indicator is another such explicit tool.

[BTW. A toggle to an outside view would explain things in an instant, if via a smooth transition theres no doubt (a cut-scene can be confused with a jump to a different aircraft, eg. your wingman) --- but I deliberately avoid talking about outside views here, keeping a focussed discussion about in-cockpit viewing and spatial awareness.]

--

Regarding "stable horizon" --- I can't fully grasp what you both mean by that --- particularly since I assume the horizon tilt must be dependent on camera bearing.

Andy, regarding your article --- looking forward to it! (I've dropped in and looked for it a few times already. :-)

Kurt --- I think I'm in sync with you on the "across-the-pilot" view. Did you read my thread on the Flanker 2 formum "External view --- proposal"? I envisage a toggle bringing you out of the pilots head backwards along LOS to an SA view. Here's an example (clarifying the situation in the image above):

[This message has been edited by Attila (edited 12-16-1999).]


Posts: 579 | From: Scotland | Registered: Nov 1999  |  IP: Logged
Kurt Plummer
Member
Member # 358

posted 12-17-1999 11:13 AM     Profile for Kurt Plummer   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
Hey Attila,

You got it in one!

To me one of the BIG advantages of the '3rd person marionette' viewing perspective is that it can use REAL (aircraft icon) graphic data rather than some idiocy of cockpit 'art' equivalent.

However good the latter are getting they typically distort or simply block-out the needed relative positioning airframe spacing and AOT positioning while they also aren't as 'pretty' (insert screencapture here! either.

IMO, such are also a realm suitable for 'experimentatal modification', to the players perspective tastes, which I always find to be something worthy of praise in a manufacturer going the extra mile.

Namely, (centered forward view) do you 'look best' from over the pilots left shoulder or right or right over the top of the seat headbox?

It /will/ vary, by player preference and dominant eye I bet.

OTOH, IMO, you will need to be careful how far back you go because when you look back forward you need to do it without much 'zoom' (disorienting) effect to get back inside to a HUD-line that is pretty much flat to the horizon.

If you stay outside all the time, you're getting into 'top down' perspective slant effects which can mask your look-down cues (AGL at low level for instance) and ruin your horizonline distance 'projection' of targets.

So stay inside or at least 'close' to the cockpit/ESeat and keep the features tiny enough that you are essentially looking around the nose as much as over or through it.

All's you need to do then is convince people that an /external/ HUD is a 'good thing', and IMO, you've got it made.

Now the noroll-stabilized camera and canopy line bit bothers me a little.

First, the BEST way to tell a man he's looking high or low or level is to TELL HIM.
Don't make him move his head or his eyes to look at 'minihuds'. In real life, looking down at corners induces the leans and in sims it at least distracts you which is lethal.

Rather, again give the player a choice on whither he wants DATA put around the target equating with:

"I'm looking at such o'clock"

and

"So elevation"

like

4
Target
Hi

or

4Hi
Target

or

4TargetHi

Data is constant and so can be portrayed without much worry over masking or rotation or sizing worries, right near the target.

SIZE the target within a given diamond or box surround that equates to a nm-distance value 'range' figure for the estimated target acquisition distance and subsequent 'refined' closure. This makes it possible to keep the Titanic In Flight temptation at a minimum without alienating old/poor eyes.

It also restricts the Data at a fixed perimeter distance away from any giveb dot or silouhette shape perse.

It's main use is of course as a passive ranger in sighting of non-radar targets and in guestimating the turn in spacing in the 2-6Kfoot 'close' arena.

Small box is an XXnm range distant target, a LARGE box is a close one. Give numeric manual equivalencies of course but keep things fairly loose, some days you'll see the bear and some days he'll see you...

While experience and youth and aspect deflection and a good initial radar 'guesstimate' on where to look also helps.

Now horizon lines... To me, you get fully wrapped up in a maxrate roll and 'tracking' the horizon line with another line on the canopy gets to be a 'mind altering experience'.

Which are you trying to stop motion/orient
motion on? The line of the horizon or the canopy line on the horizon?

Especially because you can /bet/ the threat isn't going to 'make it easy' on independent move complication of your tracking, as much as energy allows, he going to be going up and down like a porpoise in heat.

It's better to just 'know' (data) where you're looking at and make your moves match the bandit and the natural cues. IMO.

