Forums: Skydiving Disciplines: Wing Suit Flying:
Wingsuit Formation Analyzing Software

 

First page Previous page 1 2 3 4 5 Next page Last page  View All

kallend  (D 23151)

Apr 1, 2010, 11:42 AM
Post #51 of 112 (1106 views)
Shortcut
Re: [tvandijck] Wingsuit Formation Analyzing Software [In reply to] Can't Post

Tom, I'm having a really hard time logging in to your "mantis" site, and the emails it claims to send me have not arrived.


SuperGirl  (D 30091)

Apr 1, 2010, 12:03 PM
Post #52 of 112 (1099 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
Tom, I'm having a really hard time logging in to your "mantis" site, and the emails it claims to send me have not arrived.

works fine for me. maybe the email never made it past your spamblocker... if you used your university email to sign up, your department might have some fancy virus scan in place that likes to occasionally give too many false positives... I know my dept email sometimes has that issue...


(This post was edited by SuperGirl on Apr 1, 2010, 12:11 PM)


SuperGirl  (D 30091)

Apr 1, 2010, 12:09 PM
Post #53 of 112 (1097 views)
Shortcut
Re: [tvandijck] Wingsuit Formation Analyzing Software [In reply to] Can't Post

Tom,
I was just taking a look at the Mantis bugtracker you set up...
To report an issue, I have to assign it to a "project", but no projects are listed for me to choose from, and as a "reporter" I do not have the privileges to create any new projects.

-a

Edit: issue fixed. thank you, Tom.


(This post was edited by SuperGirl on Apr 1, 2010, 1:31 PM)


kallend  (D 23151)

Apr 1, 2010, 12:28 PM
Post #54 of 112 (1088 views)
Shortcut
Re: [SuperGirl] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
In reply to:
Tom, I'm having a really hard time logging in to your "mantis" site, and the emails it claims to send me have not arrived.

works fine for me. maybe the email never made it past your spamblocker... if you used your university email to sign up, your department might have some fancy virus scan in place that likes to occasionally give too many false positives... I know my dept email sometimes has that issue...

I looked in there, it's all real spam (like offers of $10M from Nigeria, Viagra ads, etc.).


Birdman_Mike  (D 25676)

Apr 1, 2010, 12:32 PM
Post #55 of 112 (1086 views)
Shortcut
Re: [mccordia] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
There's also several windows shell options for mac available. Not meaning more work on this side for now. We dont have funds, all time spent on this is 'for the love of the game' so macs arent the first priority. But once we get to a more complete version, Tom said mac versions are also a definate possibility..

How about making a version for the iPad? Mark the heads and draw the lines on the touch screen. Of course, the first iPads are just coming out this weekend. So, maybe this idea is a bit premature Cool

Purple Mike


SuperGirl  (D 30091)

Apr 1, 2010, 12:36 PM
Post #56 of 112 (1082 views)
Shortcut
Re: [Birdman_Mike] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
How about making a version for the iPad? Mark the heads and draw the lines on the touch screen. Of course, the first iPads are just coming out this weekend. So, maybe this idea is a bit premature Cool

Purple Mike

a version that will run in a web browser should run fine on the iPad... especially if it's javascript like we talked about...


tvandijck  (D 30374)

Apr 2, 2010, 8:46 AM
Post #57 of 112 (1032 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
Tom, I'm having a really hard time logging in to your "mantis" site, and the emails it claims to send me have not arrived.

does it still not work? I added you as reporter, the same way I added Andrea, and she just added a first report.

As for the emails it claims to send, you are right, I don't get those either.. I think my hosting provider actually blocks the server from sending anything.

that said though, the emails are pretty useless anyway.

Anyway, after someone signs up, I have to assign that person to a project. As this is not the only project I work on, so there is more stuff in there then I want visible for everyone.. So you cannot start right away.
I do however send out email afterwards manually to confirm your registration. I hope you did receive that at least.


