Pick a datum, any Datum.....

[+paul.blitz]

Ok, I am using OziExplorer, and there are potentially 4 map-related things that have to be set:

1) map datum

2) map projection

3) waypoint list datum

4) individual waypoint datum

 

As my map comes from Multimap, and I use the grid to calibrate the map, (1) will be OSGB, (2) will be BNG. So far so good

 

Individual waypoints get quoted here as WGS84, so (4) is WGS84.

 

But what do I set (3) to? OSGB? WGS84?

 

I have just done a big map of N Yorkshire, and put in a waypoint that I know (in the middle of York, which I know pretty well) and it's a little bit out (ok, maybe only 200 or 300 metres, but I want PERFECTION!!!)

 

Paul

 

Team Blitz

 

No, I gave YOU the spare batteries....

[+jeremyp]

I don't know the answer to your question. Isn't a waypoint list just a list of waypoints? How does it have a datum?

 

But, your error is consistent with confused datums. i.e. if you entered a waypoint in OSGB36 when it should be WGS84 or vice versa, you'd expect it to be off by 100 - 200 metres.

 

-------

jeremyp

The second ten million caches were the worst too.

http://www.jeremyp.net/geocaching

[+washboy]

Excellent, isn't it (just difficult to find adequate maps)?

 

I'm not sure I properly understand your difficulty, Paul. Your waypoints will be correct to whichever datum you created them in (probably WGS84). This attribute is (crucially) part of the OE waypoint file data structure.

 

The chances are, your waypoint file will be set as WGS84 and, when the waypoints are loaded, OE "recalculates" to the datum of the loaded map. If you wish to, you can override that (in the waypoint list dialog) to set the waypoint datum to whatever you like. The point is that it's only the figures in the waypoint list display which are affected by this.

 

I suppose the real question is, how/where did you create the waypoint?

 

Caveat: I'm clearly not Des Newman (creator and author of OziExplorer) so my interpretation here is just that - an interpretation of what OE does

 

=====

There's no such thing as a free lunchbox!

[+paul.blitz]

quote:

---

Originally posted by washboy:

Excellent, isn't it (just difficult to find adequate maps)?

---

 

I cut & paste mine from Multimap's 1:25000 ones. Takes ages, but the maps are fairly good. So far I have one that covers Winchester / Southampton, another for York & Northwards, and a small one around Aylesbury: shout if you'd like a copy.

 

quote:

---

I'm not sure I properly understand your difficulty, Paul. Your waypoints will be correct to whichever datum you created them in (probably WGS84). This attribute is (crucially) part of the OE waypoint file data structure.

---

 

Ah, ok... so you can enter waypoints in any format you like, but they get converted to the datum you define for the waypoint list....that explains something I have seen.

 

quote:

---

The chances are, your waypoint file will be set as WGS84 and, when the waypoints are loaded, OE "recalculates" to the datum of the loaded map.

---

 

Yes, I'd worked that bit out.... So I guess if I keep my waypoints all as WGS84, and the map as OSGB I'll be fine

 

I suppose the real question is, how/where did you create the waypoint?{/QUOTE]

 

I manually enter them....

 

Thanks for your thoughts... sounds like I'm doing it all right then!

 

Paul

 

Team Blitz

 

No, I gave YOU the spare batteries....

[+washboy]

quote:

---

Ah, ok... so you can enter waypoints in any format you like, but they get converted to the datum you define for the waypoint list...

---

 

Yes, you can enter waypoints using any datum and they get displayed in the waypoint list translated to whichever datum you select for the list display. The important bit is that they are converted to the datum of the loaded map just before they are plotted on it. It's a little like numbers on a spreadsheet being helpd in a common format "behind the scenes" but displayed according to the number format defined for any given cell.

 

So (back to waypoints), when you enter a waypoint, you need to specify the datum of the co-ords you're supplying but the co-ords on the target display (be it in the waypoint list or on a map) are the result of an on-the-fly transformation to the datum of the display. Apologies for such an ugly explanation.

 

quote:

---

So I guess if I keep my waypoints all as WGS84, and the map as OSGB I'll be fine

---

 

Certainly you'll want to keep the maps you mention as OSGB but, as for your waypoints, it's up to you which datum you'd rather work with. Co-ords from GC.com will have to be entered as WGS84, of course, but if you create your own waypoints based on what you see on an OSGB map, then it's probably more convenient to enter them as OS Grid Refs than use a seperate convertor to get WGS84 co-ords.

 

If you haven't already cottoned to this, it'll help to notice that the "add waypoint" entry dialog is different depending whether you're displaying the waypoints in the list as degrees or BNG (the two icons just to the right of the "sort" icon). If you're displaying degrees then you have to supply the waypoint co-ords in Lat/Long (degrees) and if you're displaying BNG then you get to supply the old Eastings & Northings ("TQ12334 12345").