The same goes for 3/9. When comparing nosepoint to make your move into him, the littlest point-deflection will change clock numbers and lead/lag in a heartbeat.

You can't afford to be guessing and orientating relative to a 3-9 that may be much further forward or aft and at some weird relative (bank) 'height' (plane above or plane below) _the target_ where you're eyes oughta be 100% focussed anyway.

PRECISE maneuver aggression is a function of knowing where you started (relative airspeeds and altitude) and /from the clock progression/ 'where' your opponent is at NOW, in terms of residual energy and lift after his nose behavior.

You can't know time plus rate if you don't have an immediate idea how fast the angles and space are changing, relative (hands all crossed up, the Rolex a smoking ruin... .

Now for the roll-stabs. It depends on how you do it. If you're /tracking/ a target, of course not. You've gotta have one eye on locked on his pitot tube and the other covering his boards/nozzles and his stabs, _at all times_.

OTOH, if you're doing a snap-look or a scanlock check (see my other posts) you want to see the /activity/ not track a specific aircraft.

There's something called a 'SODAR Loop' that used to be taught. It's stands for 'See (acquire) Observe (study) Decide (which is juiciest or has the biggest teeth?) Act (initiate defense or attack) React (modify your play to his response).

From a neutral start on numbers and nose point, he who stays at least one complete cycle (four moves) 'ahead' wins. Almost all the time.

Yet the most important element, SO, the element which -assures- this 'neutrality' of first positioning, is /totally inadequate/ in sim rendering today.

Why? Because we have 'padlock from radar' on Star Destroyer sized targets and -truly stupid- opfor AI that have apparently never even /heard/ of bilevel splits or severe offangle intercept conversions outside the primary cone to stress the merge sensor volume control.

Given that it's not lose sight but /fail to sight/ that really kills you, the only way to get SO back in the loop is to make BIG scans of the surrounding sky and tighten up the radar antenna index and bar/scanwidth volume controls.

Normally your scan is from a level position but whether as a result of late maneuvering to entrance (new bandits) or a position change (after a TTurn or belly check for instance) you just can't see much if you're _always looking at dirt_.

In a developed circle fight, a 'level' (look across, stabilized) horizon is still important because the man at the top rules everybody under him and the unseen hawk is gonna be looking for a hotside shot and a quick escape out the other side.

Unless his name is SAM he's NOT likely going to be doing this by coming straight up from the ground!

But he could very well be coming from a large spacing or altitude separation, 'outside' your seen-enemy-here expectations on the elbow of dueling radius.

Again, looking _across_ the horizon (perhaps mildly fishbowled due to 'canopy distortion', snicker lets you scan up as well as down, to LOS range. On ALL TARGETS.

With my sytem, the scan puts up a 'memorized' SA globe (mimicks the minds visual short term memory) display of these tracks in the spherical volume of airspace that your eyes have just scanned.

(Pad)'lockon' is a function of ## key clicking the memory-track of top interest.

At which point, certainly, you get the full elevation (roll indexed) tracking headmovement going.

Aggressor candidates out at Nellis used to be flown to death (3-5 missions a day) until the top few Darwined the others by being able to fly a constant speed and bank angle and AOA with their heads cranked across their opposite shoulders for _two_ windup and maxrate reversed turn circles.

They then had to be able to translate this into a precise on/off lead or lag play based on an intimate knowledge of the /other guys/ airspeed and bank and AOA. Again from the opposite 'side' (banked turn startpoint) of the fight. Knowing (calling) where they were going to be rolling out for guns, missiles, snap or tracking.

They would then stand before a board and with the flightdata blanked out on the TACTS replay have to 'do the debrief' to within something like ten knots and two units and _1 degree_ rate of turn on both opponents.

Without ever having seen their instruments because they were outside the cockpit the whole time.

It was a stone /bitch/ to do and about 70% of the initial 'survivors' got whacked here.

But it was necessary because you're trying to become a teacher and you can't teach if you don't know how everything happened, impartially and without any question at all of your accuracy.

No sim pilot is ever going to be that good.
A combination of things that can be summed up as 'motivation' ('flight hours', visual acuity/reference points and seat-of-pants acceleration sense) is just not there.

But my system at least lets you 'pretend real good' by having a taste of the skyscan discipline needed to SEE the target coming and OBSERVE (where's everybody at, relative to me, in the SA globe) which is the greater threat.

