Logger device for F2B - does it exist?

Hello,

I would like to have a logger device with which I could plot the flying path of my plane. Quite the same kind of graphs as here, but 2D is enough:
http://www.eagletreesystems.com/press/GoogleEarthPress.pdf

The flying path should be drawn and the correct path should be plotted with different color. I think GPS is not accurate enough for this purpose, so it should be some radio-wave based system. Is this kind of system already available?

Thanks for info!

Best Regards,
Jukka Juslin from Finland

I don't think anything exists as you describe. Also, I'm not sure commercial GPS is accurate enough or capable of tracking (update of position) quickly enough to be useful. (I assume you are talking about a control line model airplane? If you are wanting to track a long distance RC or FF journey, the GPS would work fine.)

Someone correct me if I'm wrong about the commercial GPS capabilities. If I am wrong, there are certainly some wonderful (tiny) GPS solutions (chip sets) around nowadays, and what you are visualizing is clearly a possibilty.

[photo not recovered: 23926.jpg]

"There are no exceptions to the rule that everybody likes to be an exception to the rule." -Charles Osgood

Juslin

I was in Finland most of last week, in Hamina and Vaalimaa - it would have been nice to have met you. Where do you live? Helsinki?

Charles

>Juslin
>
>I was in Finland most of last week, in Hamina and Vaalimaa -
>it would have been nice to have met you. Where do you live?
>Helsinki?
>
>Charles

Hello Charles!

I live in Helsinki. We have a CL place (very good) at Helsinki Malmi airport.

Would be nice to meet, if anybody is going to visit Finland - please let me know! Best time to visit is during the summer...

Jukka

Jukka

We have a machine in Vaalimaa which occaisionally needs repair / service. This last visit (repair) was at very short notice. If I go again I will email you, it would be good to see you. I can usually organise to stay over the weekend, if necessary

Charles

Larry:

I think you have the answer already--the sensitivity and resolution of high-end GPS is probably sufficient but the update rate is not fast enough.

The 3D video processing systems CAN do this however. The same technology that is now used to analyze the kinematics of athletes and racehorses.

Several years ago I configured a test system to track a reflectivs set of spheres (like ping pong balls) placed on the exterior of a car. The car was on a four-post servo-hydraulic shaker. The shaker was re-playing the vibration signature of a railroad car transporter.

The system captured the video in real time. Post processing, using the wireframe model of the car could produce a 3D track of any point on the wiremesh--from the data captured from the three video cameras and the three reflective points.

It was pretty cool and could readily capture what we do with our planes---but $,$$$,$$$ project. But they don't have the problem of the cars smacking into one another on the trains anymore!! It was a real problem!

FWIW

Curt

>I don't think anything exists as you describe.

I wish it did.

Randi

Hi Jukka,
(I hope that this post works now. I wrote a reply and it did not work, so if this comes out twice, then I hope that I can remember what I said the first time!!)
It is technically possible to create a tracking system for our stunt pattern. It will just take someone to actually do it that has the experience, time......and money. There will be very little commericial value in such a project. The ball trackers used for cricket and tennis, such as Hawkeye, use a number of video camera's that feed into capture cards, and then some fancy software translates the co-ordinates into a 3-D path. Another method would be to use a small transmitter on the model and then place a number of receiver aerials around the perimeter of the circle. The phase differences between the aerials can then be calculated and the flight path can be plotted.

I also believe that inertial sensors can be used in the airplane and the information can then be stored or transmitted for plotting. I continually work on trying to find a method that is practical and not too expensive. This would give us a wonderful way to train judges with real-time feedback of the flight path. I use video from a fixed point to do this, but it takes time, and requires manual tracing. Each manoeuvre takes me around 10 to 15 minutes. Just by the way, you do not need a high resolution like 1 cm to do this. Video frames are 1/25th of a second, and this relates to steps of more than 80 cm each, depending on the speed of the model. I've attached a shot of a hourglass performed by Serge Delabarde at the 2004 world champs, and I've shown the dots for each step. That sort of resolution is good enough to see the flight path. It also shows just how far off the rule book track that we fly some of the manoeuvres.

