Blender Test #3

[Tim Wescott]

I really need to get out to the shop.

Thanks to Tim Just for inspiring this photo.  Too bad I edited him out.

[Ken Culbertson]

Tim, this much easier to watch.  What corner radius were you using and how far back is the "camera" from the edge of the circle.  Keep it up!

Ken

[john e. holliday]

Good thing I'm not judging this as the bottoms of maneuver were way lower than level flight and the hour glass is much too wide.   The corners don't look right the way they snap through a turn.

[Paul Walker]

This is clearly not a Tim flight. Somebody's simulator.

[Ken Culbertson]

Good thing I'm not judging this as the bottoms of maneuver were way lower than level flight and the hour glass is much too wide.   The corners don't look right the way they snap through a turn.

I didn't notice the bottoms, you are right but not all of them.  It appears to be only the rounds. I did watch in slow motion and the corners are, at best about 1 plane length.  Most modern planes would fold in a corner that tight at full speed.  The hourglass is too wide for the corners we fly.  I measured the triangle and the hourglass, and the bases are the same.  This would only be true if you had 0' corners, which he appears to have.  The top flat of the hourglass is totally dependent on your corner size and at a 0' corner it really is that long.  What makes the top appear smaller is the larger corner.  Just imagine being under the old rules ans losing points for corners that were too tight. 

Still, this is the best I have seen, and it will get better! 

Ken

[Arlan McKee]

I kept trying to turn the volume up.

[Brett Buck]

Good thing I'm not judging this as the bottoms of maneuver were way lower than level flight and the hour glass is much too wide.   The corners don't look right the way they snap through a turn.

   It's a test, guys!  The other thing that looks unusual is that the airplane looks to be about 50% bigger than most stunt, that's why it looks so cramped on round maneuvers. It's a test, there will be other tests, too.

     Brett

[Motorman]

Can you do one for Old Time?

[Howard Rush]

Very cool.  Even as is, it shows you the illusion to shoot for.

[EricV]

   It's a test, guys!  The other thing that looks unusual is that the airplane looks to be about 50% bigger than most stunt, that's why it looks so cramped on round maneuvers. It's a test, there will be other tests, too.

     Brett

Maybe its modeled after Windys Sweeper...
EricV

[Steve Helmick]

While I liked the smog in the background, the color scheme of the airplane made it a bit difficult to see, and I missed the sweet sound of an internal combustion engine. The hourglass was definitely too wide for my taste. IMO, the smog, color scheme and lack of definition of the far side of the circle makes me undecided about the bottoms and level flight altitude.  Steve

[Brett Buck]

Maybe its modeled after Windys Sweeper...
EricV

   And that is exactly how it looked!

     Brett

[Dennis Moritz]

Orestes? Joking

[Tim Wescott]

Good thing I'm not judging this as the bottoms of maneuver were way lower than level flight and the hour glass is much too wide.   The corners don't look right the way they snap through a turn.

Yes, the hour glass sucks -- as do all of the square maneuvers.

I'll check the bottoms, the code is generating them at 1.5 meters (5 feet), but I may be messing that up with the video rendering.

Tim, this much easier to watch.  What corner radius were you using and how far back is the "camera" from the edge of the circle.  Keep it up!

The camera is about 80 or 100 feet back from the edge of the circle, to capture the whole flight and not be horribly distorted.  Corner radius is way wrong -- it's 0.3 meters.  But, setting it much bigger makes it too swoopy.

Keith Renicle said something about needing to start the plane turning before it's center of gravity actually goes around the corner, or the corner looks looser than it really is.  So I need to try that (I'm not looking forward to that).

... The other thing that looks unusual is that the airplane looks to be about 50% bigger than most stunt, that's why it looks so cramped on round maneuvers. ...

I'll check that, but it should be a hair less than 60" span and 45" long.

I was assuming that it looked cramped from a combination of exactly 45 degree maneuvers, and the camera being back far enough to catch the whole scene.  But it's something to double-check.

The airplane is made out of four distorted "spheres", because that was easy.  The paint scheme is white and blue stripes, because that was easy, too -- I was going for something that could be seen, and it turns out that getting Blender to render realistically is like pulling teeth.

While I liked the smog in the background...

Hey, that's genuine Muncie humidity!  Not a bit of smog.

This is clearly not a Tim flight. Somebody's simulator.

Well, not Tim Just.  I snapped a picture of him in practice at Muncie last year, which was lucky because I needed a background picture that had been snapped at eye level to make a natural-looking flight.

[Tim Wescott]

I don't know if anyone else saw it, but the round eights look like the intersections are lazy -- even though the simulation is two circles that are exactly kissing each other.  You can see that it's right if you slow the video down.

I'm wondering if that's just a human perception thing, and also why everybody who's good enough for that to show seems to either over turn a bit or not quite hit the intersections.