 

NB: It's essential that you select the correct datum in the "Position Datum" dropdown list on the entry form. However, it's irrelevent what you've selected from the "Datum" dropdown list on the waypoint list - that only affects which numbers display within the list and that's why the numbers change (sometimes subtly) depending upon the particular datum you've selected.

 

Apologies if grandmother & eggs applies here. It wasn't at all obvious to me, when I first used OE, so that's why I labour the point.

 

quote:

---

sounds like I'm doing it all right then!

---

 

Well, I'm still unclear whether or not you specified the appropriate "Position Datum" when you entered the waypoint but I'm sure you'll now be able to figure out if you did anything wrongly.

 

BTW, YHM

 

=====

There's no such thing as a free lunchbox!

[+paul.blitz]

Been playing more with Ozi Explorer, and have worked out why things didn't look correct... like you move cursor onto a grid line, but the displayed x/y location was about 100 mtrs out.

 

I set up with the map calibration set as British grid / OS BG, whic is, of course 100% correct for the multimaps.

 

Of course, when you READ the displayed location (in Nat Grid format, as wel as deg & mins) it DOES clearly say on the bar "WGS84".....

 

So whilst the readout is in grid format, its displaying on the WGS84 datum, rather than the map's OS datum.....

 

DOH!!!

 

Paul

 

Team Blitz

 

No, I gave YOU the spare batteries....

[+washboy]

If you calibrated your map with:

 

Map Datum = Ord Srvy Grt Britn

Map Projection = (BNG) British National Grid

 

then the bar (you mean the one just below the toolbar?) should say "Ord Srvy Grt Britn" to the right of the figures. If it says "WGS 84", it suggests that the map was actually calibrated with the WGS84 datum instead of OSGB. The correct calibration for Ordnance Survey maps is as above.

 

However, even if you calibrated correctly, OE's Lat/Lon readout (on the "bar" and on "flyover hints") will be OSGB Lat/Lon.

 

So, if you create a waypoint with WGS84 Lat/Lon, OE will translate to OSGB and it will appear on the map in the correct place but the Lat/Lon on the bar will appear wrong - 'cos it's OSGB Lat/Lon NOT WGS84 Lat/Lon.

 

Nuisance, isn't it?

 

=====

There's no such thing as a free lunchbox!

[+Pharisee]

quote:

---

Originally posted by washboy:

 

So, if you create a waypoint with WGS84 Lat/Lon, OE will translate to OSGB and it will appear on the map in the correct place but the Lat/Lon on the bar will appear wrong - 'cos it's OSGB Lat/Lon NOT WGS84 Lat/Lon.

 

---

 

I'm not sure if you've helped Paul but you've sure as hell confused me

 

Until I read your last posting, I'd always thought that latitude and longitude were... well.. sort of absolute. If you were so many degrees north or south of the equator and so many degrees east or west of the meridian then THAT'S WHERE YOU WERE.

Are there really different size degrees or starting points other than the equator and the meridian?

If I've missed something fundamental and appear really thick then sorry but it's probably because I am

 

John

__________________________________________________________

 

The Team... A group of people who, individually, can do nothing but collectively.. decide that nothing can be done.

[+Huga]

Pharisee:

Yes, if you are somewhere, then you are a certain number of degrees north/south of the equator and a certain number of degrees east/west of the Greenwich Meridian.

 

BUT, we can't measure the lat/long of every point on earth....when GPS reports your position, it's relative to all the satellites, and then it projects that position to the surface of the earth. this requires the shape of the earth to be known...it's not spherical. It's slightly oblate (fatter around the equator).

 

So the different datums have different shaped earths - there's the Airy spheroid, the OSGB36 spheroid and the WGS84 spheroid, to name but a few. Each has a different shaped earth and slightly different starting points to fit the spheroid to the earth more accurately. The WGS84 zero-longitude line and the OSGB zero-longitude line are in different places, by about 100m or so.

 

OS maps are defined according to the OSGB36 (1936) datum, if I remember correctly, but the default navigation for worldwide stuff is the american WGS84 (1984..?) datum.

 

Make sense?

 

--

**Mother is the name of God on the lips of all children**

[+Pharisee]

quote:

---

Originally posted by Huga:

 

OS maps are defined according to the OSGB36 (1936) datum, if I remember correctly, but the default navigation for worldwide stuff is the american WGS84 (1984..?) datum.

 

Make sense?

 

---

 

Yep... Thanks for explaining it. (Trust the Yanks to play to different rules )

 

John

__________________________________________________________

 

The Team... A group of people who, individually, can do nothing but collectively.. decide that nothing can be done.