If you attended my presentation at the F2B workshop in Belgrade at the Euro champs, then you would have seen the many traced pictures that I used to good effect. If anyone else has a system ready-to-roll, then I am interested! This is an interesting project.

[photo not recovered: 25244.jpg]

Attachment #1, (jpg file)

Is that an hourglass? So much for 5 foot radius corners!

(Incidentally, YES, trees DO live on oxygen. They just happen to PRODUCE more oxygen than they CONSUME,DURING THE DAY! Ever wonder what they do at night? Wonder who comes up with these things.. Don't live on oxygen, indeed... )

Yup, so much for 5ft. corners!! I have plenty of traced squares as well. They all look similar, and the corners are usually around 16 ft. radius. I've added a picture of a square that scored an average of 8 points at a recent world champs. It sure looked quite good in real time, but when you compare it to the rule-book FAI square, then what sort of score would you give it??

Judges training with reference to the actual shapes will sharpen their skills, an we will see an initial lowering of scores in general. If we can assure the pilots that they can indeed perform the maneuver shapes to the rules, and not to any particular whims and fancies of judges, then the competitive stunt will be improved.

[photo not recovered: 25258.jpg]

Attachment #1, (jpg file)

>Yup, so much for 5ft. corners!! I have plenty of traced
>squares as well. They all look similar, and the corners are
>usually around 16 ft. radius. I've added a picture of a square
>that scored an average of 8 points at a recent world champs.
>It sure looked quite good in real time, but when you compare
>it to the rule-book FAI square, then what sort of score would
>you give it??
>
>Judges training with reference to the actual shapes will
>sharpen their skills, an we will see an initial lowering of
>scores in general. If we can assure the pilots that they can
>indeed perform the maneuver shapes to the rules, and not to
>any particular whims and fancies of judges, then the
>competitive stunt will be improved.
>
>

Keith,

Excellent information!

I was thinking that somebody probably has done this already and no surprise it is you. Its a pity I was not at Belgrade. Will you be in France at Landras, having some workshop at 2008?

I would opt for radio frequence and receiver in the plane. Good thing is that it is doable - mostly probably matter of who will do it first. I had dreams before that I could use extra smoke from the plane to do the same real time - but after my smoke bomb experiments I found out that it doesn't really work.

I hope this will progress! I feel this is more important for me than eagle tree motor loggers.

Jukka Juslin From Finland

>Keith,
>
>Excellent information!
>
>I was thinking that somebody probably has done this already
>and no surprise it is you. Its a pity I was not at Belgrade.
>Will you be in France at Landras, having some workshop at
>2008?
>
>I would opt for radio frequence and receiver in the plane.
>Good thing is that it is doable - mostly probably matter of
>who will do it first. I had dreams before that I could use
>extra smoke from the plane to do the same real time - but
>after my smoke bomb experiments I found out that it doesn't
>really work.
>
>I hope this will progress! I feel this is more important for
>me than eagle tree motor loggers.
>
>Jukka Juslin From Finland
Hi Jukka,

I am hoping to be in France next year, and Peter Germann will be arranging for another workshop. After the success of the Belgarde workshop, we will be hoping that more pilots attend. What happened at the workshop was that we held it on the Saturday before the Euro champs, so the pilots that arrived preferred to get out to the field and fly, set-up their equipment etc. We therefore had just a handful of pilots. What them happened was that news of the presentation spread, and some pilots felt that they had missed out on something good. A few of them asked Peter Germann if we could repeat the 3-D graphics part of the workshop again one evening after flying had finished. The Serbian organizers were so helpful with everything, and once again arranged for the conference room, plus PC and projector, and we repeated the presentation. I was well prepared for the usual objections about the square shape, and how everything is just subjective, leave the event alone, we are happy with it just as it is etc. etc. etc. To my surprise, there were zero negative comments, only good constructive feedback. In France, we will most likely try to do this in the evening.

