| View previous topic :: View next topic |
| Author |
Message |
najevi Capitaine

Joined: 20 Dec 2008 Posts: 248 Location: Gold Coast (Australia)
|
Posted: 16 Aug 2009 8:36 Post subject: Questions about track files and coordinate systems |
|
|
I am trying to determine the most efficient way of displaying race courses at vsk.wikia.com Since there could easily be several hundred to a few thousand, it will not be very efficient or wise to simply upload JPG/PNG files. (Anyone who has installed the RC44 boat model will have discovered how the JPG image per Track.Gbx method of documenting a race course is terribly hungry on disk space.)
If you have a good understanding of the coordinates used by the VSK5 custom race editor (and presumably stored in the <track-name>.Gbx files) then please do your best to answer the following.
1. How big is the 3D-world map for each of the 16 sailing venues that come with VSK5? (The Sydney map is approx. 3200x2400 pixels at maximum zoom out and the zoom magnification seems to be 7-times so approx. 22400x16800 pixels at maximum zoom in.)
1a. Is this correct? [only for Sydney, different for others]
1b. Is it the same for all venues?) [No. Not all venues are the same size.]
2. (answered) Is there more than one map for each of the 16 sailing venues? (I've seen a few different Nordic and Tropical maps.)14 of 16 seascapes are fixed/locked terrain. Nordic and Tropical seascapes are a blank canvas allowing user customization of terrain. Nordic and Tropical maps will therefore need to be uploaded to the wiki as needed using PNG or JPG images.
3. ( answered ) Are the maps from prior version of VSK re-usable on VSK5?
3a. Might this explain why I seem to see many different maps for the one venue? - VSK4, 32AC and VSK5 maps can be used with the VSK5 client
- VSK3 maps might be usable
- The many different maps for the same venue is only for Nordic and Tropical venues - explained elsewhere
4. ( answered ) Regarding the block based map building that Maitai et al used to create the "map of the world" 3d-world map - is that widely used by other community members?Used primarily for solo play however some clubs (notably Virtual-Winds) race multiplayer online with that world map regularly.
5. What coordinate system does the editor use? [simple square grid]
5a. Is it pixel based?
5c. Where is the origin? (Top left pixel is a common standard.) [near bottom left]
5d. Are negative coordinates ever used? [assume yes]
5e. Is the scale of px:Nm the same for each sailing venue? [No. Not all the same.]
6. (answered) Do you know how to extract the coordinates of each mark/buoy from the Track.Gbx file?No. Only Nadeo can assist with dumping these coordinates to ASCII form.
Why do i ask?
I am trying to determine the most efficient way of displaying race courses at vsk.wikia.com. I like the idea of leveraging the Google Maps concept and have access to that extension. It involves tiling each 3D-word into 256x256 pixel tiles and dynamically selecting only those tiles needed to display a race course area. An overlay feature with simple polyline and icon style graphic capability is then used to display the marks and legs of a course in solid colors.
What I am not sure about is how to get my hands on a high-resolution map of each sailing venue ... showing the entire venue. (It appears that any of several file formats are suitable - including .dds) The 3D-world view includes perspective that cannot be reproduced in the Google maps model but I suspect it is still the more efficient approach to take.
If this subject interests you then please read the following wiki forum post and offer your feedback/advice. ( http://vsk.wikia.com/wiki/Forum:Is_there_a_Google_Maps_guru_reading_this%3F )
Regards,
Last edited by najevi on 21 Aug 2009 17:29; edited 4 times in total |
|
| Back to top |
|
 |
kiwi_bardy Amiral
Joined: 28 Jun 2007 Posts: 416 Location: New Zealand
|
Posted: 16 Aug 2009 10:09 Post subject: Re: Questions about track files and coordinate systems |
|
|
| najevi wrote: |
2. Is there more than one map for each of the 16 sailing venues? (I've seen a few different Nordic and Tropical maps.)
|
Tropical and Nordic maps start off as just empty ocean.... they are 'construct-a-map' venues... the map creator adds land in any pattern they like - that's why you have seen several different maps.
As per this thread, the Nordic (approx 20 Nm x 20 Nm) and Tropical (approz 5Nm x 5 Nm) maps are different sizes... |
|
| Back to top |
|
 |
najevi Capitaine

