| View previous topic :: View next topic |
| Author |
Message |
guest
Joined: 08 Oct 2006 Posts: 3949
|
Posted: Sun Dec 02, 2007 11:54 pm Post subject: maps shift with big scale |
|
|
Hello,
I'm a new and fan user of TB, with a SE W660i and last TB version. It's working very well, I will only use it for my next road/bike/paraglider trip !!
I have a problem of map display shift (about 1.5km) with several maps (from an Atlas) I made with the web site gm2tb. With a big scale (10km/cm) of those maps, I have a shift, with a small scale (500m/cm), the position is correct...
Do you have an idea about this problem ?
thx a lot!!!
Romain
ps: I can not send here my gpx log of the trip, I'm not allowed to attach it ...
|
|
| Back to top |
|
 |
temptrekbuddy
Joined: 27 Nov 2006 Posts: 193 Location: Poland
|
Posted: Mon Dec 03, 2007 2:27 pm Post subject: Re: maps shift with big scale |
|
|
| guest wrote: |
Hello,
ps: I can not send here my gpx log of the trip, I'm not allowed to attach it ... |
zip it then.
_________________ Automatic Free 2GB backup -> Mozy; with this link 256MB extra. |
|
| Back to top |
|
 |
kruch Site Admin
Joined: 02 Jul 2006 Posts: 5088
|
Posted: Mon Dec 03, 2007 2:44 pm Post subject: |
|
|
| Isn't this a known issue that gm2tb calibration for big scale isn't accurate.... ?
|
|
| Back to top |
|
 |
ck3d
Joined: 08 Dec 2007 Posts: 2
|
Posted: Tue Dec 18, 2007 7:29 pm Post subject: |
|
|
In the middle of large scaled map-sets TrekBuddy has got latitude deviations.
The position is correct displayed at bottom or top of map-set.
I think there is a mapping problem in TrekBuddy!
I tested it with gmap and osm.
|
|
| Back to top |
|
 |
Ranger
Joined: 12 Dec 2006 Posts: 4726
|
Posted: Tue Dec 18, 2007 8:39 pm Post subject: |
|
|
First check your maps in OziE, for example using tracks.
I saw a lot of maps wrong calibrated.
If you have map working in OziE well and not working in TB, just attach sample or link to sample.
No way to check problem without maps you used.
Problem with gm2tb is well known, projection info is not included in calibration file, so only detailed zoom levels can work.
Up to this time no solution from tool author is available.
|
|
| Back to top |
|
 |
ck3d
Joined: 08 Dec 2007 Posts: 2
|
|
| Back to top |
|
 |
kruch Site Admin
Joined: 02 Jul 2006 Posts: 5088
|
Posted: Thu Dec 20, 2007 3:53 pm Post subject: |
|
|
| I'll try to compare two large scale Google maps of the same area - one created by gm2tb, and another by GoogleHackPureJava2 (if it still works against current Google maps).
|
|
| Back to top |
|
 |
kruch Site Admin
Joined: 02 Jul 2006 Posts: 5088
|
Posted: Thu Dec 20, 2007 4:40 pm Post subject: |
|
|
OK... I could not get Google maps, but got VirtualEarth maps. I'm convinced there is a bug in gm2tb.
On attached pictures, you may see that location of small town Beroun south-west of Prague. (approx lat 49deg 57min lon 14deg 4min)
Coordinates from alternative downloader and online map service (mapy.cz) matches, but the same location is shifted by many kilometeres on gm2tb map.
Since gm2tb maps does not start/end on Google map tiles boundaries (unlike map from GoogleHackPure java and its successor), I suspect there a bug in calibration in gm2tb, where some calculation is therefore needed to get map corner points that do not match map tiles corner points. In fact, gm2tb does not have to do this recalculation, since Points in the Ozi calibration file do not have to be at the map corners...
| Description: |
|
| Filesize: |
51.29 KB |
| Viewed: |
8700 Time(s) |

|
| Description: |
| "Beroun" on purehack2 map |
|
| Filesize: |
30.81 KB |
| Viewed: |
8700 Time(s) |