Back to the tracking thing. One of our regular judges is a highly qualified electronics engineer that has designed many tracking systems for vehicle and people tracking. As Brett and Larry mention below, the inertial system has many inherent problems and most likely would have to have some sort of reset system to compensate for drift after each maneuver. The camera system is a good way, but you would need quite a few to cover the hemisphere. The radio system meausuring phase differences is most likely the simplest to give a basic path.

As I stated before, I see the main use in judges training, and of course, pilot coaching as well. I certainly do not believe that it is a good idea to use this sort of system in actual judging. At least not now. The F2B working group that was started by Peter Germann is working hard to improve the judging standard at international level, and a giant step forward will be achieved when pilots and judges all share the same understanding of the maneuver shapes. This should be easy because we only have a single axis airplane. As I said in my Belgrade presentation....."it's just like golf, all you have to do is to hit a small white ball down 18 holes. How difficult can that be??" Anyone that has tried to play golf will know that there is a lot more to it. The same goes for our most beloved sport. The problem comes in because the model traces paths on a ball, or sphere, so the paths are all circular, or should I say, spherical.

Somebody wise said that "Old habits die hard." This is the case with stunt flying. The many misconceptions that regularly come up, are like old wives tales that have been carried on from generation to generation. The stunt rules have been modified through the years, but the basic pattern has been the same since George Aldrich thought it up. Years ago in international competitions, there were not that many outstanding competitors, and the really good pilots stood out above the rest. These days, if you watch the top 15 fly-off at the world champs, everyone looks like a winner. The pilots everywhere have access to much better equipment, and they all put in serious amounts of training. Unfortunately we have not provided the judges with equipment to help them to keep up with all of these improvements.

The F2B workgroup has representative from all major role playing countries, and we all realized that we could really improve the event by creating a common understanding of the rules. It's no good for pilots to train and coach to fly the pattern to the rules when the judges from each country have their own ideas about the basic shapes. One of the real problems in curing this, was translation from English. Rules usually look like legal documents, and people do not often read them in detail, especially if they believe that that they already know exactly what to do anyway. Interactive 3-D pc graphics, however, can illustrate all the pricinciples that we need to know with animated pictures. This crosses the language barrier perfectly.

My presentation worked so well in Belgrade, and the old CL-Sim has also helped many to understand the basic principles. So the next step is have a real-time feedback system to help with judges training. It also makes judging much more interesting, as it's something new to learn, and learning is fun. My apologies for taking so many words to explain why I would like a tracking system. (Maybe this is why I get on so well with Ted Fancher!......I suppose it has something to do with our passion for the subject)

And that's a very GOOD corner. Years ago, I stepped through a Windy video of a particularly good flight, frame by frame. Using a nominal 70' line, and the frame rate, I calculated a 17 foot radius for the corners. Several people insisted that I was simply wrong, I HAD to be, of course!

If we DID have a tracking device, it would be condemned by many, convinced that their flights just HAD to be much better.. Or its software would be blamed. The world of competition Stunt cannot deal with a scale, let alone the cold reality of flight shapes.. ;->

The prime users of such a device would likely be individuals, who were working very hard to practice and improve their pattern - much like several use video cameras nowadays. But for judging a contest? Never!

Tell me I'm wrong.

[photo not recovered: 23926.jpg]

"When the cup is full, carry it even." -Scottish Proverb

>And that's a very GOOD corner. Years ago, I stepped through a
>Windy video of a particularly good flight, frame by frame.
>Using a nominal 70' line, and the frame rate, I calculated a
>17 foot radius for the corners. Several people insisted that I
>was simply wrong, I HAD to be, of course!
>
>If we DID have a tracking device, it would be condemned by
>many, convinced that their flights just HAD to be much
>better.. Or its software would be blamed. The world of
>competition Stunt cannot deal with a scale, let alone the cold
>reality of flight shapes..