Joined: 20 Dec 2008 Posts: 248 Location: Gold Coast (Australia)
|
Posted: 16 Aug 2009 16:56 Post subject: |
|
|
Thank you. I found more information in the Manual after being prompted by that thread. (I also found reference to the wind strength - something I recall having questions about in the past.)
Understanding that 14 of 16 seascapes/venues have fixed/locked terrain I am encouraged to pursue a tile-based approach for displaying courses at those venues leveraging the Google Maps API. - Nordic and Tropical will have to be treated as the exception to the rule and require JPG/PNG/GIF uploads at a suitable resolution.
This plan still hinges on extracting coordinates for the racing course marks from the Track.Gbx file. So there are still a few more "miracles" needed.
If any user (or Nadeo) has pieced together a composite image of the entire map area for the 14 venues then please post some URLs for download. |
|
| Back to top |
|
 |
kiwi_bardy Amiral
Joined: 28 Jun 2007 Posts: 416 Location: New Zealand
|
Posted: 16 Aug 2009 23:20 Post subject: |
|
|
| najevi wrote: | This plan still hinges on extracting coordinates for the racing course marks from the Track.Gbx file. So there are still a few more "miracles" needed.  |
It would seem to me that it also assumes that the Nadeo rendering of the land and waterways is reasonably accurate at each of the venues.
I know that it is quite accurate at some sites, like for instance Marseilles, and some parts of La Trinite - however other parts are very inaccurate, or at least artistic license has been taken. Just a couple of examples are:
1. The Gulf of Morbihan at La Trinite where the islands and right hand end of the gulf are reasonably 'creative' interpretations of the real Gulf's terrain features; and
2. Parts of the Auckland map - South East of Motutapu Island, for example, is open water where in reality there is the Buckland's Beach Penninsula, Brown's Island, Motuhuie Island and Waiheke Island there.
I don't think this is a show stopper, BUT if you had a course in the bottom right hand corner of the Nadeo Auckland map South and South East of Rangitoto and Motutapu islands, say, the google map approach you are talking about (assuming the other miracles happen) would likely result in some courses there having marks on land or legs passing through islands that are not there on the Nadeo map. This could be confusing.
These are just two examples where this may be an issue. I am sure there are more. The point is the approach is likely to result in (in at least some instances) maps that do not look like the course in the game. It's a great idea, and certainly saves a lot on the manual work of generating course thumbnails...
...would it be possible to write an application that takes the coordinates and applies them to a pre-adjusted map perhaps to address this limitation? |
|
| Back to top |
|
 |
najevi Capitaine

Joined: 20 Dec 2008 Posts: 248 Location: Gold Coast (Australia)
|
Posted: 17 Aug 2009 1:15 Post subject: |
|
|
In case others are of the same misunderstanding as Bardy, I should clarify that the Google Maps API I am referring to is a generic set of tools that will accept any image as it's map. In the case of the 14 sailing venues that come with VSK5, the image(s) I plan to use are a full resolution image of the Nadeo generated map for each of those 14 locations. For examples of this having been done for another online game wiki refer to the wiki forum post mentioned above and follow the links provided there.
The coordinates of buoys contained in the Track.Gbx file (and hopefully to be dumped in ASCII readable form in either a Result.csv file or similar) will therefore be easily applied to such a map and should yield an identical chart of the course minus any water and air traffic that the track designer may have programmed. (I may suffer a few headaches over the algorithm for drawing a correct path through gates. Even the Nadeo generated path sometimes fail to correctly illustrate when a gate buoy must be rounded. I am not sure that I know the precise rules for this but somebody does.)
That is why I am asking for sources of the Nadeo created maps. With enough time and patience those could be generated manually however I have to believe that these images are sitting on a disk somewhere on the planet just waiting to be emailed or downloaded.
ETA sometime in October depending on how much help I get from someone experienced with the API.
I do plan on using the real world maps/satellite images/etc. of traditional Google maps at each Venue page just to show where in the real world each seascape can be found. However that is intended more as a piece of trivia or curio for the geography buffs and world travelers. It is independent of the display of racing course data. Setting this up should take less than a week. ETA sometime in September.
So you see the real world maps will not come into play with respect to displaying race courses but will come into play when illustrating each Venue page. FYI I did some google research to come up with the following list of lat/long data for each of the venues but I'd appreciate someone else checking that these are accurate. http://vsk.wikia.com/wiki/Category:Venues
A few of the venues had multiple same named places so a little sleuthing was necessary. Feel free to make any corrections necessary. |
|
| Back to top |
|
 |
kiwi_bardy Amiral
Joined: 28 Jun 2007 Posts: 416 Location: New Zealand
|
Posted: 17 Aug 2009 1:58 Post subject: |
|
|
Ahhhhhh - realisation dawns.
;o) |
|
| Back to top |
|
 |
admiral Amiral