And then, with your ideas, we orient _On The Target_ (top/bottom wing axis='horizontal', inlane/out/shot axis='vertical') and with minimal clutter as everything becomes purely focused-visual on his position through the canopy arc and to some extent through your hand on the throttle.

Other stuff is subliminal based on (real) horizon lines and time after first-turn gut feeling. It may be 'out the corner of your eye' but it is NOT in the screen corners.

WHEW! Okay so let me say again, I think you've got it going but don't get caught in the conventional mindset. Andy, be gentle with me, it's been a long damn day...


Kurt Plummer


P.S. If I had to do add something it would be these TWO:

1. SRM Swirl Chart. Nose to Nose, it's basically pie chart with an inner and outer RMin/Max boundary and fixed diagonals from say 20-40 and 60-100 on each side dominated by Fuzing and Lookup and in somecases Bore requirements.

But you crank in some turn and everything goes to hell. Because the 'windows' that you can shoot through turn with or against (aspect) the direction of turn.

MANY SRM shots are missed, not because the game 'models them poorly' but simply because there is no way to teach the player wuzzup for the shot requirements. No two shots are the same close in so there's no real curve except reading the (inaccessible to us mortals) books.

But if you overlay this swirl chart /then/ you get real time shot calcs. However the numbers are generated at least they would be reliable (training or novice mode) aids.

2. Elbow and Anchor. These I'm weaker on but imagine the fight having two critical 'entry' points. One is the distance /ahead of/ the nose at which you make first-break into the enemy with intent to kill your opponent. This is the aggressor/anchor point.

Now add the further distance at which the enemy starts to respond (he doesn't have to accept the nosepoint, especially with a helmet weapon it may not even be wise). This is the total-elbow.

The first sets the size (nm separation) of the circles and hence the spacing at which you're going to fight (and the weapon type you may have to first-defeat).

The other is the /timing point/ within the circles that represents the likeliest 'hinge' of lead at which current maneuver capability (Ps) will give either you or the enemy the shot IF you do 'xx' (max the lead, slip to lag, go pure).

So. Your looking at your target and up pops this AE (Anchor Elbow) training circle display of say 10nm initial size equivalency.
Much like a missile reticle and fairly large to allow for 'developments'.

Within this circle are two radial lines representing your current aspected tracks, presumeably heading towards each other.

Doesn't matter which 'real' compass direction your coming in at, its the /orientation between the lines/ and how far/fast you progress along them that determines things.

When you both break-in, The distance ahead of each vector and the direction of turn before they cross determines where the anchor is set and where subsequently the circlewar will start.

He who sets the circle will 'snap size' it to say a 2nm guns progression or a 4nm SRM progression (which is a good reason to come in controlled if you've got the lift to support a 'slow' point with ALASCA weapons).

If you say blow by each other's 3/9 (or otherwise break obliquely aft) then the circle doubles (concentric) and EACH SIDES 360 (or whereever the nose-back-to-initial-
on) point shows up as little dual clockpoint sliders around the rim.

He who has the tighter turn will have his circle inside the others while him as has the best energy:rate exchange will see his nose-on marker come at an 'earlier time' (clock orientation again) than the other guy.
Depending on who has what missile and how they /play the fight from there/ decides which is the most important factor (know your bleed rates and don't lose bore, flyout or fuzing limits).

Him as is losing both had better make some radical adjustments in geometry or energy rate (unload and extend on the far-opposite side for instance).

Now if you both break /in the lead/ then the fight goes single circle and here the elbow becomes the dominant factor.

This time YOU are at the presumed center of the circle and the size of two radial lines to the outer rim is -directly related- to the spacing between you. Smaller the circle, the less the space.

The other dominant factor is the weapons selected because these select what the two radius lines /mean/.

Now the first line (colored red) is the bad guy lead off your nose (AON). The second line (green) is your tail line (AOT). In between the two lines are marked zones for lead (pull your nose ahead of his), pure (hold nose positioning) and lag (your nose behind his) pursuit which represents the amount of G/fps your putting into the fight.

In the single battle, everything is about the avoidance of 'collision' (spatial overshoot) whilst maintaining 'position' (weapons-on lead) for the finite /time/ it takes to get the shot.

So if you're pulling sufficient G to put the man in the 'LEAD' position then your tailline should be moving towards a perfect (180`) opposition from your nose line and you should blast him as you saddle up.