We could easily deal with it and EXACTLY THE SAME PEOPLE would win that do now. All it would illustrate is that some people come closer to others.

It's a conceit (mostly by non-competitive fliers) that we fly A LOT better than they really do. Even good flights by top contenders have MANY BIG ERRORS on every flight, and top contenders are generally well aware of it. There are no perfect flights, not even close.

>The prime users of such a device would likely be individuals,
>who were working very hard to practice and improve their
>pattern - much like several use video cameras nowadays. But
>for judging a contest? Never!
>
>Tell me I'm wrong.

Uh, you're wrong. Video cameras see some things very well and others not at all well. There's no reason to do it, of course, because all of the competitive pilots and all of the good judges ALREADY KNOW that the maneuvers look like the tracing above. So what? The goal is to find who came the closest, not to prove that no one is very close.

And if you think it would reduce the controversy, well, there you'd be wrong, too. As soon as you had a perfectly accurate track, then you would have to determine a score from it. Which requires a scoring algorithm. Which would have to assume some sort of weighting of the errors from perfect track. Then, the guys who finished out of the money and are chronic conspiracy buffs would determine that the scoring algorithm was weighted to *hurt* them and *help* the winners. Which is not far from what they currently believe. Others would argue vociferously (as they do now) that stunt was bogus and invalid because the minimum radius was 13 feet instead of 5 - still ignoring the other aspects. Others would go out and try to get their 5-foot radius, get it down to 11, then complain at loud volume that they deserve to win - even though getting the radius down created so many other errors that they had worse flights. In other words, *business as usual*.

We pretty much know that what Keith's tracings tell us is true. Measuring it live wouldn't change ANYTHING.

Brett

Now Brett,

Are you trying to divert this discussion to some subject of "conspiracy"?

Sometimes you remind me of a dog who's been kicked so many times he flinches and tries to bite any time anyone walks by.

No one disputes that video cameras don't see everything, do they? But they really don't have to. My calculation of 17 feet for the radius was simple. I assumed the flyer (Windy) was using 70 feet lines, and added a couple of feet to that for a radius for his arm, then calculated the circumference of the flight circle. I timed a level lap just prior to the maneuver, and from that I got a speed. Then I stepped the video, frame by frame through the transition between horizontal and vertical. Knowing the frame rate (30 Hz) and the number of frames, I knew the duration. With the speed known, I calculated the distance travelled through the QUARTER circle for the corner. That allowed me to come up with a radius of 17 foot. Nothing terribly exotic about it, I repeated it several times. In spite of what a video camera cannot see, it was very plain that the radius was ~17 foot.

Now, let me straighten you out about how this particular under-achiever flier feels about the mighty gods of Stunt. I KNOW they are great flyers, I've watched them many times, and I give them credit. And I know some of them are just brimming with natural talent, and don't have to practice much. I also know that others work pretty hard at it, burning 75 gallons or more of fuel a year. And a few use video cameras to tape and critique their own flights. So, at least SOME of the flyers do practice quite a bit to gain their great skill. And I believe these guys would be the first to try to use such a logger to improve their pattern.

Other than this, my take is that these noblemen of Stunt pretty well put their pants on one leg at a time, just like the rest of us mooks.

So, forget the projections of conspiracy theories, Brett. We all know just how fabulous you guys are, and we aren't trying to cheapen it by insinuating that you just practice so much you automatically beat us.

Another clue for you. In spite of your suspicions, I've never heard ANYONE seriously suggest that with different judges or scoring methods that the same anointed gods of Stunt wouldn't STILL win the contest, nor that there is anything inherently unfair about the best fliers winning the contests. To suggest otherwise is a distortion of reality.

We're not all after you Brett, nor are we even mildly jealous of your ability, we don't begrudge you of anything; please realize that every discussion is not about some conspiracy theory.

L.

PS - I saw one of your recent flights on the Infinity on a Windy video, and it was FINE. Clearly you are a champion, Brett. Accept it.