SuperGirl  (D 30091)

Apr 2, 2010, 8:57 AM
Post #58 of 112 (1027 views)
Shortcut
Re: [tvandijck] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
I think my hosting provider actually blocks the server from sending anything.

nope, it's not that, cause I got all the emails.


kallend  (D 23151)

Apr 2, 2010, 9:44 AM
Post #59 of 112 (1014 views)
Shortcut
Re: [SuperGirl] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
In reply to:
I think my hosting provider actually blocks the server from sending anything.

nope, it's not that, cause I got all the emails.

Yes, but you have "Super" status, SupergirlWink.


SuperGirl  (D 30091)

Apr 2, 2010, 9:45 AM
Post #60 of 112 (1008 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
In reply to:
In reply to:
I think my hosting provider actually blocks the server from sending anything.

nope, it's not that, cause I got all the emails.

Yes, but you have "Super" status, SupergirlWink.

mommy always told me I was "special" Tongue


kallend  (D 23151)

Apr 5, 2010, 9:09 PM
Post #61 of 112 (976 views)
Shortcut
Re: [SuperGirl] Wingsuit Formation Analyzing Software [In reply to] Can't Post

Here's an 8-way. (1) is using the angle/line method, (2) is using my method***. Both show the same problem, the fliers on the left side are clearly way too close to the centerline. In fact, the software couldn't figure out which jumper went with which dot for one of the jumpers.

*** because Tom's software hasn't fully implemented my method yet, I had to use the "free rotation" feature to align the markers with the "dots", BUT as the software doesn't rotate the background picture along with the markers, I removed that, just leaving the markers and dots in place.
Attachments: 8-way1.jpg (63.6 KB)
  8-way2.jpg (46.5 KB)


tvandijck  (D 30374)

Apr 5, 2010, 10:11 PM
Post #62 of 112 (967 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
Here's a nice 16-way from the Elsinore warm-up jumps.

New version available online, same link as the opening post. I used it to quickly get the attached result on the elsinore 16-way.

or quick link: http://www.tomvandijck.com/flock/

changes:
* added stretch, and rotate to 'kallend' method.
* scaled markers with wingsuits.
* made wingsuits transparent.
* added quick edits for angle/distance tolerances.
* added visuals to show whether error was due to angle or distance.
* couple bugs here and there.

enjoy.


(This post was edited by tvandijck on Apr 5, 2010, 10:20 PM)
Attachments: elsinore16_scored.jpg (19.8 KB)


tvandijck  (D 30374)

Apr 5, 2010, 10:14 PM
Post #63 of 112 (964 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
Here's an 8-way.

nice screenshots ;)

If you click on "Project / Export..." you can export the image with the markers, lines and dots. without getting the full UI in your picture...


kallend  (D 23151)

Apr 6, 2010, 5:43 AM
Post #64 of 112 (939 views)
Shortcut
Re: [tvandijck] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
In reply to:
Here's an 8-way.

nice screenshots ;)
.

I wanted to show the positions of the sliders. Does "export" show any of the tolerance values?


Premier DSE  (D 29060)

Apr 6, 2010, 7:19 AM
Post #65 of 112 (918 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

No, export doesn't show values (although I suppose they could be exported as EXIF data).
Personally, I like the way you're taking a screenshot rather than exporting. It certainly is possible to overlay that data on the final image, maybe that's a feature to consider.
Tom, what about adding a data stripe to the exported image so that settings are immediately visible?


kallend  (D 23151)

Apr 6, 2010, 7:37 AM
Post #66 of 112 (914 views)
Shortcut
Re: [tvandijck] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
In reply to:
Here's a nice 16-way from the Elsinore warm-up jumps.

New version available online, same link as the opening post. I used it to quickly get the attached result on the elsinore 16-way.

or quick link: http://www.tomvandijck.com/flock/

changes:
* added stretch, and rotate to 'kallend' method.
* scaled markers with wingsuits.
* made wingsuits transparent.
* added quick edits for angle/distance tolerances.
* added visuals to show whether error was due to angle or distance.
* couple bugs here and there.

enjoy.

I just tried the latest version on the 8-way I posted a couple of posts ago. A definite improvement, Tom. Certainly faster than doing it by hand!

No way you can get all the heads in the green dots by rotating or stretching unless you allow the dots to be HUGE.