OTOH, you should also be closing the range on the enemy by cutting across the circle and if he can maintain the closure, eventually the radius will get smaller and the distance that his leadline has to travel will get shorter and he will squirt out the other side like a greased grape.

If you go Pure then you should be stabilized and the circle should be fairly even sized while you wear down his energy.

If you pull Lag, you're looking to /gain/ some space and you'd better hope it's not because you haven't the speed to match his nose lead. If the AOT and AON merge and trade places then HE is the threat.

NOTE, this is NOT a 'lead computing gunsight, times ten'. Because YOU are the one doing the computing (strategizing) and it's all based on KNOWING where your at, relative to his nose (again, why I want specific clock plus elevation measurement).

What the AE circle does is keep up your awareness of space and time as a function of your (invisible) nosepoint and non-stereo closure.

The 'elbow' (optimally 90` opposition of nose and tail at initial break) is adjusted SOLELY based on what you see the target doing and the weapons you've got to employ.

_You_ then decide to take offbore, headon or tailchase shots, as needed.

I guess what I'm trying to say is that the numero uno of dogfight errors I see people make in combat fsims is the belief that if they pull the joystick through the stops 'just a little harder', they will win.

It ain't so joe because you've got to manage closure as well as angle off and if you only succeed in blowing by the guy, you've wasted all that effort and airspeed on /nothing/.

Dunno if it will work exactly like I hope or where exactly I would put it on screen (might clutter things to have it overlay the target) but again, I think it could be a good training aid at least...


Okay, now I am done, thankfully, my fingers are killing me. KP


Posts: 672 | From: | Registered: Sep 1999  |  IP: Logged
Andy Bush
Member
Member # 12

posted 12-17-1999 08:53 PM     Profile for Andy Bush   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
Folks...this doesn't happen very often, but this time I really am at a loss for words...

Where's the aspirin...


Posts: 595 | From: St Louis, Mo | Registered: Sep 1999  |  IP: Logged
Spiff
Member
Member # 222

posted 12-17-1999 09:11 PM     Profile for Spiff   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
2,792 words. Wow.
Posts: 313 | From: Texas | Registered: Sep 1999  |  IP: Logged
Attila
Member
Member # 853

posted 12-18-1999 05:05 PM     Profile for Attila     Send New Private Message   Edit/Delete Post   Reply With Quote
He he --- I think a reference to the chinese proverb "a picture says more than a 1000 words" is appropriate. Kurt would still need two or three pictures, though... :-)

Kurt, you are clearly debating much more than the topic title of this thread suggests. Still, it would be interesting to see some illustrations of your ideas for viewing and explicit SA aids (your scanlock/SA-sphere?). (May I suggest an edit of your post, and cut/paste to another thread with some 'tea-spoon' explanation of your ideas for us mortals? :-)

I am not sure what you think about canopy markings though. I get a feeling that you're not the panning type, manually nor automatic padlock. Am I right in assuming you prefer a forward view, hence with a stable horizon reference enabling controlled maneuvering? And combine that with an instant target view (toggle or overlays?) with added symbology/annotation, as well as a explicit symbolic tool (aka 'elite' space sim) illustrating a 'mental picture' of the situation around you? The latter is hard to visualise from your post (althoug it reminds me of some screen shots from Mig Alley).

I tried to focus (here) on increasing spatial awareness with normal 3D viewing and environmental modelling (ie. the 3D models). Canopy markings is part of such environmental modelling, hence an implicit tool, not an explicit tool (like mini-HUD and suggested SA-sphere). Of course explicit aids have their place --- but I argue that they come second to implicit tools for increasing spatial awareness.

As for the whole question of SA (full situational awareness), position and movement of threats, and BFM maneuvering --- that's an important debate, but somewhat distinct to the question of spatial awareness. The latter is question of knowing/feeling/understanding aircraft and pilot head attitude and movement in relation to the 3D world --- no more, no less.

Also, I tried to limit this discussion to how to improve spatial awareness in-cockpit, thus ignoring the external view for the moment (although I am an advocate of what you call a '3rd person marionette' view toggle).

PS. Your BFM tools sound very interesting also. What about a thread (with pictures! :-) on this topic? I am all for better tools for practice and analysis of BFM --- flight path visualisation, added symbology etc. (How many of us mortal sim pilots know where the ideal elbow/entry point is? :-)

PS2. Andy --- maybe you could use some of your publishing expericence/writing skills to help Kurt with some editing/presentation? Don't give up so easily --- a lot of valuable ideas and knowledge in here... :-)

PS3. Kurt, what's your background in flying/sims?