[photo not recovered: 23926.jpg]

"Every formula which expresses a law of nature is a hymn of praise to God." -Maria Mitchell

I believe in 17 foot turn radius. This is interesting information.

Perhaps one hard point in the f2b program is the time when you come down from hour glass maneuvre and try to make that 3 foot turn radius...

Keith is doing wonderful work on this!

Jukka

I remain very interested in this subject.

I was going to mention that although one could theoretically take a 3-axis accelerometer, and integrate the force measurement (F = ma) to get velocity, and integrate that to get position, as a practical matter, this is difficult. Yes, they have inertial guidance systems for ICBMs and perhaps some airplanes, but have you seen the units? They are huge and wildly expensive..

The problem is noise, which integrates like crazy itself..

Knowing all this, and knowing how fine some of the accelerometers are nowadays and how fast and accurate some A/D converters are, it would still be interesting to try, wouldn't it? But I seriously doubt that it is workable in any practical package.

It seems to me that the coordinated video cameras doing something like the golf ball tracker you mentioned are an obvious practical way to do it, and perhaps could be implemented for a few thousand bucks.

I've wondered about a very bright white LED on the model, blinking at about 10/second, and film it with a camera, stopped WAY down, fairly wide angle lens, on a tripod, with the shutter held open for multiple seconds. Would that provide a useable "track"?

When you solve it, let us all know how it's done!

L.

"Unless a man undertakes more than he possibly can do, he will never do all he can do." -Henry Drummond

[photo not recovered: 23926.jpg]

>I remain very interested in this subject.
>
>I was going to mention that although one could theoretically
>take a 3-axis accelerometer, and integrate the force
>measurement (F = ma) to get velocity, and integrate that to
>get position, as a practical matter, this is difficult. Yes,
>they have inertial guidance systems for ICBMs and perhaps some
>airplanes, but have you seen the units? They are huge and
>wildly expensive..
>
>The problem is noise, which integrates like crazy itself..
>
>Knowing all this, and knowing how fine some of the
>accelerometers are nowadays and how fast and accurate some A/D
>converters are, it would still be interesting to try, wouldn't
>it? But I seriously doubt that it is workable in any practical
>package.


It may be matter of semantics, but what really kills it is *drift* and scale factor asymmetry, for a purely inertial system. Vibration and noise by definition have no net DC components. Where vibration and noise get you is when the scale factor is not symmetrical, so it integrates more on the + swing than the - swing. I think you will also need a gyro package to make sense of the accelerometers - and they have the same problems plus the added problem of G-sensitive drift.

What you need to get around this is some external "observer" that will give you "truth" position as some relatively low rate, and use that to tweak the the gyro/accel measurements.

I am not really and expert on the consumer/low-dollar/microsize end of the sensor spectrum but even the very best existing inertial references would be very hard-pressed to provide the kind of accuracy needed for keeping track of a stunt plane for 8 minutes.

External sensors (high-def video cameras and lots of them) would be a lot more practical, but you need some sort of live version of Keith's tracking method. His is already probably accurate enough if you added sufficient fixed references and multiple angles but you would need to do it "on the fly' instead of manually frame-by-frame. What's missing in Keith's method is a way to correct for "off-angle" maneuvers.

Brett

I think it could be done with two or three cameras.

And I believe you could use infra-red imaging and daylight filters on the video that Larry suggested with the blinking LED.

Keith:

Is it possible to now enter a lrger turn radiuss as the default turn into CL-Sim? I don't recall my copy having that ability.

I use video at least a couple times a season for self coaching and being able to get a reference picture from the Sim would be a great help.

Curt

On White Sands and other missile ranges, they use a tracking system consisting of cinetheodolites, essentially high speed cameras and telescopes. Their locations are precisely surveyed, the range provides synchronized timing to .1 ms or better. Each frame of film records pointing (azimuth and elevation) and time information. Post mission, the film is all "read" on a machine which digitizes the pointing data, and adds fine pointing information called "boresight", which the operator provides by positioning crosshairs on the image. All this data is fed into a multi source triangulation solution, which yields position accuracies to a meter or less (impressive considering what is being tracked). The more "cines" involved, the better the solution; low elevation angles reduce accuracy, etc.

