Free maps for Garmin

[andygates]

Now, what on Earth is a violet navaid used for?

As far as I am aware it is either red, green, white or yellow!

Exactly. But the Garmin icons include blue and purple.

[Panoramix]

Now, what on Earth is a violet navaid used for?

As far as I am aware it is either red, green, white or yellow!

Exactly. But the Garmin icons include blue and purple.

I know for sure that USAians swap the green and the red so they may have extra colours as well!

[andygates]

Ooh, Mapquest have started a flirtation with OSM.  There's a web map here: MapQuest Maps - Driving Directions - Map -- though cyclepaths and tracks are excluded (facepalm!).  And they've bunged some loot into the OSM piggybank too. 

[andygates]

Aha! Violet is a chart convention for "a light" -- colour to be specified in the description code.  From the days of limited-colour reproduction, one assumes.

See: Lights, buoys - aids to navigation.

[Séamas M.]

Or an easier option is to download a map from the worldwide routable page linked a couple of posts back. It includes an installer for MapSource (I assume it works for Training Center as well?).

This worked. 

If you download the file: OSM-Garmin-Munky-UKIE-Tiles.zip it includes the tdb file as well as the map tiles etc. So you can use it with MapSource, Training Center etc.

This didn't, but that's probably just me being thick. 

[andygates]

After a week of staring at an orienteering map or two, I'm tempted to try an orienteering style for Garmin.  Obviously useless during events (GPS isn't generally allowed) but it could be handy for survey-walking at the hyperfine detail level orienteers work in.

Plus, it's cool and it's an excuse to finally nail that fine-detail contour thing... 

The orienteering build of OSM, OpenOrienteeringMap, is here: OpenOrienteeringMap (in progress - no contours yet).

[hbunnet]

Plus, it's cool and it's an excuse to finally nail that fine-detail contour thing... 

The orienteering build of OSM, OpenOrienteeringMap, is here: OpenOrienteeringMap (in progress - no contours yet).

This looks great.

Is it possible to get finer detailed contours? I thought the ones on the SMC site were at the limit of NASA resolution, am I wrong? (hope so).

[andygates]

They might well be at the limit of useful resolution - if the elevation model is estimated between those points, then it'll draw contour lines around those estimations and that won't neccessarily be true or useful.  They may still be pretty, just unreliable.  Will have to play around.

Opencyclemap uses 10m contours, as does SMC. 

[Greenbank]

They might well be at the limit of useful resolution - if the elevation model is estimated between those points, then it'll draw contour lines around those estimations and that won't neccessarily be true or useful.  They may still be pretty, just unreliable.  Will have to play around.

The contour lines themselves are interpolations.

Contours are now derived from a finer height grid. The 3 arcsecond * 3 arcsecond SRTM grid (about 60*90m at our Latitude) has been subdivided into 2 wide by 3 high (thus about 30m square) points, which are interpolated using a 36-point spline (i.e. taking account of the nearest 6*6 grid of SRTM datapoints). The resulting finer grid was then contoured as before (using 3DEM and DEM2TOPO). This gives 'rounder' contours, which take account of local gradients and gradient changes .. however small features will still be missing, since you can only 'round' what was seen to be there, not 'invent' things which were too small to get measured in the first place.

So the DEM data height of a single point is based on an interpolation of the real SRTM height points in the  surrounding 300mx480m area.

[andygates]

I didn't realise it was *that* low-res!  Seems to work okay for ordinary hiking/biking, at least in practice. 

[bikefish]

Im using 5 m interval contour in Australia - not perfect (drop offs not shown well) but useful as not many other land features in much of the country.. might be ASTER based.  I think parts of USA are at 10 feet. I would guess most of world is at 20 m.

[Somnolent]

I use Andy's superb Munky maps, and have done for a while, but I've only just noticed this, when I installed Mapsource on a new machine, lots of roads seemed to be discontinuous.  Here is an example:

Here is the Landranger view:



and the one from the OSM site



but when I look at it in the latest incarnation of Mapsource, a portion of the A3090 has "disappeared"



On the screen of the Etrex Legend it looks as though the "Monarch's way" footpath (which follows the A3090 at this point) is the top layer, and it's not obvious there is a road "underneath"... which is not a big problem.

It would be nice to have the roads "continuous" in Mapsource though.  Is it something I've done wrong?  Or an issue with the latest issue of Mapsource (6.16.2) ?   Or a choice of layers by Andy on the "munky maps" ?

[Wowbagger]

A few months ago we were cycling through Witham and I noticed that a chunk of very important road, which had definitely been there previously, had disappeared. This could be ineptitude on the part of a well-meaning editor.

I once inadvertently drained Abberton Reservoir. I was editing it to make the banks line up properly with Ordnance Survey Open Data and the banks were in a terrible tangle with power lines, roads and other lines. My solution was to delete and reinstate part of the bank. What I didn't realise was that a lake has to have a continuous boundary else it won't be filled in in blue. I couldn't work out a way in Potlatch to correct my mistake, even though I could see the colour draining from the tiles. Dez had to go in using the live editor (which my computer couldn't use because it was too slow) to correct my mistake.

[Somnolent]

It's not as if the road has been deleted.... it's there in OSM, and it's there in Mapsource (which will quite happily route along it) - it just isnt visible on the screen.