Take care.

[This message has been edited by Attila (edited 12-18-1999).]


Posts: 579 | From: Scotland | Registered: Nov 1999  |  IP: Logged
Andy Bush
Member
Member # 12

posted 12-18-1999 06:30 PM     Profile for Andy Bush   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
Attila

Good post.

I lay no claim to being a 'great communicator'...I'll leave that to Ronald Reagan!!

Kurt

I know I've crapped in your mess kit in the past...I apologize. I realize you do have good things to say and good ideas that are of interest to us all. I'm interested in your ideas...I just can't understand what they are most of the time. As I said above, I'm no one to give writing lessons to anyone, but I would suggest that you stick with commonly accepted grammar, punctuation, and terminology. You have a talent for making up acronyms and nicknames that would be neat if only the rest of us knew what the heck you were talking about.

You have much to bring to these discussions. I value your technical background. Please take Attila's suggestion in the kindest way possible...he wants to hear what you have to say and so do I.

Andy


Posts: 595 | From: St Louis, Mo | Registered: Sep 1999  |  IP: Logged
Attila
Member
Member # 853

posted 12-18-1999 07:01 PM     Profile for Attila     Send New Private Message   Edit/Delete Post   Reply With Quote
Hi Andy,

You're quite right --- my comments about Kurt's posting is just an attempt to 'focus a lot of light to a sharper image'... :-)

And I didn't mean to suggest that you should act like a translator for Kurt --- just that he seem to touch on many interesting things about BFM as well as ingenious ideas for sim viewing, flying and analysis. It would be really cool if this became a source for one of your excellent articles! From your articles and posting you've proved that you have the skills to explain these things to a non-pilot like myself... :-)


Posts: 579 | From: Scotland | Registered: Nov 1999  |  IP: Logged
Attila
Member
Member # 853

posted 12-18-1999 09:48 PM     Profile for Attila     Send New Private Message   Edit/Delete Post   Reply With Quote
Andy and Kurt,

I'm a bit confused --- you both seem to say that the view from the virtual cockpit in sims doesn't look like the real thing. As real pilots I guess you know what you're talking about. But I don't understand what you mean.

Imagine you sit in the cockpit for real, then put on some head gear that limits your view to a small square similar to what we simmers see on our monitors (also ignore the stereo and focus abilities of your eyes).

Isn't what you now see and the virtual cockpit view visually the same? (My background in 3D is substantial enough to know that this is what all the calculations the computer does every frame try to achieve.)

In the real cockpit, you move your head to point your LOS --- we pan the virtual camera with keyboard keys (or point-of-view hat, mouse or trackball). The only difference is that you implicitly know the attitude of your head at all times from balance and muscle feedback. We simmers need visual feedback on our monitors.

Tilt (or camera/head roll) is particularly interesting, because simmers can't control it, and have difficulty deducing it from the visuals. This is important because it affects the tilt of the horizon, thus the understanding of aircraft attitude. (I guess some sims simulate no tilt, others try to mimic real head movements, eg. some tilt in the extreme over-the-shoulder situation).

My simple thesis is that canopy markings can help give us simmers the visual feedback to achieve the spatial awareness you have about head and aircraft attitude in the real cockpit. If the markings are well done, even a snapshot will instantly clarify head attitude and the horizon outside will then clarify the aircraft attitude (all the variables in the "viewing equation" solved!).

It might still need a lot of training though to achieve the awareness without thinking about it, but all variables would be there on the screen. And as with mice --- with training we will learn... :-)

Or did I miss something?


Posts: 579 | From: Scotland | Registered: Nov 1999  |  IP: Logged
Major Tom
Member
Member # 1256

posted 12-18-1999 10:40 PM     Profile for Major Tom   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
GAH!!!

Kurt!

That's one big dang message! I'm not even going to bother reading it, I get 4 lines down and you are already not making any sense to me. Please, try to condense your posts after you write them and for christ sakes, dont use so many spaces. It's damn irritating.

And what exactly is your background in aerospace? I hope I'm not prying and I hope I dont seem mean or vindictive by asking you that.

