The Cyclone Jitters, or the large coordinate blues
- Scott.Warren
- V.I.P Member
- Posts: 364
- Joined: Tue Jan 30, 2018 3:14 pm
- 6
- Full Name: Scott Warren
- Company Details: WSP Canada
- Company Position Title: Project Manager - Geomatics Engineer
- Country: Canada
- Linkedin Profile: Yes
- Location: Dartmouth, Nova Scotia, Canada
- Has thanked: 44 times
- Been thanked: 111 times
- Contact:
The Cyclone Jitters, or the large coordinate blues
youtu.be/T7JcjL9nWaE
Anyone have a solution for the jumpy jittery dancing that happens when you use large coordinate systems in an LGS or even Cyclone for that matter.
This is something that happens with some regularity in Cyclone Core, and now I'm seeing it in an LGS. I can often resolve it by setting a new origin-Coordinate system so 0 is close to the site, but i don't like that work around.
Just seeing if anyone has any thoughts on cause or solutions.
The video has a COE on the same coordinate system as the Cloud, but it doesn't drop in the correct place, or even stay in the same location when moving the camera around.
Is this a floating point issue with the software?
- smacl
- Global Moderator
- Posts: 1409
- Joined: Tue Jan 25, 2011 5:12 pm
- 13
- Full Name: Shane MacLaughlin
- Company Details: Atlas Computers Ltd
- Company Position Title: Managing Director
- Country: Ireland
- Linkedin Profile: Yes
- Location: Ireland
- Has thanked: 627 times
- Been thanked: 657 times
- Contact:
Re: The Cyclone Jitters, or the large coordinate blues
Possibly, or with the input format used. Unstructured E57 suffers from resolution limits where large floating point numbers are used without first reducing them to a scaled integer base. From the libE57 documentation section on best practices
If you use raw floating point for large coordinates rather than scaled integers in E57 you hit serious resolution problems. I note the format of E57 supplied by Leica products changed some time back from Cartesian to Spherical, most likely as a workaround to this issue. I've seen similar problems in RCS files.The standard only allows doubles to be used for cartesian bounds coordinates; however point data should use scaled integers to be efficient. This mismatch will cause round off errors that will be found by the validate57.exe tool. To fix this problem the bounds need to be passed through the same calculations as a scaled integer.
- Scott.Warren
- V.I.P Member
- Posts: 364
- Joined: Tue Jan 30, 2018 3:14 pm
- 6
- Full Name: Scott Warren
- Company Details: WSP Canada
- Company Position Title: Project Manager - Geomatics Engineer
- Country: Canada
- Linkedin Profile: Yes
- Location: Dartmouth, Nova Scotia, Canada
- Has thanked: 44 times
- Been thanked: 111 times
- Contact:
Re: The Cyclone Jitters, or the large coordinate blues
For this i used 2 scanners (p20 and BLK360), the COE and LGS exports were done using the same Cyclone Modelspace, and both used the native provincial coordinate system system, there were no additional Coordinate systems saved for this one.
When the NS coordinate system was modernized recently, they tacked a 2 on the front of all the eastings, probably doesn't help here.
Sample coordinate sizes:
E=25573243.929
N=4947917.172
Not sure E57's have a part to play in the workflow I used here. I'd be surprised if E57's were lurking inside the LGS.
When the NS coordinate system was modernized recently, they tacked a 2 on the front of all the eastings, probably doesn't help here.
Sample coordinate sizes:
E=25573243.929
N=4947917.172
Not sure E57's have a part to play in the workflow I used here. I'd be surprised if E57's were lurking inside the LGS.
-
- I have made 30-40 posts
- Posts: 33
- Joined: Tue Apr 21, 2009 9:16 am
- 14
- Full Name: Bart Tylutki
- Company Details: KSE
- Company Position Title: HDS Manager
- Country: USA NJ
- Been thanked: 3 times
Re: The Cyclone Jitters, or the large coordinate blues
Yeah, it happens with large coordinate system in orthogonal view while looking at things from bottom side mostly. Walkaround for Cyclone is quite simple: set local coordinates system for working process then reset to default before export. After setting up temp local small coord system close window and reopen - it should be fine now.
View>coord system> set origin (at picked point)
Hope this helps.
View>coord system> set origin (at picked point)
Hope this helps.
- Scott.Warren
- V.I.P Member
- Posts: 364
- Joined: Tue Jan 30, 2018 3:14 pm
- 6
- Full Name: Scott Warren
- Company Details: WSP Canada
- Company Position Title: Project Manager - Geomatics Engineer
- Country: Canada
- Linkedin Profile: Yes
- Location: Dartmouth, Nova Scotia, Canada
- Has thanked: 44 times
- Been thanked: 111 times
- Contact:
Re: The Cyclone Jitters, or the large coordinate blues
That's what I tend to do when In Cyclone, I guess it means I have to manage a local system in each LGS if I want to add 3D Models .
I also noticed that its usually in Ortho view that it wobbles, not so much in perspective view.
I also noticed that its usually in Ortho view that it wobbles, not so much in perspective view.
-
- I have made 30-40 posts
- Posts: 33
- Joined: Tue Apr 21, 2009 9:16 am
- 14
- Full Name: Bart Tylutki
- Company Details: KSE
- Company Position Title: HDS Manager
- Country: USA NJ
- Been thanked: 3 times
Re: The Cyclone Jitters, or the large coordinate blues
This issue was in Cyclone probably since the beginning. When I brought it up to Leica attention a couple years ago, they didn't even know what I was talking about and blaming video card glitching. Funny
BTW, where on earth is this place (E=25573243.929 N=4947917.172) it's feet I assume?