The range's high quality position data is used to calculate velocity and acceleration, which is also extremely accurate (derivatives versus integrals).

I believe a similar concept could be implemented on a small scale using 3 or more video cameras, which starts to sound very much like the golf ball tracker system already described.

Things are possible, for certain. Whether they are practical or not is a different question.

[photo not recovered: 23926.jpg]

"The wise see knowledge and action as one; they see truly." -Bhagava Gita

>I believe a similar concept could be implemented on a small
>scale using 3 or more video cameras, which starts to sound
>very much like the golf ball tracker system already
>described.

That would have the possibility of working (although cost-prohibitive), with the one remaining problem being how you map the optical image to determine the CG of the airplane VS some other point. It might at first seem silly but the if your processing tracks a wingtip or tail instead of the CG it's an error on the order of a few feet.

Of course tracking cameras are out of the question - you need passive (non-moving) cameras with enough coverage to get any possible variation on where the maneuvers are positioned. There's no way to track something with this kind of angular acceleration with something like a Nunn-Baker camera.

Brett

We used small, retroreflective spheres about .75" dia affixed to the object for a target. The area was flooded (each camera location had a IR flood) with IR and the cameras tracked the IR image.

Of course ours was setup indoors--never tried the IR flood thing in daylight.

Curt

> That would have the possibility of working (although
>cost-prohibitive), with the one remaining problem being how
>you map the optical image to determine the CG of the airplane
>VS some other point. It might at first seem silly but the if
>your processing tracks a wingtip or tail instead of the CG
>it's an error on the order of a few feet.
>

Exactly right. The cinetheodolite data gets boresight correction, with a HUMAN looking at the image and deciding where to put the crosshairs. Although I briefly worked on a proposal for a system to locate the centroid of a "plume" in a video frame using histograms for vertical and horizontal "stripes" on the image, again I think it is impractical.

> Of course tracking cameras are out of the question - you
>need passive (non-moving) cameras with enough coverage to get
>any possible variation on where the maneuvers are positioned.
>There's no way to track something with this kind of angular
>acceleration with something like a Nunn-Baker camera.
>

Also true, look at the difficulties for even an experienced operator to track a flight with a video camera!

However, I *do* know how to do it, using one of my yet to be patent-applied-for techniques. I could describe it to you, but I would have to kill you.. Suffice it to say that the first sentence of your last paragraph relates directly to my secret method.

> Brett

[photo not recovered: 23926.jpg]

"In the republic of mediocrity, genius is dangerous." -Robert G. Ingersoll

>However, I *do* know how to do it, using one of my yet to be
>patent-applied-for techniques. I could describe it to you, but
>I would have to kill you..

So, it's a win-win for everybody!

This sort of "pattern recognition/centroid of object in a passive CCD field" is a fairly well developed science but of course you are correct about it's sensitivity.

Brett

>Are you trying to divert this discussion to some subject of
>"conspiracy"?

Uh, no. Are you?

>Sometimes you remind me of a dog who's been kicked so many
>times he flinches and tries to bite any time anyone walks by.

What the heck are you talking about?
My point was that adding technology isn't going change human nature or the nature of the event.

<< snip strange rant about something unrelated to my post??>>

>Another clue for you. In spite of your suspicions, I've never
>heard ANYONE seriously suggest that with different judges or
>scoring methods that the same anointed gods of Stunt wouldn't
>STILL win the contest, nor that there is anything inherently
>unfair about the best fliers winning the contests. To suggest
>otherwise is a distortion of reality.

Oh really? Aside from every two months here and every afternoon on the old UHP, you mean? And in 10 emails a week? Or for the last 20 some-odd years, like a drumbeat? Of course it's nonsense, but to pretend that there aren't people who think they are getting the shaft (and are willing to shout it to the rooftops) is absurd.