That formation passed the grid test, but doesn't pass either the line/angle test or the "Kallend" test without huge tolerances being allowed.


kallend  (D 23151)

Apr 6, 2010, 1:04 PM
Post #67 of 112 (871 views)
Shortcut
Re: [SuperGirl] Wingsuit Formation Analyzing Software [In reply to] Can't Post

 
OK, some food for thought.

Here are 3 pictures of the same HYPOTHETICAL 64-way diamond supposedly with 90 degree angles at the corners.

1. The formation - pretty damn ugly IMO.

2. The lines and angles analysis with tolerances as shown (these are smaller than the default values). No red lines. Is this an acceptable formation as the method suggests?

3. The "kallend" analysis with stretch and rotation as shown according to the method, and a generous tolerance (around 40%) on position. (I estimated the tolerance by eye right now as I'm not sure the "Size" listed by the program is the % tolerance). Around 1/2 the "jumpers" are out of position, some are way out. Is this an unacceptable formation as the method suggests?

Edited to add - Tom has done a really nice job putting this together and making it available for testing.


(This post was edited by kallend on Apr 6, 2010, 1:16 PM)
Attachments: 64way-1.jpg (85.8 KB)
  64way-lines angles.jpg (90.8 KB)
  64way-jk.jpg (91.9 KB)


tvandijck  (D 30374)

Apr 6, 2010, 5:32 PM
Post #68 of 112 (841 views)
Shortcut
Re: [DSE] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
Tom, what about adding a data stripe to the exported image so that settings are immediately visible?

I got a feature request for that in mantis indeed... just haven't gotten around to adding it... I'll try it somewhere next weekend if the weather sucks again ;)


kallend  (D 23151)

Apr 7, 2010, 6:53 AM
Post #69 of 112 (811 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
Here's an 8-way. (1) is using the angle/line method, (2) is using my method***. Both show the same problem, the fliers on the left side are clearly way too close to the centerline. In fact, the software couldn't figure out which jumper went with which dot for one of the jumpers.

*** because Tom's software hasn't fully implemented my method yet, I had to use the "free rotation" feature to align the markers with the "dots", BUT as the software doesn't rotate the background picture along with the markers, I removed that, just leaving the markers and dots in place.

Here is the 8-way re-analyzed using the latest version of Tom's program. A large amount of lateral compression ("Stretch" number = 82) is needed to be able to get everyone in green dot.

We do not at this time have a consensus on whether formations that stretch or compress along one dimension compared to the original design should be allowed, and if so, what limits on stretch or compression should be imposed. The grid allows none, the angle/line method implicitly allows it on account of the tolerance on the angle, and in the kallend dot method it is optional. Comments, anyone?
Attachments: 8-way3.jpg (83.2 KB)


kallend  (D 23151)

Apr 7, 2010, 9:43 AM
Post #70 of 112 (787 views)
Shortcut
Re: [tvandijck] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
In reply to:
Here's a nice 16-way from the Elsinore warm-up jumps.

New version available online, same link as the opening post. I used it to quickly get the attached result on the elsinore 16-way.

or quick link: http://www.tomvandijck.com/flock/

changes:
* added stretch, and rotate to 'kallend' method.
* scaled markers with wingsuits.
* made wingsuits transparent.
* added quick edits for angle/distance tolerances.
* added visuals to show whether error was due to angle or distance.
* couple bugs here and there.

enjoy.

Tom, those are very small tolerance dots you've put on that 16-way.

How hard will it be to display the dot radius as a % of the inter-jumper spacing? The numbers currently shown are fine for making adjustments to the picture but it's not clear what they mean.

Also, I'm unclear exactly what the "Stretch" number means. Is it the the width scale as a % of the length scale, or something else?


tvandijck  (D 30374)

Apr 8, 2010, 12:05 AM
Post #71 of 112 (738 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
Tom, those are very small tolerance dots you've put on that 16-way.
yes, I liked how small I could get them, with just showing 2 people outside of the dots.. it's in no way to mean those 2 guys did a poor job or anything.. I'm not in any position to make that call.... it was just an example..