[Wowbagger]

It's not as if the road has been deleted.... it's there in OSM, and it's there in Mapsource (which will quite happily route along it) - it just isnt visible on the screen.

Can you clarify please? Are you saying it's not visible on the screen because perhaps the map is being viewed in too small a scale, or for some other reason? The road I was referring to, in Witham, had been deleted from OSM, whatever scale you looked at it.

[fuaran]

I've noticed similar problems with the Munky maps in Mapsource, with some roads and paths not showing at all. It seems to mostly be roads/paths that are part of cycle route relations?
As you say, they are not visible, but can be routed along. I've not tried loading them onto my eTrex, so I don't know if they show there.

I've got a few other OSM based maps that work fine in Mapsource, so I assume its something specific to the Munky map style files?

[Polar Bear]

I noticed whilst on holiday in Shetland that all the roads seemed to have disappeared.   I fiddled with a few settings and hey presto, they all came back.   Off the top of my head I cannot recall what it was but I spent quite a while sifting through the settings until they returned.
 

[Chris S]

I've seen disappearing roads before.

Is it a syntax or formatting error in the original OSM encoding?

There's a syntax checker isn't there, for finding this stuff?

[andygates]

That Hursley road is golden, though.  Most odd.  That Mapsource will route it and just not display it suggests that it's a Mapsource display error, but I can't for the life of me think why.

I think Fuaran has me with cyclepaths that are also roads -- I'll see if I can tweak the priority of that so that roads win. 

Chris, there's all sorts of checking tools.  Keep Right is the bestest IMO, it'll highlight all sorts of weirdness and deviation from expected standards: http://keepright.ipax.at/  -- here's its view of Hursley and the errors are vanilla: keep right!

[Panoramix]

Good website Andy,

When I go down the cyclepath in this area, the routing goes mental! I can't figure why, any clue? It will tell me to take perpendicular roads all the time as if it was trying hard to avoid the cyclepath!

[Somnolent]

 I'll see if I can tweak the priority of that so that roads win. 

I'm no expert..... but I have a suspicion that may be the solution.

Here is another example - near Stockbridge, Hampshire.



The A3057 is not at all discontinuous on openstreetmap.org, but take a look at a closer view in Potlach.... North of the upper "break" the cycleway is separate from the A3057 primary, but they are contiguous for that brief distance to the north & south of the RAB where the A30 meets the A3057.



Interestingly the Hursley one isnt a cycleway "overlapping" the road, it seems to be the Monarch's Way long distance footpath.

[frankly frankie]

It would eliminate one variable if, in Mapsource, you set the detail level to 'highest' instead of 'higher'.

I've noticed similar problems in Mapsource, and I don't use the Munky maps but my own build and style file - so I think it's more down to some more basic problem with the line and area attributes but I can't see anything in Potlach.  And it's the same regardless of Mapsource version - I use 6.13.7 for preference but do have 6.16 installed as well - no difference.
I'm thinking of an area just SW of Bayeux where, in Mapsource, some sections of primary road disappear under areas of forest or orchard, while other sections of the same roads sit on top as they should.

[TimO]

I need to get more au fait with the various tools, and mechanisms used to create these maps.

I found various weirdnesses with Andy's Munky Maps when using the auto-routing from home to work.  On Tooting Bec Common it seems to not recognise that there is a cycle route there, and keeps on trying to send me down the roads several hundred feet off of the Common.  On Clapham Common it repeatedly wants me to turn off of the cycle path, onto footpaths.  Along the side of Battersea Park it wants to send me on a long route through the park, rather than along the perfectly acceptable, and shorter, cycle path up the edge of Queenstown Road.

This isn't a criticism of Andy, who has done a fine job of something which I'm sure is full of subtle issues and complexities, but it would be nice to be able to sort out why these anomalies are happening.  Looking at the mapping on OpenStreetMaps suggests that the various bits are identified correctly as cycle paths and foot paths, so there's probably some fine details which I don't understand which is responsible.

[Panoramix]

I need to get more au fait with the various tools, and mechanisms used to create these maps.

I found various weirdnesses with Andy's Munky Maps when using the auto-routing from home to work.  On Tooting Bec Common it seems to not recognise that there is a cycle maps there, and keeps on trying to send me down the roads several hundred feet off of the Common.  On Clapham Common it repeatedly wants me to turn off of the cycle paths, onto footpaths.  Along the side of Battersea Park it wants to send me on a long route through the park, rather than along the perfectly acceptable, and shorter, cycle path up the edge of Queenstown Road.

This isn't a criticism of Andy, who has done a fine job of something which I'm sure is full of subtle issues and complexities, but it would be nice to be able to sort out why these anomalies are happening.  Looking at the mapping on OpenStreetMaps suggests that the various bits are identified correctly as cycle paths and foot paths, so there's probably some fine details which I don't understand which is responsible.

Mmmhhh, same thing as me! Reverse engineering is obviously tricky. Any pointer Andy what would be worth investigating?

[corshamjim]

FWIW I've seen weirdness using velomap.org on my garmin recently, notably where it tried to route me on to a cycle route from the A4 for 100 yards or so only to u-turn back on to the A4.  I couldn't see anything wrong on OSM afterwards when I checked in potlatch to see if there was anything obvious amiss there.