Search the archives for "Anatomy of a Team Trials", or the Genie letter (both of which you definitely *have* seen) and then come back and tell me you never heard of such a thing. Nothing much has changed since then.

>
>We're not all after you Brett, nor are we even mildly jealous
>of your ability, we don't begrudge you of anything; please
>realize that every discussion is not about some conspiracy
>theory.

After me? How would that work again ;-)? It never crossed my mind. My biggest competition issue is my own failings, not external forces! I'm not too worried about other people beating me, only me beating myself. Bear in mind, I have *lost* more contests at all levels than most people have participated in!

My post was mostly directed towards your assertion that no one, like the "Lords of Stunt" or what ever silly characterization you want to use, would accept or believe the evidence of measurement. My point was that everybody *already knows about how close we come* and it has been well-understood since the 50's.

I wrote an entire Design column about how we didn't come anywhere close to 5 feet just last year, complete with full derivation. It's been measured time and again. If some people took offense to or didn't believe your measurements then they don't really understand the physics. I don't doubt it at all. I just assert that it doesn't really matter.

Even if you could measure the aircraft CG to 1/2" for the entire flight, it's not going to consequentially change the results, the event, or the human dynamics behind it. We'd just find something else to argue about.


Finally, a meta-comment - I'm really surprised, Larry. Usually I expect more subtlety (which could be a lot more effective) from someone of your experience. Very ham-handed.
Try these memes next: "Brett is in al Qaeda", "Brett is Fancher's Puppet", "Brett does business with Michael Vick" or something like that. Might take, you never know!

Brett

You need to turn loose of that paranoia, Brett.

PS - I agree with your comments about inertial systems.

[photo not recovered: 23926.jpg]

"If you see a whole thing - it seems that it's always beautiful. Planets, lives.. But close up a world's all dirt and rocks. And day to day, life's a hard job, you get tired, you lose the pattern." -Ursula K. LeGuin

[photo not recovered: 25244.jpg]

Actually, it's not that the corners are 17 feet rather than 5 that surprises me - I'd known that (and it's been mentioned here) a long time ago. What surprises me is that the 17 foot corners, and the maneuver, look so LOOSE. That's no hourglass, that's a snowman. And that's no square, that's a round loop with slightly flattened sides:

[photo not recovered: 25258.jpg]

Yet, when I watch videos of Ted flying squares, they do look square (if it weren't for the parallax, of course). Don't know what it is - maybe the fact that it's moving fools the mind.

Hi Isky,

You are dead right. If you watch the video, it looks much better. In fact, at the workshop, I used an avi file of this same outside square of Ted's before I showed the traced pattern, and most folks wanted to see the movie again after showing it. The best comment came from our top-class French flier, Serge Delabarde. You may not know, but Serge has quit flying serious competition now and has volunteered to judge. He was judging in Belgrade and also attended the workshop. The hourglass that I showed here is from Serge, and he quickly commented. "Aha, now you all know why I gave up flying competitions..........it is because my flying is so bad!!"

It is hard to retain a complete flight path in memory. Especially something like the 4-Leaf clover with 4 different segments that are spread out over such a big section of the hemisphere. It's also amazing to me that you spot so much more once you have a better idea of the actual shape and how it should appear from outside the sphere. It's just like those pictures that they use in those motivational lectures. Most times we simply do not see what the actual image is until someone tells us, and then it's so blatantly obvious that we could kick ourselves. Most of us only see what we want to see. Interesting stuff!

>You need to turn loose of that paranoia, Brett.
>

I think you are projecting, again.

Brett

>Yup, so much for 5ft. corners!! I have plenty of traced
>squares as well. They all look similar, and the corners are
>usually around 16 ft. radius. I've added a picture of a square
>that scored an average of 8 points at a recent world champs.
>It sure looked quite good in real time, but when you compare
>it to the rule-book FAI square, then what sort of score would
>you give it??
>
>Judges training with reference to the actual shapes will
>sharpen their skills, an we will see an initial lowering of
>scores in general. If we can assure the pilots that they can
>indeed perform the maneuver shapes to the rules, and not to
>any particular whims and fancies of judges, then the
>competitive stunt will be improved.
>
>[photo not recovered: 25258.jpg]
>

I would like to ask who is flying in this picture? Should
be possible to know?

