stunt simulator

In few recent posts on http://www.rcmodely.sk/Controlline/diskusia.php are links to Keith R. downloadable simulator and other files.

Most recent version in resolution 1024x768 is downloadable here:

http://www.netax.sk/hexoft/sim/sim.zip

I tried this. It seems more like a video. The airplane flies itself and does manovers. I have no control over the airplane. After several minutes the computer locks up and I have to reboot. Is this how it is suppose to work???

Sim shows definition of our pattern. The key is to look from static point using several types of camera and/or move around the circle to see figures from different angles or from center of circle. That what you see is what you have to see if figure is properly flown. It is not a game it is tool for pilots and judges.

If those "several" minutes is end of program, then it enough to press one of control keys. ESC will take back from sim.

I reccomend to read those attached documents.

LAST EDITED ON Dec-04-04 AT 06:06 AM (CST)
 
It hung my computer every time.
I RELAODED IT AND IT RUNS FINE.
Ed

does it happen on both versions? I mean high and low resolution. Can you write something more about your computer? Type, OS, memory, screen resolution ....

I believe Keith will respond soon.

This is the best time consumer to come along in years. I love it. It works very well.

Keith, Igor,

This is fantastic. It works perfectly... The perfect pattern, finally! I can watch it for hours from all different camera perspectives... Wow... And the 1,5 m Radius turns.. Impossible but great!!
Some suggestios: add a model library and add some noise!!!
And how about a camera view from the model towards the pilot??

Great job!
I love it!

HI

Keith did a great job on this, looks like it would be much more useful if you had the plane do 12 to 15 ft radius corners, and also mine will not show the 4 leaf ?

Regards
Randy Smith

I have watched and watched this....good stuff here!
I would like to see "x"'s or "poofs" where each maneuver starts and stops.

See Ya

LAST EDITED ON Dec-05-04 AT 02:47 PM (CST)
 
In the user's manual it's explained why the clover isn't shown. View one looks like what I get with my helmetcam except my maneuvers aren't that straight or low. But mine is noisier! Getting some of the different views is neat, like watching the pattern from near the top of he circle.
When Bob Reeves drags out the plane mounted video camera again I'm hoping he'll tape a flight looking down the wing at the pilot.
.
.
Well it worked fine at work but my home computer locks up when trying to run it. I keep getting an "Invalid Page Fault" error. Dang!

WOW!!! WOW!!! ..... WOW!!!

Tim

That is sooo COOL !! Works great. I'll be he spent a lot of time developing this. Now, if someone can just hook it up to my joystick, I could make it crash!

LURK NIRVNA!! This is F-A-N-T-AS-T-I-C!
I was able to adjust angles until view was very close to actual pilot position. It was almost like really flying except my bottoms are a bit more consistent(:

Can't wait until all the angles are worked out so the view of sizes and shapes are fine tuned. What a great tool!!!!!
Please Gimme more................

Pretty cool stuff, I just watched the first full flight and am ready to mess with the view controls more now. But need to read the info first. ARRGGGH. Hats off to Keith on this project.
Lee TGD

Hi all,

Thanks for the nice comments. I think that step one is to download the basic manual that explains why it was done and how it works. As per my previous posts, I have been holding back on releasing it because it is still a bit rough and unfinished without the clover. This 3-D stuff does not come naturally to me, so I am on one hellavu learning curve. If I wait until it is perfect, it probably would never be released!

Basically it was an attempt at producing a simulation of the FAI pattern that is flown 100% to the rules, in order to teach perspective when training judges. I was trying to lay to rest some of the age-old arguments on how the maneuver shapes look from the judges view point. Strangely, this has caused some more arguments, and all I have done is to animate each maneuver over their respective shapes that are tagged onto a sphere. If the pattern is not supposed to look like this, then we should change the rules.

I also found the 5 ft. corners to be mean! The other thing that we don't see, is the slight curve on top of the square stuff. This is most likely due to the fact that we fly much bigger radius corners, and there is just no time to fit the curve (flying parallel to the ground)in as well.