My personal background in aircraft ammounts to the completion of ground school and about 80 hours in small aircraft. It's not a whole lot of flying time, but its more than most people will ever get behind the yoke/stick. I'm currently enrolled and attending Embry Riddle Aeronautical Univeristy's flying program and I am happy to say, enjoying it a whole bunch.

Andy is honest about his background, Attila isn't afraid to admit he has pratically no time behind the wheel, and I've let it all hang loose. So why dont you tell us who you are and where you are comming from Kurt?

Atilla and Andy-

3 things immesnly help with the maintaining of situational awarness in the virtual cockpit.

-reflections
that vary enough so you can roughly tell what portion of the cockpit your head is pointing at
-virtual attitude indicator
helps a great deal in determining the relative relationship between the aircraft and the horison
-directional indicators (ala WWII fighters)
in padlock an arrow pointing to the nose of your aircraft is a quick and easy way to determine the best possible way to out manuver your foe. while looking strait ahead, an arrow towards the enemy helps in exactly the same way and at the same time allowing you to pay closer attention to your instruments.

This in my mind is the winning combination to success in virtual skys.

Agree? Have anything to add?


Posts: 1352 | From: Prescott, AZ | Registered: Dec 1999  |  IP: Logged
Attila
Member
Member # 853

posted 12-19-1999 12:46 AM     Profile for Attila     Send New Private Message   Edit/Delete Post   Reply With Quote
Major Tom,

I find your first point, reflections, the most interesting in this context. Your second and third point are explicit aids, and my view is that they don't necessarily enhance awareness, as much as providing mechanical maneuvering tools. That said, anything that helps is welcomed.

Regarding reflections --- giving us a rough indication is not enough if the camera/head tilts. You need more precise canopy markings in combination with the horizon outside to infer the spatial situation (eg. some horizontal lines to give you tilt reference).

If the camera/head has no tilting, a rough idea might be enough. That leads me to think that camera tilt is just an avoidable complication factor. It is perfectly possible to swivel the camera/head without tilt, and I guess many sims already do (although that simulates a rather robotic neck :-).

Reflections may replace the functionality of the lift-line because of symmetries in the reflections when looking up. But around 3 and 9 o'clock, at medium high pitch angle, reflections are a bit vague. Knowing whether you're looking into the forward or aft hemi-sphere is important for maneuvering, I would think.

Maybe the combination of reflections and a 3-9 line would be a good solution?

PS. I have no time at all in the cockpit of an aircraft --- just to make that clear. I'm just a software developer with an interest in sims (development and flying :-).


Posts: 579 | From: Scotland | Registered: Nov 1999  |  IP: Logged
Major Tom
Member
Member # 1256

posted 12-19-1999 03:02 AM     Profile for Major Tom   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
The bad thing about trying to create totaly realistic cockpit reflections is that I can just focus with my eyes right through them in a real life. But that is impossible to do in a sim.

Come to think of it, every solution to the situational awareness problem in simulations is a cheat in one way or another.

What I think would make this business a whole lot simpler would be to realistically restrict the virtual pilots head movements. Second would be to represent a more accurate picture of what a pilot sees in an aircraft. All to often in simulations it seems like (at least to me) that I'm staring at the world through a cardboard paper tube tunnel.

When simulations start to be enginered for 1600x1200 resolutions, I feel that we will be able to create a more realistic view what a pilot can see. When panning around the cockpit in most sims I really do feel like I'm looking through a tube that is acting as my eyeball. Only the tube is moving at 1/1000th of the speed of my eyeball and it's getting damn annoying.

Point your head in any given dirrection, then move your eyeball around and see how quickly you can lock onto and track something. You do it almost instantainiously. What I think sims have been doing wrong, through no fault of their own due to lack of resolution options up until now, is that they try to simulate what they eyeball can percieve at any given moment. But they fail to simulate periferal vision, what you can see beyond what you're looking at at a given moment.

Simulation developers need to simulate a virtual periferal vision at resolutions up t and exceeding 1600x1200. Obviously there would be no need to simulate natural autofocusing and bluring effects, because the eye is already doing that naturally.

Well, it's 2:00AM and I probably didn't make much of any sense. But on the off chance that I did make something that sounded remotely like coherent talk, please feel free to respond.