Could you Keith send more of these plotted patterns?

Thank you,

Jukka Juslin from Finland

The square outside loop is by Ted Fancher, and like I said, it looks pretty good in real time. No model can turn the corner like the rule book version. The model skids and the actual path is delayed by the weight and inertia. Brett calculated that we could not get under around 13 ft. radius with our present model weights and speeds. I was told in Belgrade that the Yatsenko models turn so tight that they must be under at least 10 ft. or 3 metres. I've attached a picture below that shows Yuri Yatsenko's standard "Knight" model flown by Yuri himself. The model is around 1.3 metres long so you can work out for yourself, that this corner radius is over 5 metres.

[photo not recovered: 25263.jpg]

In the next shot, I've superimposed a 16 ft radius square in green, a FAI square in orange, and also a section of the hemisphere grid. Now you can see the total perspective from the camera's perspective. Most of the European competitors fly the tops at more than 45 degrees. The round loops are the same, and then when it comes to the vertical eights the sizes come down. In fact the vertical eight is usually the maneuver done closest to the rules. Here is Yuri's first qualifying round vertical eight from Belgrade.

[photo not recovered: 25265.jpg]

This is very well done, and as you can see here, I've traced both of the vertical eights. They are really well superimposed and the 45 degree intersection is about right as well. I think that we generally fly this better because we have the reference of the middle of the two loops, and the top loop dead overhead.

Attachment #1, (jpg file)
Attachment #2, (jpg file)

>The square outside loop is by Ted Fancher, and like I said,
>it looks pretty good in real time. No model can turn the
>corner like the rule book version. The model skids and the
>actual path is delayed by the weight and inertia. Brett
>calculated that we could not get under around 13 ft. radius
>with our present model weights and speeds.

Correct, that's one obvious limiting factor on the technical capability of the airplane. But almost the next sentence pointed out that no one is getting anywhere close to the technical capability of the airplane and we are all backing WELL off the minimum because we can't control it at 13 ft radii.

Nothing in your tracking is very surprising to me. I REALLY like the tracking system! It's the first time I have seen a good illustration of what I see when I judge or coach.

Brett

These graphics are super! Thanks for posting this material.

[photo not recovered: 23926.jpg]

"Society produces rogues, and education makes one rogue cleverer than another." -Oscar Wilde

Thanks guys,
There's another similar thread on the "other side" (Stunt Hanger)by Peter Nevai. He describes using a belt-mounted PC with a Heads-up-display system to project a grid for the pilot. It sounds practical, but as we all know too well, there are more practical things that come into play when we actually get down to trying to make it all work. I just need to find a faster way of getting feedback than manually tracking the video footage. Like I said before.....it's all very interesting!

Just a thought. Perhaps the reason for the "illusion" is the 7 degree nose-up attitude Igor's photograph (as seen elsewhere) shows. The airplane would achieve the correct attitude for the exit of the corner some point in time before the corner ends. In other words, the airplane is "sinking towards its belly" in the last few degrees of the turn. To our eyes, it would look as though the turn was completed way before it actually was.

If this skidding helps the illusion, perhaps the thing to do would be to decrease flap movement so that you get more of a nose-up attitude during a corner. A flapless plane might even look as though it were turning sharper even though the actual turn is larger.

I think so ... it is far easier to watch heading than real path of CG. We spoke about it in Belgrade also. Keith will remember - it was question of wind influence to judging ... judge must really expolore ONLY path of the model and not angle of its fuselage (trying to deal with wind), or smoke path or whatever ... and yes it is DIFFICUL and it needs TRAINING. So I really welcome what Keith do. Judging needs training same way like flyers do, judges need tools and judges need know-how.

[photo not recovered: angle.jpg]