Joined: 27 Oct 2003 Posts: 336 Location: Nederland
|
Posted: 17 Aug 2009 9:27 Post subject: |
|
|
| Quote: |
I know that it is quite accurate at some sites, like for instance Marseilles, and some parts of La Trinite - however other parts are very inaccurate, or at least artistic license has been taken. Just a couple of examples are: |
IMO of similar, or even more interest are accuracy of the currents, and wind to each venue. If these could be verified then actual tidal maps and wind grids could be used. Which wouldn't be as cumbersome as going through the track editor to check current and or wind conditions. |
|
| Back to top |
|
 |
maitai Amiral
Joined: 04 May 2006 Posts: 354
|
Posted: 21 Aug 2009 9:40 Post subject: |
|
|
I ahve search a lot on that quation when doing the worldmap. Although I made a few discoveries, I don't think it's enough to extract reliably mark coordinates. I am afraid we have to wait for Nadeo help on that one. _________________ www.virtual-winds.com |
|
| Back to top |
|
 |
admiral Amiral

Joined: 27 Oct 2003 Posts: 336 Location: Nederland
|
Posted: 21 Aug 2009 11:56 Post subject: |
|
|
Obviously what is needed is all map coordinates expressed in Longitude and Latitude. It's up to Nadeo to implement. An Easy way out could be to just have latitude and longitude for the north-west corner of every venue, All else can then be derived from the grid size.. transfer length and with in hundreths or thousends of degrees.
When the grid is small enough projection errors, a flat area in vsk vs a spherical part of the globe, should be neglectable.
AFAIK all vectors (Currents, Winds) are allready in ° and Kts or m/s
basic venue 20 x 20 miles = 0.333° x 0.333° at the equator
Last edited by admiral on 21 Aug 2009 12:18; edited 1 time in total |
|
| Back to top |
|
 |
admiral Amiral

Joined: 27 Oct 2003 Posts: 336 Location: Nederland
|
Posted: 21 Aug 2009 12:06 Post subject: |
|
|
neglecting projection errors may not be that smart:
http://en.wikipedia.org/wiki/Geographic_coordinate_system
Copy and Paste for the lazy:
On a spherical surface at sea level, one latitudinal second measures 30.82 metres and one latitudinal minute 1849 metres, and one latitudinal degree is 110.9 kilometres. The circles of longitude, meridians, meet at the geographical poles, with the west-east width of a second being dependent on the latitude. On the equator at sea level, one longitudinal second measures 30.92 metres, a longitudinal minute 1855 metres, and a longitudinal degree 111.3 kilometres. At 30° a longitudinal second is 26.76 metres, at Greenwich (51° 28' 38" N) is 19.22 metres, and at 60° it is 15.42 metres.
The width of one longitudinal degree on latitude \scriptstyle{\phi}\,\! can be calculated (to get the width per minute and second, divide by 60 and 3600, respectively) where Earth's average meridional radius approximately equals 6,367,449 m. Due to the average radius value used, this formula is of course not precise. You can get a better approximation of a longitudinal degree at latitude by:
where Earth's equatorial and polar radii, 6,378,137 m, 6,356,752.3 m, respectively.
Length equivalent at selected latitudes in km
Latitude Town Degree Minute Second ±0.0001°
60° Saint Petersburg 55.65 km 0.927 km 15.42m 5.56m
51° 28' 38" N Greenwich 69.29 km 1.155 km 19.24m 6.93m
45° Bordeaux 78.7 km 1.31 km 21.86m 7.87m
30° New Orleans 96.39 km 1.61 km 26.77m 9.63m
0° Quito 111.3 km 1.855 km 30.92m 11.13m
I do think that latitude conversions can be the same for any venue however the venues shoudn't be to large. (This will never work for a world map..)Though to get an average maybe the middle point is better then the north-west corner as i suggested in the earlier post.
Frankly maybe some who know the venues well are able to determine latitude and longitude for each venue? |
|
| Back to top |
|
 |
najevi Capitaine