Posts: 1352 | From: Prescott, AZ | Registered: Dec 1999  |  IP: Logged
Major Tom
Member
Member # 1256

posted 12-19-1999 03:05 AM     Profile for Major Tom   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
Or in otherwords we need to change the system of what we see and how we look around in our virutal offices in the sky :-)

The deck is fixed and no matter how hard we try to cheat the current system, we'll never win. We may win a hand or two, but there is no real hope for the virtual cockpit view system as we know it.


Posts: 1352 | From: Prescott, AZ | Registered: Dec 1999  |  IP: Logged
Vector
Member
Member # 463

posted 12-19-1999 03:15 AM     Profile for Vector   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
Don't you guys get it? Kurt posts all this stuff and then laughs when he sees us mere mortals try and understand it

------------------
-\/ector, Flight Sim Enthusiast


Posts: 903 | From: Comox, BC, Canada | Registered: Oct 1999  |  IP: Logged
Kurt Plummer
Member
Member # 358

posted 12-19-1999 05:31 AM     Profile for Kurt Plummer   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
AB, At, MT,

Get me an EML, I've put together some stuff if you're interested.

It's primitive but it should get across a couple of the ideas I'm hoping to present.

At the rate I draw it's gonna be halfway to the /next/ Millenium before I finish... I guess your used to that?;`)


Kurt Plummer


Posts: 672 | From: | Registered: Sep 1999  |  IP: Logged
Andy Bush
Member
Member # 12

posted 12-19-1999 09:33 AM     Profile for Andy Bush   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
Folks

These last posts have been good reading, and I like the direction everyone is going. I'll sum it up by saying that we all are trying to make a good thing better.

Let me add one more parameter to the discussion...not a new one by any means, but, as far as I'm concerned, one that is at the very heart of BFM. I intentionally make reference to BFM since a padlock can have uses other than air to air.

Effective BFM is BFM flown out-of-plane, ie split plane maneuvering. To date, padlocks offer a technique that is best suited to in-plane maneuvering...pure pursuit or 'g for brains'. What ever improvements we or others can suggest to the padlock view must see this as the number one 'problem' to be fixed. Pilot orientation while looking thru a canopy that has no nose (or attitude) cues is also important, but it does no good to know where we are looking if there is no way to do anything with it.

As for the issue of reflections and 'lift lines', I'd just as soon have the canopy scribed with lines...in effect turn the canopy into sort of a WW2 bird cage, if that were to get us to where we want to go. Would this be a 'cheat' as Maj Tom mentioned? Certainly. But, so is everything that we are presently using. It would be hard for someone to defend the 'pointing arrows', 'lift lines', or reflections, while not being willing to accept other SA improvement techniques.

Goodness...if we keep going on like this, we may end up using the external...

Andy

Kurt

If you have something that you would like me to look at, send it to [email protected]


Posts: 595 | From: St Louis, Mo | Registered: Sep 1999  |  IP: Logged
Major Tom
Member
Member # 1256

posted 12-19-1999 06:44 PM     Profile for Major Tom   Author's Homepage     Send New Private Message   Edit/Delete Post   Reply With Quote
I'd be willing to accept anything that would enhance my virtual awareness.

Today I went down to the local flight museum, after some persuasion they let me into the cockpits of a few of the fighters they had on display. Theoretically, even if I had no inner ear or ability to determine the position of my head, I'd be able to determine where my head was pointing in the virtual cockpit. This is because with periferal vision, while staring strait up in the F-4 Phantom II cockpit, I was able to see both the top of the hud and the arch(s) that hold the glass of the canopy in place, all out of the corner of my eye. And I could still focus on the HUD relatively well if I actually bothered to look down with my eyeballs.

I went in a few other aircraft, experimented a bit more and found that no matter where your head is pointing in the cockpit, you can always see some part of the cockpit out of the corner of my eye (other than reflections, lift lines, or imaginary red arrows :-) )

The conclusions I drew from my little expirement today only confirm that we really haven't been seeing what real fighterpilots see in real life.

I believe we now have the technology and the resolutions to portray a truely accurate flight sim picture.


Posts: 1352 | From: Prescott, AZ | Registered: Dec 1999  |  IP: Logged

All times are MST (US)  

   Close Topic    Move Topic    Delete Topic next oldest topic   next newest topic
Hop To:

Contact Us | COMBATSIM.COM Home

COMBATSIM.COM, INC. All Rights Reserved.

Powered by Infopop Corporation
Ultimate Bulletin Board 6.04b

Sponsor
2014 COMBATSIM.COM - All Rights Reserved