Puzzled



B

Baz

Guest
Yesterday I planned a route in Derbyshire using Memory Map, the route
was 8.3 miles, I then uploaded the route to my Garmin Geko and went and
walked the route.
At the end of the walk the trip odometer on the Geko read 8.1 miles, but
the strange thing was, when I got home I downloaded the track to Memory
Map, MM showed the track to be 10.1 miles even though it followed the
route exactly.
Am I doing something wrong? Or just expecting too much from my gear?
 
A

Adrian Godwin

Guest
Baz <[email protected]> wrote:
> Yesterday I planned a route in Derbyshire using Memory Map, the route
> was 8.3 miles, I then uploaded the route to my Garmin Geko and went and
> walked the route.
> At the end of the walk the trip odometer on the Geko read 8.1 miles, but
> the strange thing was, when I got home I downloaded the track to Memory
> Map, MM showed the track to be 10.1 miles even though it followed the
> route exactly.
> Am I doing something wrong? Or just expecting too much from my gear?


You're expecting too much.

All those measurements are different, and the errors between them
can be quite large.

The route plan is the sum of all the straight-line segments you
drew on the map.

The odometer is the sum of all the differences between every
pair of position measurements, each one partially estimated
(unless you stopped every second to let the receiver go into
averaging mode).

The track distance is similar to the odometer but using only
a fraction of the measurements (because the receiver doesn't
store every point it measures).


-adrian
 
A

Adrian Godwin

Guest
Baz <[email protected]> wrote:
> Yesterday I planned a route in Derbyshire using Memory Map, the route
> was 8.3 miles, I then uploaded the route to my Garmin Geko and went and
> walked the route.
> At the end of the walk the trip odometer on the Geko read 8.1 miles, but
> the strange thing was, when I got home I downloaded the track to Memory
> Map, MM showed the track to be 10.1 miles even though it followed the
> route exactly.
> Am I doing something wrong? Or just expecting too much from my gear?


You're expecting too much.

All those measurements are different, and the errors between them
can be quite large.

The route plan is the sum of all the straight-line segments you
drew on the map.

The odometer is the sum of all the differences between every
pair of position measurements, each one partially estimated
(unless you stopped every second to let the receiver go into
averaging mode).

The track distance is similar to the odometer but using only
a fraction of the measurements (because the receiver doesn't
store every point it measures).


-adrian
 
D

Dominic Sexton

Guest
In article <[email protected]>, Baz
<[email protected]> writes
>Yesterday I planned a route in Derbyshire using Memory Map, the route
>was 8.3 miles, I then uploaded the route to my Garmin Geko and went and
>walked the route.
>At the end of the walk the trip odometer on the Geko read 8.1 miles,
>but the strange thing was, when I got home I downloaded the track to
>Memory Map, MM showed the track to be 10.1 miles even though it
>followed the route exactly.
>Am I doing something wrong? Or just expecting too much from my gear?


Firstly the odometer on the GPS will most likely under-read. The
software has to decided when the receiver is moving and when it isn't
but appears to be (position changes due to errors and changes in the
satellites used for the fix). Therefore it is highly likely that that
total track distance will be longer that the odometer reading.

Secondly the route you planned in Memory Map will be much straighter
than the track recorded by the GPS. If you zoom in you will see that the
GPS track will wiggle. If the GPS has a poor view of the sky (blocked by
your body, terrain, thick tree cover) this will be much worse than if it
has a good view of the sky. I would recommend carrying the GPS in the
top pocket of a rucksack, in a pouch on a shoulder strap or even under a
hat in this weather. That way it will be in a good position to make the
most of the satellites that are available.

These two will account for some of the discrepancy however the extra ~
25% you have seen here does seem to be larger than I would have
expected....

--

Dominic Sexton
 
D

Dominic Sexton

Guest
In article <[email protected]>, Baz
<[email protected]> writes
>Yesterday I planned a route in Derbyshire using Memory Map, the route
>was 8.3 miles, I then uploaded the route to my Garmin Geko and went and
>walked the route.
>At the end of the walk the trip odometer on the Geko read 8.1 miles,
>but the strange thing was, when I got home I downloaded the track to
>Memory Map, MM showed the track to be 10.1 miles even though it
>followed the route exactly.
>Am I doing something wrong? Or just expecting too much from my gear?


Firstly the odometer on the GPS will most likely under-read. The
software has to decided when the receiver is moving and when it isn't
but appears to be (position changes due to errors and changes in the
satellites used for the fix). Therefore it is highly likely that that
total track distance will be longer that the odometer reading.

Secondly the route you planned in Memory Map will be much straighter
than the track recorded by the GPS. If you zoom in you will see that the
GPS track will wiggle. If the GPS has a poor view of the sky (blocked by
your body, terrain, thick tree cover) this will be much worse than if it
has a good view of the sky. I would recommend carrying the GPS in the
top pocket of a rucksack, in a pouch on a shoulder strap or even under a
hat in this weather. That way it will be in a good position to make the
most of the satellites that are available.