I only go up to 2500000 and your numbers are 10x larger. That would explain such HUGE "jitter" from your video.
BTW, where on earth is this place (E=25573243.929 N=4947917.172) it's feet I assume?
I only go up to 2500000 and your numbers are 10x larger. That would explain such HUGE "jitter" from your video.
- gsisman
- V.I.P Member
- Posts: 898
- Joined: Fri Oct 07, 2016 1:51 pm
- 7
- Full Name: Steve Long
- Company Details: Montgomery County DOT _ MD
- Company Position Title: Land Survey Supervisor
- Country: United States
- Skype Name: gsisman1
- Linkedin Profile: Yes
- Has thanked: 767 times
- Been thanked: 149 times
Re: The Cyclone Jitters, or the large coordinate blues
Yeah, he said they tacked an extra 2 as in 20 million! That's a large coordinate! We use state plan all the time, but in my area we have none that high just single millionsbtylutki wrote: ↑Wed Apr 27, 2022 9:11 pm This issue was in Cyclone probably since the beginning. When I brought it up to Leica attention a couple years ago, they didn't even know what I was talking about and blaming video card glitching. Funny
BTW, where on earth is this place (E=25573243.929 N=4947917.172) it's feet I assume?
I only go up to 2500000 and your numbers are 10x larger. That would explain such HUGE "jitter" from your video.
- Scott.Warren
- V.I.P Member
- Posts: 364
- Joined: Tue Jan 30, 2018 3:14 pm
- 6
- Full Name: Scott Warren
- Company Details: WSP Canada
- Company Position Title: Project Manager - Geomatics Engineer
- Country: Canada
- Linkedin Profile: Yes
- Location: Dartmouth, Nova Scotia, Canada
- Has thanked: 44 times
- Been thanked: 111 times
- Contact:
Re: The Cyclone Jitters, or the large coordinate blues
That's a bingo.gsisman wrote: ↑Wed Apr 27, 2022 10:55 pmYeah, he said they tacked an extra 2 as in 20 million! That's a large coordinate! We use state plan all the time, but in my area we have none that high just single millionsbtylutki wrote: ↑Wed Apr 27, 2022 9:11 pm This issue was in Cyclone probably since the beginning. When I brought it up to Leica attention a couple years ago, they didn't even know what I was talking about and blaming video card glitching. Funny
BTW, where on earth is this place (E=25573243.929 N=4947917.172) it's feet I assume?
I only go up to 2500000 and your numbers are 10x larger. That would explain such HUGE "jitter" from your video.
https://geonova.novascotia.ca/moderniza ... technology
Nice part is you can tell at a glance which coordinate system its in, but tacking on the extra digit as a prefix seems to be adversely affecting Cyclone.
- Scott.Warren
- V.I.P Member
- Posts: 364
- Joined: Tue Jan 30, 2018 3:14 pm
- 6
- Full Name: Scott Warren
- Company Details: WSP Canada
- Company Position Title: Project Manager - Geomatics Engineer
- Country: Canada
- Linkedin Profile: Yes
- Location: Dartmouth, Nova Scotia, Canada
- Has thanked: 44 times
- Been thanked: 111 times
- Contact:
Re: The Cyclone Jitters, or the large coordinate blues
Naw homie, its in meters!btylutki wrote: ↑Wed Apr 27, 2022 9:11 pm This issue was in Cyclone probably since the beginning. When I brought it up to Leica attention a couple years ago, they didn't even know what I was talking about and blaming video card glitching. Funny
BTW, where on earth is this place (E=25573243.929 N=4947917.172) it's feet I assume?
I only go up to 2500000 and your numbers are 10x larger. That would explain such HUGE "jitter" from your video.
https://epsg.io/8083
New Scotland is a weird and wonderful place.
- Phill
- V.I.P Member
- Posts: 653
- Joined: Wed Feb 06, 2008 7:19 am
- 16
- Full Name: Phillip Nixon
- Company Details: SKM
- Company Position Title: Surveyor
- Country: Australia
- Location: Sydney
- Has thanked: 7 times
- Been thanked: 45 times
Re: The Cyclone Jitters, or the large coordinate blues
I think its to do with cross multiplying numbers. There are only so many digits a computer number can be and once thats exceeded the computer just says E### and the rest just get assigned zeros. I think its called a floating point overflow.
Without doing a giant example say a variable has a maximum number of digits of 4 (They are much larger than 4) and you give it the number 1234567, it stores it as 1234e3 so when it is used in math the number is 1234000. When doing big coordinate math you might do 1 or more cross multiply of those numbers and it float overflows so when it get back to making a move on screen its jumpy cause it lost precision while doing the math. When you make a local system the math is smaller (ie 100x100 is a smaller number than 25000000x25000000), which is why it is fixed by making a local coordinate system
It happens a lot in computing - if you've ever put big number coordintes in cloud compare it asks you to reduce them, and for years Autocad has had trouble making hatches in big coordinates which I think might be also related
Cheers Phill
Without doing a giant example say a variable has a maximum number of digits of 4 (They are much larger than 4) and you give it the number 1234567, it stores it as 1234e3 so when it is used in math the number is 1234000. When doing big coordinate math you might do 1 or more cross multiply of those numbers and it float overflows so when it get back to making a move on screen its jumpy cause it lost precision while doing the math. When you make a local system the math is smaller (ie 100x100 is a smaller number than 25000000x25000000), which is why it is fixed by making a local coordinate system
It happens a lot in computing - if you've ever put big number coordintes in cloud compare it asks you to reduce them, and for years Autocad has had trouble making hatches in big coordinates which I think might be also related
Cheers Phill