|
| Description: |
|
| Filesize: |
40.53 KB |
| Viewed: |
8700 Time(s) |

|
|
|
| Back to top |
|
 |
olegich
Joined: 07 Aug 2007 Posts: 70
|
Posted: Tue Jan 01, 2008 4:16 pm Post subject: |
|
|
kruch
It appears on maps with more than 7 vertical tiles(256x256).
More tiles-more shift (scales are equal).
Also with the same dimension and the same scale but different vertical borders.
pic#1: 61.270N 57.704N (small shift, crossroad at top of the map)
pic#3: 63.234N 59.889N (shift has grown)
TB_____________________OZI
added: i have to say that 7 tiles for maps with zoom=10 (in google notation), that i've attached and in high latitudes (due to mercator projection)
| Description: |
|
 Download |
| Filename: |
OZImaps.zip |
| Filesize: |
382.86 KB |
| Downloaded: |
34 Time(s) |
| Description: |
|
 Download |
| Filename: |
tracks.zip |
| Filesize: |
777 Bytes |
| Downloaded: |
17 Time(s) |
| Description: |
|
 Download |
| Filename: |
TBmaps.zip |
| Filesize: |
777.17 KB |
| Downloaded: |
30 Time(s) |
|
|
| Back to top |
|
 |
start78
Joined: 17 Oct 2007 Posts: 186
|
Posted: Sat Jan 12, 2008 7:41 am Post subject: |
|
|
Hi kruch,
since you already made the first steps to solve this issue, wouldn't you think it a good idea to contact the gm2tb-creator and help him solve it?
I made an attempt to find out what is wrong here. Until now i only know that he created gm2tb by reverse engineering from mapexplore. (I was in e-mail contact with him. His mail adress is easy to find in his awokenmind-blog)
According to his statistics there are many people using gm2tb. Perhaps joining forces with him can be usefull for both sides?
My first idea (perhaps you remember) was a manual calibration in tb. I think the upper left corner of a large map should be correctly calibrated. Perhaps if you specify just one more point on the map inside tb it could calculate the misscalibration and correct it. This would be a poor workaround since it doesn't solve the original problem and blotes the program...
<edit>
| Ranger wrote: |
| ...projection info is not included in calibration file, so only detailed zoom levels can work... |
Until now i didn't understand what you meant in the above linked thread, sorry. Can you think of a way how gm2tb can get the necessary projection line from gmaps? If i can tell the creator that this line is missing perhaps he can add it to his programm.
I just now wrote him an e-mail and told him about the missing projection infos. Perhaps he is able to come up with a solution.
</edit>
_________________ Nokia 5800XM (FW 21.0.025 T-Mobile Ger) / TB 0.9.89 / Atlas by TBAC |
|
| Back to top |
|
 |
Ranger
Joined: 12 Dec 2006 Posts: 4726
|
Posted: Sat Jan 12, 2008 10:34 am Post subject: |
|
|
Hi start78,
First of all, I use working well Mercator projected maps in TB.
Map calibration is not just playing with numbers, without understanding what they represents. Bug in tool is known, so Kruch have nothing to do with it in TB. I can say again - before you use GM map in TB, try to check how it works in OziE. Doesn't work. From TB point of view all is clear - if datum/projection info is not included in calibration file, TB defaults to WGS/UTM. Question is what kind of coordinates gm2tb will write to calibration file.
Regards,
Ranger
|
|
| Back to top |
|
 |
start78
Joined: 17 Oct 2007 Posts: 186
|
Posted: Sat Jan 12, 2008 12:24 pm Post subject: |
|
|
I don't know much about calibration and projection. And i think the creator of gm2tb doesn't, too. But you and kruch seem to know.
That's why i'm asking all the time. If you can explain me what is necessary or tell me where i can easily get this knowledge, maybe i can figure out what to do and tell the gm2tb-creator about it.
Is it so hard to calculate the needed projections if you have a calibration point and a zoom level in gmaps?
Gm2tb is already one of the top results if you do a google-search about trekbuddy and many people seem to use it (over 600 maps per day) because you can create a ready-to-use tar-map without any knowledge about map-creation or any program installed on yout PC. I don't know how to make such a tool but i would really like to help fixing this annoying bug so the community (and me, of course) has a handy tool for creating tb-maps in no-time.
I know a forum isn't about asking stupid question that could have been answered by doing a google-search...
_________________ Nokia 5800XM (FW 21.0.025 T-Mobile Ger) / TB 0.9.89 / Atlas by TBAC |
|
| Back to top |
|
 |
Ranger
Joined: 12 Dec 2006 Posts: 4726
|
Posted: Sat Jan 12, 2008 1:44 pm Post subject: |
|
|
Hi again,
Technical knowledge about maps is one, but you touch few other thinks.
TB is mainly designed for outdoor. For such activities you need suitable maps.
First, I don't see too much value of GM road maps in outdoor activities. But probably you are right, street and road maps can be interesting for GPS beginners to start, if they couldn't find better map sources.
Remember, Google Terms restricts use of their maps with other navigation software.
Second, as an example, last year we started (in my country) cooperation with commercial topo maps producer to have TB ready maps suitable for outdoor activities. And we have first bundle of TB with electronic & paper maps available now
More info about it you can find in TB wiki.
Which way to choose ?
Regards,
Ranger
|
|
| Back to top |
|
 |
start78
Joined: 17 Oct 2007 Posts: 186
|
Posted: Sat Jan 12, 2008 3:42 pm Post subject: |
|
|
The coop with compass is a good idea. But it costs money and isn't available everywhere.
About using gmaps for outdoor use: yes, they lack topo infos but at least for my region they are extremely detailed concerning ways and paths.
As long as the gm2tb-maps are for free and TB is for free i don't see many conflicts with Google Terms.
Why should treckbuddy be reduced to outdoor navigation? I already made very good experiences supporting navigation per paper roadmaps (ask my girlfriend).
This goes Off-Topic...
_________________ Nokia 5800XM (FW 21.0.025 T-Mobile Ger) / TB 0.9.89 / Atlas by TBAC |
|
| Back to top |
|
 |
Ranger
Joined: 12 Dec 2006 Posts: 4726
|
Posted: Sat Jan 12, 2008 5:04 pm Post subject: |
|
|
OT
| start78 wrote: |
The coop with compass is a good idea. But it costs money and isn't available everywhere.
As long as the gm2tb-maps are for free and TB is for free ... |
Right. It isn't available but can be ?
You don't see restrictions, but Google ... asked some apps designers to remove GM support. Surprised ?
| start78 wrote: |
Why should treckbuddy be reduced to outdoor navigation? I already made very good experiences supporting navigation per paper roadmaps (ask my girlfriend).
|
Reduced ? No.
I don't know how long you read this forum, but more than year and half ago Kruch wrote topic why TB was designed ...
Road navigation is a bit supported, but cannot be compared with specialized road navi tools available today (with local maps, autorouting, voice directions, etc.)
|
|
| Back to top |
|
 |
|