Joined: 20 Dec 2008 Posts: 248 Location: Gold Coast (Australia)
|
Posted: 21 Aug 2009 16:44 Post subject: |
|
|
Maitai - when you say that you could not figure out the mark coordinates from the Track.Gbx file then I definitely sit up and pay attention. It would be ideal if Nadeo was to provide these coordinates (along with other track data) in ASCII form along with each Track.Gbx file. Admiral and Bardy are both aware of my wish list in that regard.
Just to clarify my intention with this set of questions. At this time, I have no interest in plotting race courses on a map of the real world. My interest extends only to plotting marks to overlay on a map of each of the Nadeo provided venues/seascapes.
This means that latitude and longitude measured in degrees and minutes are not required. Spherical coordinates are not required. Simple square grid, pixel based coordinates of each mark on the course are all that is necessary.
Maitai, thanks for your other answers via PM.
The only remaining question of immediate interest is where can I obtain high resolution images of each of the venues/seascapes? I suspect Hylis et al will need to come to the rescue on this point.
Last edited by najevi on 21 Aug 2009 17:58; edited 1 time in total |
|
| Back to top |
|
 |
maitai Amiral
Joined: 04 May 2006 Posts: 354
|
Posted: 21 Aug 2009 17:39 Post subject: |
|
|
What I meant is I was able to identify one single bouey coordinates, by comparing binarily 2 gbx with only a small move of a single bouey. That allowed me to figure out the coordinates system at that time, but I am not sure I remember what it was (I think 0,0 is just the left/down corner of the grid, but I am not sure and too lazy to do it again). The problem was to identify the other marks because they can be anywhere in the gbx file, there is not structure as such, it's just a collection of objects in whatever order as I can tell. And you don't even know how many marks you are looking for....
To make a long story short, without Nadeo's help it's hopeless I think, or it will take a huge effort to "crack" it for a somehow small reward.
PS: this knowlegde would also probably allow to inject one player's replay track in another replay based on the same map, allowing for ghost replay. And to calculate some stats (distance and therefore average speed on water, number of tacks, etc). A very useful thing when playing "time-record" course type in solo mode. But again Nadeo's help is necessary here too. _________________ www.virtual-winds.com |
|
| Back to top |
|
 |
admiral Amiral

Joined: 27 Oct 2003 Posts: 336 Location: Nederland
|
Posted: 22 Aug 2009 9:43 Post subject: |
|
|
| Quote: | | where can I obtain high resolution images of each of the venues/seascapes? |
Not only high resolution but also accurate images can be obtained from any
navigation system. Actual oceanographic maps.
The coordinates of the marks on any vsk-map can be obtained by an birds-eye image of the course, which can be scaled to be superimposed on oceanographic-maps obtained from a navigation system.
For this it where easies if latitude and longitude of a map are known.
Automation of the whole process is possible but bends on the ability to
transfer mark coordinates, not in the least the amount of resources thrown at it.  |
|
| Back to top |
|
 |
najevi Capitaine

Joined: 20 Dec 2008 Posts: 248 Location: Gold Coast (Australia)
|
Posted: 21 Jan 2010 8:43 Post subject: |
|
|
| najevi wrote: | | The only remaining question of immediate interest is where can I obtain high resolution images of each of the venues/seascapes? I suspect Hylis et al will need to come to the rescue on this point. |
To be very clear I am not looking for any form of real world map or oceanographic survey charts, weather charts or satellite photographs.
The only satellite view that is necessary for this project is that same map view we are presented with when we press the tilde (~) key from within the game. Maximum resolution/detail is best because the MediaWiki software and the GoogleMap API already take care of caching and using low-resolution tiles to achieve fast loading web pages.
If I had the patience of a saint (and I do not) then I would spend a month of Sundays pedantically panning and then saving screen shots (at maximum magnification) then painstakingly stitching these together for a composite image. Most of you would agree that is a thankless task.
So Hylis/Florent and team: - Can any one of you extract a maximum magnification (highest detail) satellite view map of each of the 14 (fixed) venues?
- If you can then are there any intellectual property concerns you'd have with allowing these to be used at the wiki to facilitate the efficient display of user created courses? (Features of the GoogleMap API will be used to overlay mark coordinates and lines showing the path from start line to finish line. Demo here.)
I don't remember if you can notice any difference between low tide and high tide in that satellite view but if there is a noticeable difference then the better map to have will be the low tide map. _________________ it's better to spin and fin' than to fit and quit!  |
|
| Back to top |
|
 |
admiral Amiral

Joined: 27 Oct 2003 Posts: 336 Location: Nederland
|
Posted: 05 Feb 2010 17:59 Post subject: |
|
|
How much can you get on a screen with the free camera at the highest point?
If it only takes like 4 screenshots at like 1900 x 1200 that need to be pasted that may proof faster... then somebody tweaking that camera hight to get a high-res shot of the whole venue. |
|
| Back to top |
|
 |
|