[+washboy]

quote:

---

I'm not sure if you've helped Paul but you've sure as hell confused me

---

 

Every time I think about this stuff it makes my head hurt

 

I hope Paul understood what I was talking about. The terminology I used was specifically related to the terms used in OziExplorer and may not be the best way to think about the general subject of datum offsets.

 

So, if you don't use OziExplorer, apologies and please ignore this guff

 

But if you're interested and fancy a bit of brain-twisting, have a look at A Guide to coordinate systems in Great Britain

 

=====

There's no such thing as a free lunchbox!

 

[This message was edited by washboy on December 12, 2002 at 07:09 AM.]

[+paul.blitz]

quote:

---

If you calibrated your map with:

 

Map Datum = Ord Srvy Grt Britn

Map Projection = (BNG) British National Grid

---

 

My map CALIBRATION uses OSGB/BNG, with the actual calibration points set using BNG (which I have selected as my alternate grid)

 

On the setups, in SYSTEM I set the Datafile Datum to WGS84; in MAPS I set datum=OSGB, Alt Grid=BNG; in GPS I set WGS84.

 

My Waypoint List I always use WGS84.

 

quote:

---

...then the bar (you mean the one just below the toolbar?) should say "Ord Srvy Grt Britn" to the right of the figures.

---

 

My bar shows:

 

" BNG WGS84"

 

(where the "WGS84" textis in blue, rest in black)

 

If I get Ozi to add grid lines (just found them!) then the grid is NOT aligned to the map, but is about 100m south, 200m east of the OS map's gridlines.

 

Now, the waypoints I have created (from Geocaching WGS84 positions) look to be positioned spot on on the display.

 

This says to me:

 

a) I loaded a map, and defined my calibration of that map to be using the correct OSGB positions

 

all the rest of stuff (waypoints, grids...) ends up using wgs84... but correctly "translated" to the map positions. I'm guessing this is because I told it to use WGS84 as its datafile datum?

 

Paul

 

Team Blitz

 

No, I gave YOU the spare batteries....

[+paul.blitz]

quote:

---

quote:

---

...then the bar (you mean the one just below the toolbar?) should say "Ord Srvy Grt Britn" to the right of the figures.

---

 

My bar shows:

 

" BNG WGS84"

 

(where the "WGS84" textis in blue, rest in black)

---

 

I am on my home PC now, and discovered that the top bar was showing "Ord Srvy Grt Britn"... so I turned on the grid, and it lined up with the map grid.... so I thought, and played.....

 

I then tried double clicking on the bar.... and hey, up popped a box to "select display datum". So I selected "WGS84", and not only does it then display "WGS84" in the bar, the grid jumped to be in the position I decribed in above message!

 

The waypoints do NOT move when you do that, just the displayed x/y locations, and the grid overlay!

 

So again this shows that the WAYPOINTS (all set to WGS84 on my system) are absolute, as is the map's datum (OSGB), and ozi "converts" between them.

 

The bar & grid can then use whatever datum you want!!!!

 

Wow, the guy who wrote it sure knows what he's doing!

 

Paul

 

Team Blitz

 

No, I gave YOU the spare batteries....

[+The Merman]

As I understand it, the OSGB is based on figures that were set down by the Astronomer Royal Sir George Airey in 1936.

Subsequent space mapping has shown his figures to be out by as much as 100 metres.

 

The OS would find it absolutely impossible to redo all the UK maps and as the OSGB does not rely on lat long but grid to find position, then it is for mapreading purposes irrelevant.

 

Until the outset of Satnav, the trusty compass was all you had, and was always good enough to take the average boyscout to within a few feet of what he was looking for when out orienteering.

 

I woke this morning and my boat was not rocking...for one horrid moment I thought I lived in a house!

[+washboy]

quote:

---

I then tried double clicking on the bar.... and hey, up popped a box to "select display datum".

---

 

I've been wanting just that facility - and it was there all along! Thanks for finding it, Paul

 

So, does the display datum difference explain why your York waypoint was misplaced? I got the impression that your waypoint was actually misplaced on the map (determined my the map's calibrated datum), not just that co-ords as displayed were out (determined my the map's display datum).

 

=====

There's no such thing as a free lunchbox!

 

[This message was edited by washboy on December 14, 2002 at 03:30 AM.]

[+paul.blitz]

quote:

---

I've been wanting just that facility - and it was there all along! Thanks for finding it, Paul

 

Yes, it IS a nice feature... I was just checking to see how good the new OS coords are on the cache pages, so moved the cursor to the given location.... 100m SE!! AHH!! set to OSGB, and then it was spot on!

 

So the Degrees locations are for WGS84, the grid ref is to OSGB!

 

So, does the display datum difference explain why your York waypoint was misplaced?