[Steve Helmick]

There's something strange (to me) about live perception vs. video facts. I can't see a radius on a square corner when Paul or Howard do it, but if I was to watch a video of the same flight (by Paul), the radius would be easy to see. I'm not sure why, but I'm not sure I've ever seen a video of Howard flying. Anyway, this leads me to think there is definitely a humon (sic) perception problem in watching a pattern, whether live, video, or simulated.    Steve

[Ken Culbertson]

There's something strange (to me) about live perception vs. video facts. I can't see a radius on a square corner when Paul or Howard do it, but if I was to watch a video of the same flight (by Paul), the radius would be easy to see. I'm not sure why, but I'm not sure I've ever seen a video of Howard flying. Anyway, this leads me to think there is definitely a humon (sic) perception problem in watching a pattern, whether live, video, or simulated.    Steve

I do a lot of video and study it in slow motion.  The 8 intersections are a perfect example.  What you see if they are done perfectly looks wrong but if you cheat the intersection just a little by starting the switchover just before the plane becomes vertical, they look perfect.  Maybe that is just me, we all see tings differently.

Ken

[Brett Buck]

There's something strange (to me) about live perception vs. video facts. I can't see a radius on a square corner when Paul or Howard do it, but if I was to watch a video of the same flight (by Paul), the radius would be easy to see. I'm not sure why, but I'm not sure I've ever seen a video of Howard flying. Anyway, this leads me to think there is definitely a humon (sic) perception problem in watching a pattern, whether live, video, or simulated.    Steve

   Seeing it in person is a completely different experience than watching a projection of it, even in 8K. That's why all the incessant nit-picking of the reference frames, seeing it from the center of the circle or the upwind side, spherical projections, etc, is a complete waste of time - unless you plan on judging it from the projections.

     I am pretty sure that the AMA drawings are more-or-less OK as is, but trying to take that and express it mathematically does entail picking a reference frame and applying some rigor to the problem.

     The difference between doing it for real with people, and trying to do it some other way, is a critical matter for stunt, and I think it would be a terrible, bordering on fatal, mistake to try to cut the judges out of the system. They are better and make for a better event than any hypothetically perfect "objective" system. I point to virtually every other genuinely objective CL event as evidence.

     Brett

[Tim Wescott]

     The difference between doing it for real with people, and trying to do it some other way, is a critical matter for stunt, and I think it would be a terrible, bordering on fatal, mistake to try to cut the judges out of the system. They are better and make for a better event than any hypothetically perfect "objective" system. I point to virtually every other genuinely objective CL event as evidence.

     Brett

Yes, don't ever cut the human judges out of the equation.

Originally I started down this path as a training aid for beginning judges.  I already had some half-finished flight simulation software that I started for something else, so it was a low(ish) lift.

Also originally, I wanted to render the pattern for VR goggles, for someone standing in the judge's position.  I was under the misconception that you could turn your cell phone into a cheap VR goggle set with a $30 frame and an app.  It turns out that Google abandoned the app about five years ago, so if you want to equip a budding judge with VR goggles then someone is going to have to fork out $300 for the goggles, and have a decent gaming computer to drive them.

So I'm not going to render for VR goggles unless someone specifically asks.

   Seeing it in person is a completely different experience than watching a projection of it, even in 8K.

That's what had me on about VR goggles.  I think you could do it with a video wall, but that's a wee bit more expensive than VR goggles, and you'd have exactly one sweet spot from which to watch.

And yes -- even with VR goggles or a video wall, it still wouldn't be the same as being there in person.

[Steve Helmick]

"The difference between doing it for real with people, and trying to do it some other way, is a critical matter for stunt, and I think it would be a terrible, bordering on fatal, mistake to try to cut the judges out of the system. They are better and make for a better event than any hypothetically perfect "objective" system. I point to virtually every other genuinely objective CL event as evidence.

     Brett"

On a more practical side of it, any technology solution that eliminates judges probably isn't going to be 100% reliable, and then you need a backup plan available immediately. What happens when we come to rely on the techie solution and there's nobody who remembers how to watch stunt and write scores? Lately, I've only been judging at 3 contests a season, and it's clear to me that it takes awhile to grind the rust off.  Steve

[Doug Moon]

At first glance I was like that's Circle one Muncie. Cool. Two laps and I was like wait, WHERE'S the pilot?

 

Cool simulation. To me the colored wings made it looked folded on the down leg of the inside square 8.

Very cool the progression of this stuff over the years is awesome.
 

[Tim Wescott]

That picture was one I snapped at Muncie, and did have a pilot in it -- until I photoshopped it out.  If you look closely there's a fractured building in the distance where Tim Just's head used to be.

It turns out that putting paint jobs onto objects in blender in any sort of a controlled manner is really freaking hard.  I wanted to concentrate on getting the thing visible and into the (virtual) air quickly, so that's what you get.

What I did is the Blender equivalent of randomly switching between two rolls of Monocoat as you cover an airplane, just to get some color.