In reply to:
How hard will it be to display the dot radius as a % of the inter-jumper spacing? The numbers currently shown are fine for making adjustments to the picture but it's not clear what they mean.

not sure, I'd have to try a couple things... shouldn't be too hard, but the hard part is making numbers meaning full to begin with.. the numbers right now are pretty absolute numbers in pixels.. so dotsize == 200 just says the dots have radius of 200 pixels. but that is somewhat 'random' since photo resolution is then a big factor.. same count for distance, and most other numbers.


In reply to:
Also, I'm unclear exactly what the "Stretch" number means. Is it the the width scale as a % of the length scale, or something else?

those are indeed percentages... so
50% is width is 50% of height.
200% is width is 200% of height.
pretty simple... it's one of the few numbers that actually is relative, and thus not subjective to resolution.


tvandijck  (D 30374)

Apr 8, 2010, 12:18 AM
Post #72 of 112 (736 views)
Shortcut
Re: [kallend] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
Comments, anyone?

Just looking at the 16 way I tried, the stretch there is at 75% orso, which is thus an even bigger deviation from what you did.

Personally, I'd say totally allow that.. the formation looks uniform, is exciting to look at, and even with very low tolerances I can get all dots green.

I'd take more issue with non-uniform formations. things that look totally off, asymmetrical, etc. That 16 way, even though it's compressed that much.... totally awesome..


kallend  (D 23151)

Apr 8, 2010, 6:52 AM
Post #73 of 112 (712 views)
Shortcut
Re: [tvandijck] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
In reply to:
Tom, those are very small tolerance dots you've put on that 16-way.
yes, I liked how small I could get them, with just showing 2 people outside of the dots.. it's in no way to mean those 2 guys did a poor job or anything.. I'm not in any position to make that call.... it was just an example..

In reply to:
How hard will it be to display the dot radius as a % of the inter-jumper spacing? The numbers currently shown are fine for making adjustments to the picture but it's not clear what they mean.

not sure, I'd have to try a couple things... shouldn't be too hard, but the hard part is making numbers meaning full to begin with.. the numbers right now are pretty absolute numbers in pixels.. so dotsize == 200 just says the dots have radius of 200 pixels. but that is somewhat 'random' since photo resolution is then a big factor.. same count for distance, and most other numbers.

Well, if the distance between adjacent dots is in pixels, and the dot radius is in pixels, can't you just work from there?


In reply to:

In reply to:
Also, I'm unclear exactly what the "Stretch" number means. Is it the the width scale as a % of the length scale, or something else?

those are indeed percentages... so
50% is width is 50% of height.
200% is width is 200% of height.
pretty simple... it's one of the few numbers that actually is relative, and thus not subjective to resolution.

Thanks. So a reading of 75 means it's been squashed by (100-75) = 25% in the width direction.


Costyn  (D 92506)

Apr 14, 2010, 12:23 AM
Post #74 of 112 (666 views)
Shortcut
Re: [repcool] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
It also costs a virtual solution and windows license so it isn't an option for everyone.

I'm also a Mac user and I can higly recommend VirtualBox http://www.virtualbox.org/ as an free alternative to VMWare Fusion. Of course you still need a Windows disk...


tvandijck  (D 30374)

Apr 14, 2010, 12:35 AM
Post #75 of 112 (663 views)
Shortcut
Re: [Costyn] Wingsuit Formation Analyzing Software [In reply to] Can't Post

In reply to:
In reply to:
It also costs a virtual solution and windows license so it isn't an option for everyone.

I'm also a Mac user and I can higly recommend VirtualBox http://www.virtualbox.org/ as an free alternative to VMWare Fusion. Of course you still need a Windows disk...

Actually, the application is written in .NET, and is apparently Mono compliant... Mono can be downloaded for Linux, FreeBSD and MacOSX.. so it should be possible to run the application on all of those platforms. I'll may have to make some changes here and there, but in theory there should be no real issues.

I tried it on my old G4 Mac, but that didn't really work very well, but if anyone has a newer intel mac and wants to give it a try, I'd welcome the feedback.

you can get mono here http://www.go-mono.com


First page Previous page 1 2 3 4 5 Next page Last page  View All

Forums : Skydiving Disciplines : Wing Suit Flying

 


Search for (options)