I think that the problems with the computers locking up, or hanging, is from the type of graphics card. I use a GE-Force 5700 card, and most of the older cards will work just as well. The software, called Blender, uses Open GL, instead of Microsoft's Direct-X, so maybe this is a problem for some pc's. I know that there are some clashes with the ATI Radeon cards. Any kid that is into PC gaming can help sort out these sort of problems. The sound.....well, I have tried to do this, but right now, Blender has a problem with the Doppler shift part of the program, and as soon as it is sorted out, I will add some sound.

My own version has what I called a "Fly-cam" that is attached to the model. I can rotate it around to see just how the pilot (inside) would see things. This is fun! I am constantly improving things, so I will add it soon. The model airplane is basic, and I have learned how to draw better models and terrains now, so this will be fixed soon.

I hope that this answers some of the questions. Thanks for the input.

Keith ,

You are spoiling us.
I'll put it on some CD's :Great newyears present for my colleague stunters.... Or do I need to buy licences??

Perhaps it is a start to re assess some rules in the book as well...
Radii of turns, shape of the squares....,

Thx!
Paul

(And this message puts this topic on top again)

Wow..what a grweat tool!

My son has started to use Blender for some 3d modeling and animation.

But does Keith's sim run very slowly on everones machine? It is doing frame updates at about 1 per second on my 600Mhz P3. I have a pretty current N-Vidia Graphics card and it runs MS flight sim pretty well.

Keith's pgm is doing 1 minute laps so is pretty frustrating to view.

Is there a way to get it to go closer to realtime?

I tried changing color depth and screen resolution to minimums but it makes no difference in the way it runs.

Any hints? I saw nothing relevent in the Docs to this problem.

Thanks Keith--with the open nature of Blender, I suspect this will really take-off!

Hi again,

Paul you don't need any license from me, but I suspect that I need to include the open-source info and credits to Blender somewhere. I need to find this out when I finish it. Glad you are all enjoying it.

Curt, my older PC has the old N-vidia Riva TNT-2 chipset, and the new sim slows down on it. I still have your e-mail, so I'll send you a quicker version. The maneuver shapes are the culprits, and I need to find a way to use them without slowing down the processor. I'll send you the previous version in the meantime with the 800 x 600 graphics. This should be better. I'm a bit spoilt with my new P-4 3 Ghz plus the GE-Force 5700.

Tell your son to persevere with Blender, it is worth the effort, and certainly does not cost a whack like Maya. The old Blender Gamekit book is also worth getting as an intro into game graphics. It is a bit old now, but the principles are still sound. Enjoy!

Thnaks Keith:

Well, I loaded it onto my laptop which is a 1.2Gig celeron. The hi res Sim runs fine--near realtime. WHat a treat. My wife actually watched several full patterns with me. She did, of course, add appropriate sound effects as she pointed out the desperate place in the wingover where I planted my plane at the Nats last year!

Thanks for the great tool!

OK guys...who will be the first to take this package and set it up for virtual reality helmet and glove?

LAST EDITED ON Dec-17-04 AT 04:31 AM (CST)
 
Keith

It is brilliant and will be a great asset to my plan for competing next year. I have wondered if it is possible to preset the lap time rather than rely on the CPU speed. A little row of judges would be a useful addin for the pilot's eye view.

edit: When running the schedule from the pilots eye view I found that the overhead and vertical eights were quite strange at the transition overhead. The model seemed to miss out the center part of the manouver.

I am sure that there will be a lot of suggestion that will keep you busy

great stuff thanks

Regards

Mike

>Keith
>
>It is brilliant and will be a great asset to my plan for
>competing next year. I have wondered if it is possible to
>preset the lap time rather than rely on the CPU speed. A
>little row of judges would be a useful addin for the pilot's
>eye view.
>
>edit: When running the schedule from the pilots eye view I
>found that the overhead and vertical eights were quite
>strange at the transition overhead. The model seemed to
>miss out the center part of the manouver.
>
>
>I am sure that there will be a lot of suggestion that will
>keep you busy
>
>
>great stuff thanks

Hi Mike, Thanks for the nice comments. I find that the marker poles (F12) are fine for checking where the model is when using the pilots view. Putting objects where the real judges should be, gets in the way of the other camera views. (They are also a hassle to draw!)