---

 

Probably.... it all seems fine now!

 

quote:

---

I got the impression that your waypoint was actually misplaced on the map (determined my the map's calibrated datum), not just that co-ords as displayed were out (determined my the map's display datum).

---

 

yes, that's right.... now I understamd everything a lot better, I am happy I got it all set up right!

 

If nowt else, I have learnt a lotfrom all this.. I was reading a VERY interesting article on the OS site last night, all about how the different grids / datums came about, how it all works etc.

 

Mind you, I gave up at section 6, when the matrix maths started!

 

Paul

 

Team Blitz

 

No, I gave YOU the spare batteries....

[el10t]

quote:

---

Originally posted by Team Blitz:

Mind you, I gave up at section 6, when the matrix maths started!

---

But that's when it starts getting interesting!

 

Rich

mobilis in mobili

[+jeremyp]

quote:

---

Originally posted by The Merman:

As I understand it, the OSGB is based on figures that were set down by the Astronomer Royal Sir George Airey in 1936.

---

George Airy lived from 1801-1892. He did not define OSGB36, but he did define the Airy1830 spheroid on which OSGB36 is based.

quote:

---

Subsequent space mapping has shown his figures to be out by as much as 100 metres.

---

Err no...

 

"Different from WGS84" does not mean "wrong". The differences are due to the use of a different spheroid which was designed only to fit the Earth in the region of the UK (the ITRS80 spheroid on which WGS84 is based was designed to be a reasonable fit over the whole Earth). Also the prime meridian for WGS84 is in a different place. This does not mean that either meridian is "wrong" after all, fo a long time the French got by on a prime meridian that went through Paris and not Greenwich.

 

In fact, because the WGS84 datum takes continental drift into account, its prime meridian is actually moving West across the UK (well actually the UK is moving North East).

 

(spheroid is a mathematical term for squashed football BTW).

 

-------

jeremyp

The second ten million caches were the worst too.

http://www.jeremyp.net/geocaching

[+The Merman]

Thanks Jeremy,

 

I was trawling my mind from way back and got it a bit wrong.

 

If we learn a little bit each day then that makes us better people. That is one of the great things about this forum!

 

Bye the way Jeremy was this written by you History of the Prime Meridian I came across it while doing research (Note to self..Research first then post, not the other way round!!)

 

I woke this morning and my boat was not rocking...for one horrid moment I thought I lived in a house!

 

[This message was edited by The Merman on December 14, 2002 at 02:28 PM.]

[+Teasel]

quote:

---

Originally posted by The Merman:

As I understand it, the OSGB is based on figures that were set down by the Astronomer Royal Sir George Airey in 1936.

Subsequent space mapping has shown his figures to be out by as much as 100 metres.

---

The figure I read was 20m, which is scarily good, considering these were the days before interferometry! The 100m you quote is possibly the (deliberate) difference between the WGS84 and OSGB36 datums.

 

To quote from the above OS article:

quote:

---

Hence, the overall size of the TRF still used for British mapping came to be derived from the measurement of a single distance between two stations on Hounslow Heath in 1784 - using eighteen-foot glass rods! The error thus incurred in OSGB36 is surprisingly low - only about 20 metres in the length of the country.

---

 

[GEEK]

(A 'datum' is a mathematical model which is defined, not measured, and therefore has no inaccuracies as such. A 'TRF' is a set of measurements intended to realise this model, eg by measuring the positions of thousands of concrete pillars, or dozens of satellites, or a single tide gauge, or...

 

For example, the WGS84 datum defines a 'squashed football' whose centre is the exact centre of mass of the Earth. But by observing the orbits of satellites, we can only measure the centre of the Earth to an accuracy of around 1cm. Not perfect, but good enough to enable the US government to build us a great Geocache Placing System.

 

So differences between datums are not errors. Rather the errors are the difference between a datum and its TRF. You can also introduce errors when converting between two datums, but that's a different thread!) [/GEEK]

[+Bill D (wwh)]

Jeremyp wrote:

quote:

---

George Airy lived from 1801-1892. He did not define OSGB36, but he did define the Airy1830 spheroid on which OSGB36 is based.

---

 

I'd known for years that OS maps were based on something called an Airy spheroid, and I'd vaguely wondered how a spheroid representing the earth could be in any way air-like...

 

Duh!!!

[+jeremyp]

quote:

---

Originally posted by The Merman:

Thanks Jeremy,

 

Bye the way Jeremy was this written by you http://joe.mehaffey.com/greenwich.htm I came across it while doing research (Note to self..Research first then post, not the other way round!!)

---

 

Nope, coincidence.

 

-------

jeremyp

The second ten million caches were the worst too.

http://www.jeremyp.net/geocaching