Fairly consistently, tracklogs made on my Garmin 76CSx appear approx 25m west of where they should be when viewed in Memory Map software. I live in the Lake District.

The Garmin 76CSx is correctly set up with "British grid" as the position format and "Ord Srvy GB" as the map datum. Tracklogs are made with the garmin record interval set to 20m.

The same tracklog viewed on the Garmin UK Topo map (installed on the 76CSx) appears in the right place.

To summarise:
The identical tracklog appears in slightly different places on the two maps. It's correct on the Garmin Topo, it's slightly out within Memory Map.

I do like to understand these inconsistencies.
Can anyone shed any light on this issue?

Cheers

Geoff

Hi Geoff,

It's possibly a map datum/calibration issue, which map are you using?

You can compensate for the effect by changing the "WGS84 Datum Shift" (meters) value(s) at the lower left corner of the > Map > Edit Map Properties... panel. But it shouldn't need changing (from zero) and you run the risk of upsetting the calibration of another part of the map.

Sometimes the WGS84 and OSGB datums get confused (by the GPS) but I believe the difference there is over 100 metres.

Cheers, Alan.

Thanks Alan,

The maps I notice this small error on are
OS 1:50k Great Britain (os50_gb.qct)
OS 1:25k Northern England - Region 4 (os25_r4.qct)

I tried importing a Garmin 76CSx tracklog into Google Earth tonight. The log displayed accurately without the slight shift I notice in Memory Map.

I won't change the 'WGS84 datum shift' from 0 in 'Edit Map Properties' but I would like to understand why there is a difference.

Regards

Geoff

Hi Geoff,

What about the calibration of the blue grid lines on the OS maps? Try the "data" (at cursor) popup on the map and/or create a test Mark/Waypoint with a clearly identifiable centre (e.g. the Dot) at an exact grid line (e.g. NY xx000 yy000). Note the last digit in the 5-digit GRs are metres, but single pixels on the 1:50k maps are 5 metres across.

Cheers, Alan.

Alan,

That's a great idea! It was very revealing. The grid intersections did not have nice round grid references as they should have......they were slightly out. I think it explains my observations.

I live in the Lake District.
The map I use most of all is OS 1:25k Northern England - Region 4 (os25_r4.qct)

Here's an example from that particular map.....
I placed a 'Flag' waymark on the grid line intersection ESE of the summit of Scafell Pike
It's grid reference was shown as NY 22015 07007
I then placed a dot on the map and manually gave it a grid reference of NY 22000 07000
The dot ended up slightly west of where it should be.
There's a Memory Map screenshot of my example here
https://picasaweb.google.com/geoffmozz/MemoryMapImage?authkey=Gv1sRgCLCO0YT_qqnb4AE#5682979062437084098

This is the same error I noticed with my tracklogs. ie They would show slightly west of where they should be.

I will check the offset is the same across the whole Region 4 map and then I could use the WGS84 datum shift to correct for it.

But why is it wrong in the first place?

Look forward to hearing your take on this one.

Thanks for your help so far, very useful.

Cheers

Geoff

Alan,

Some more info....

For the OS 1:25k Northern England - Region 4 map I entered a WGS84 datum shift of
+5 metres North and +15 metres East. This shift keeps the grid intersections around the lake District to within a metre or 2 of where they should be (using the cursor as a reference).

I then checked the grid line intersections at different extremities of the region 4 map to see if the error varied across the map. It did slightly, but only by about a few metres.

The map would appear to be a little more accurately placed with the datum shift above entered.

I wonder if I'm the only one with this issue?

Geoff

geoffmoss,

Um, thats not an official Memory-Map map and as such we cannot be responsible for any calibration errors.
We have never provided OS 1:25K data in regional sections i.e. region 4
The regions were only supplied in OS 1:50K mapping.

Thanks

Hi,

Well, it could be an official Selections product which has been renamed to something slightly more meaningful than the supplied 16-digit code. ;)

Personally I have a Selections map which is about 7 metres out, interestingly also predominantly to the east.

Cheers, Alan.

Thanks guys, all very interesting stuff....
--------------------------------------------
Jason...
Actually, I am the proud owner of the Memory Map DVD of the Lake District at 1:25k.

*Well that is not the map you mentioned to begin with (in fact you mention it in 3 seperate posts)
1.)

geoffmoss wrote:

The maps I notice this small error on are
OS 1:50k Great Britain (os50_gb.qct)
OS 1:25k Northern England - Region 4 (os25_r4.qct)

2.)

geoffmoss wrote:

The map I use most of all is OS 1:25k Northern England - Region 4 (os25_r4.qct)

3.)

geoffmoss wrote:

For the OS 1:25k Northern England - Region 4 map I entered a WGS84 datum shift of +5 metres North and +15 metres East.

The Memory-Map OS 1:25K Lake District product has the map named os25_lake_district.qct (which is shown in the second image you provided), whereas the Memory-Map OS 1:50K Northern England product has the the map named exactly as you wrote except that its os50 rather than os25 - Thanks Jason*

Thanks Jason. You're quite right about the Lake District map not being the first map I mentioned. It was your comments that made me check the Memory Map DVD os25 Lake District map. It does show the error I describe. My query remains. Can you explain why this error exists? Can you check this same map yourself using the Scafell Pike example, to see if you can replicate it? The two images I link to below are from the Memory Map Lake District DVD 1:25k. Cheers, Geoff.


The Memory Map from this DVD shows the error I describe. A waypoint placed dead on the intersection of the two grid lines immediately ESE of Scafell Pike's summit shows a grid ref of NY 22019 07007
A waypoint with the grid line intersection grid reference of NY 22000 07000 is unsurprisingly positioned 19m off to the west and 7m to the south of where it should be.

I've uploaded two images which show the waypoints on the map and the map info box here
https://picasaweb.google.com/geoffmozz/MemoryMapImagesX2?authkey=Gv1sRgCK-y75-Ip-mCmwE#

Again the error varies by 2 or 3 metres over the area the map covers.

Any ideas?
----------------------------------------------------------------
Alan, thanks very much for getting me to check the grid line intersections. It's a great way of checking map calibration. If need be, I can change the WGS84 datum shift for my map. No longer will my tracklogs be curiously and consistently 'out'.
----------------------------------------------------------------

Cheers

Geoff

Yes, I'm in the same position.
My MemoryMap OS 1:25k Lake District (on CD!) needed a 10m E datum shift.
(All my other MemoryMap CDs and DVDs seem OK).

Dales