The overhead transition on a PC screen always swings the scene around. The R/C sims are exactly the same. If you observe your own head movement when you follow your model through the wingover, you will see how you have to rotate to keep the model in view. I may try and add a tiny delay in the tracking movement to make it look better.

The glitch in the vertical eight and hourglass right at the top, is a problem that I know about, and will sort it out soon. Good luck with your plans to competee next year.

Keith,WOW only words I can think of..Boris

Kieth,

I really like this and see that it can be of great use. But I have a question regarding the Triangles. when the Triangle and Hourglass figures are both displayed the angle of the sides are different and I think this to be incorrect. The rules say the sides of the Triangles are to be of equal length as are the angles, and equilateral triangle. The Hourglass according to the rules is comprized of 2 equilateral triangles connected at the vertical tips. Since the Triangle has radiused corner at 45 degrees elevation, the resultant base side would be longer then that of the Hourglass because the other sides are tangent to the top turn radius. The Hourglass has an intersection in the middle and if you were to place a turn radius tangent to the sides the top of the radius would not be tangent to the 45 degree elevation, it would fall below it. It looks to me like the Hourglass is equilateral but you tried to make the base of the Triangle match the Hourglass length. In order to fit the turn radius at the top the climb and dive sides were spread out to fit the radius tangentially. That being the case the Triangle is nolonger equilateral because the angles were made steeper to accomodate the turn radius.

If you lay this out flat you would see that the up and down sides of the Triangle and Hourglass would be parallel but not coincident, the Triangle would be wider. This can be seen with the naked eye in the rule book and if you measure it you will see what I have discribed.

Please take no offense, I think this is fantastic, but if its going to be representative of the rule book the geometry has to be correct.

Randy, base for triangels and hourglass are defined equivalent. So that round corner at 45 deg makes different angle which is defined in rules only "approximately".

But check yourself:

http://www.slovanet.sk/orsia/rules.pdf

BTW your post says for all - that sim does good job ... finally we see what is written in rules

When I saw this version first time, I wrote to Keith that those wire figures show figures much better than any picture before. I vote for judges guide in such electronic form

Igor,

You are correct in that the angles are stated "approximately" however they are both identified as a nominal 60 degrees implying equilateral triangles, so geometrically the triangle loops area still wider at the base then is the hourglass base. But both are still "1/8+ laps" long.

Its a trivial point to be sure, but as Keith said he wants them to appear geometrically correct including the impossible corner radii. Actually the FAI and AMA rules for these two maneuvers concur because they both agree that the triangle shapes are equilateral. If they are equilateral it is geometrically impossible for the sides of the triangle loops and the top and bottom of the hourglass to be the same length, the triangular loop's sides must be longer by some small degree.

>Kieth,
>
>I really like this and see that it can be of great use. But
>I have a question regarding the Triangles. when the Triangle
>and Hourglass figures are both displayed the angle of the
>sides are different and I think this to be incorrect. The
>rules say the sides of the Triangles are to be of equal
>length as are the angles, and equilateral triangle. The
>Hourglass according to the rules is comprized of 2
>equilateral triangles connected at the vertical tips. Since
>the Triangle has radiused corner at 45 degrees elevation,
>the resultant base side would be longer then that of the
>Hourglass because the other sides are tangent to the top
>turn radius. The Hourglass has an intersection in the middle
>and if you were to place a turn radius tangent to the sides
>the top of the radius would not be tangent to the 45 degree
>elevation, it would fall below it. It looks to me like the
>Hourglass is equilateral but you tried to make the base of
>the Triangle match the Hourglass length. In order to fit the
>turn radius at the top the climb and dive sides were spread
>out to fit the radius tangentially. That being the case the
>Triangle is nolonger equilateral because the angles were
>made steeper to accomodate the turn radius.
>
>If you lay this out flat you would see that the up and down
>sides of the Triangle and Hourglass would be parallel but
>not coincident, the Triangle would be wider. This can be
>seen with the naked eye in the rule book and if you measure
>it you will see what I have discribed.
>
>Please take no offense, I think this is fantastic, but if
>its going to be representative of the rule book the geometry
>has to be correct.
>
>Randy Ryan <><

Hi Randy,
Thanks for the feed-back. My apologies for the delay in this reply, but my PC was zapped by lightning, and I’ve only got back in action now. I also don’t take offense at positive criticism at all. The whole reason for doing this project in the first place, was to show a pattern that is flown 100% (or as close as I can draw it) to the rules. The problem is that there are discrepancies between the AMA and FAI rules, and of course between the way that people actually interpret the rules. Just by the way, I saw that the last corner of the hourglass in my sim does not follow the same path as the triangle. My drawing of the maneuver shapes, needs some more work! I hope to sort this out by this week-end, and then I’ll get it to Igor.
Brett Buck did a nice article on the hourglass in the Jan/Feb 2003 Stunt News (page 46/47), where he illustrated the problems, and showed the various ways of performing the triangle and hourglass. It would be wonderful if we all decided that the rules should actually tell us what is expected, and even nicer if the AMA and FAI rules agreed with each other. My sim is based on the FAI rules that we are busy trying to sort out at this time, and right now, the sim is having the desired effect of making many people see clearly that we need to improve the definitions and drawings. We all know too well that shapes with “straight” sides on spheres, can only approximate shapes drawn on flat planes, so there is always some compromise. The angles of the corners of the triangle for instance, add up to more than 180 degrees. I hate the old argument that it is all “subjective” anyway, so the rules don’t really have to be that specific. Shapes drawn on spheres are governed by some very basic math principles that are not open to subjectivity, so the rules should keep this in mind, and be written accordingly.
Take the squares for example, the AMA rules call for equal length sides, and the FAI rules state that the sides must be vertical. This makes the FAI squares have tops that are much shorter than the other sides. How many times have you read articles on how to “fudge” the squares, so that look more like real squares to the judges? I have never seen any rules that actually say this, but you still hear this all the time. Hopefully the sim will generate much needed positive debates, and we can indeed get some common understanding on the shapes. With the ability to move the viewpoint around and look at the shapes from any angle, you soon learn to understand the perspective of the shapes.
Let’s all keep working together on this, and I’m sure that we will end up with a better set of rules in the future. Thanks again for the good comments.

I recommend to read most recent messages on :

http://www.rcmodely.sk/Controlline/diskusia.php

especially my post #227

There are several others links also to simulator working at lower resolution. It will probably work better on slower computers. (unfortunately with little less functionality)

LAST EDITED ON Dec-09-04 AT 05:29 PM (CST)
 
Put it on view #1 and put the point of view on the edge of the hemisphere right in the middle of the main maneuvering plane at about 30 degrees of elevation and it will track the ship while it maneuvers around you. It's easiest setting this point of view once the flight is underway but it will keep this point of view for the next flight. You probably won't learn anything but it's a neat effect! It's getting about a 5.41 lap time on my work computer.

this would make a cool screen saver too!

Thanks
Paul

On my machine, it takes 45 seconds for the plane to "lurch" through a lap. Is that too slow? Maybe I need to trade in the Tandy TRS 80 on something a little faster. Seriously, my machine is an older HP Pavilion, so I probably need to upgrade.
Glen

Yeah, it wouldn't run at all on our old 333 Pentium, but it cooks on the this new P4 2.8gHz machine.

I get the text...But not the Fun. All I get is a BLACK SCRENx(

The speed of everything that you see on any PC games, is governed by its ability to process everything. 3-D graphics chows up processor performance like nothing else. That's why these fancy 3-D accelerator cards have been developed. Recent PC's with P-4 processors, often have Advanced Graphics Processors (AGP) as standard. You really need some sort of 3-D card to use the sim without seeing it do 1-minute lap times.

My sim with the maneuver pics, slows everything down a bit more, because the drawings are not just lines, they a thin tubes, because the Blender software cannot draw lines in the game engine. Thin lines also tend to disappear as well, so the tubes work better.

I've found that I need to put them in a different graphics layer, and then bring them out only when they are accessed. Right now they are just hidden on the same layer, but out of sight of the camera's. It will get better soon!