These two will account for some of the discrepancy however the extra ~
25% you have seen here does seem to be larger than I would have
expected....

--

Dominic Sexton
 
D

Dave Fawthrop

Guest
On Tue, 22 Nov 2005 13:28:14 +0000, Baz <[email protected]> wrote:

| Yesterday I planned a route in Derbyshire using Memory Map, the route
| was 8.3 miles, I then uploaded the route to my Garmin Geko and went and
| walked the route.
| At the end of the walk the trip odometer on the Geko read 8.1 miles, but
| the strange thing was, when I got home I downloaded the track to Memory
| Map, MM showed the track to be 10.1 miles even though it followed the
| route exactly.
| Am I doing something wrong? Or just expecting too much from my gear?

You were drunk at the time and couldn't walk straight ;-)
--
Dave Fawthrop <dave hyphenologist co uk>
Sick and tired of Junk Snail Mail?
Register with http://www.tpsonline.org.uk/mpsr/
IME it works :)
 
D

Dave Fawthrop

Guest
On Tue, 22 Nov 2005 13:28:14 +0000, Baz <[email protected]> wrote:

| Yesterday I planned a route in Derbyshire using Memory Map, the route
| was 8.3 miles, I then uploaded the route to my Garmin Geko and went and
| walked the route.
| At the end of the walk the trip odometer on the Geko read 8.1 miles, but
| the strange thing was, when I got home I downloaded the track to Memory
| Map, MM showed the track to be 10.1 miles even though it followed the
| route exactly.
| Am I doing something wrong? Or just expecting too much from my gear?

You were drunk at the time and couldn't walk straight ;-)
--
Dave Fawthrop <dave hyphenologist co uk>
Sick and tired of Junk Snail Mail?
Register with http://www.tpsonline.org.uk/mpsr/
IME it works :)
 
G

Gordon Burns

Guest
"Dominic Sexton" <{d-sep03}@dscs.demon.co.uk> wrote in message
news:[email protected]
> In article <[email protected]>, Baz
> <[email protected]> writes


>
> These two will account for some of the discrepancy however the extra ~ 25%
> you have seen here does seem to be larger than I would have expected....
>
> --
>
> Dominic Sexton


When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
route on Memory Map that is about 12 - 14 miles long. When I come back my
tracklog usually reads about 16 - 17 miles. I put this down to the inability
to plot every little twist and turn on the map. That difference is not far
short of 25%!
 
G

Gordon Burns

Guest
"Dominic Sexton" <{d-sep03}@dscs.demon.co.uk> wrote in message
news:[email protected]
> In article <[email protected]>, Baz
> <[email protected]> writes


>
> These two will account for some of the discrepancy however the extra ~ 25%
> you have seen here does seem to be larger than I would have expected....
>
> --
>
> Dominic Sexton


When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
route on Memory Map that is about 12 - 14 miles long. When I come back my
tracklog usually reads about 16 - 17 miles. I put this down to the inability
to plot every little twist and turn on the map. That difference is not far
short of 25%!
 
P

Paul Saunders

Guest
Baz wrote:
> I planned a route using Memory Map, the route was 8.3 miles
> At the end of the walk the trip odometer on the Geko read 8.1 miles
> MM showed the track to be 10.1 miles
> Am I doing something wrong? Or just expecting too much from my gear?


Almost certainly down to bad reception at some point during the walk.
Tracks are usually longer than routes since routes tend to have lots of
straight lines between relatively few points. The odometer and the track
should be a pretty close match if reception is good. If not, it may be due
to losing reception (where the odometer stops incrementing) or due to
spurious points in the track that add extra distance that you didn't
actually walk.

Dominic Sexton wrote:
> Firstly the odometer on the GPS will most likely under-read.


Perhaps, but not by much in my experience.

> These two will account for some of the discrepancy however the extra ~
> 25% you have seen here does seem to be larger than I would have
> expected....


Way larger. I've never encountered anything remotely like that, assuming
good reception of course. If the reception is bad then I don't take any
notice of the numbers.

Gordon Burns wrote:
> When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
> route on Memory Map that is about 12 - 14 miles long. When I come back my
> tracklog usually reads about 16 - 17 miles.


Usually? If that's normal for you then you're either doing a lot of walking
in deep valleys or forests, or more likely, you're not placing your GPS
where it can get and keep a good view of the sky.

> I put this down to the inability
> to plot every little twist and turn on the map. That difference is not far
> short of 25%!


I put it down to bad reception. I'm sure that I never get results that bad
(assuming good reception) so I've just had a quick trawl through my diary
and picked out some sample data. I don't usually write the track length in
my diary (since the odometer is usually pretty reliable), but I found that I
did so in 2001 for a number of trips, so I've gathered the data for every
trip where I wrote both down.

8th Jan 2001 trip 4.3km, track 4.3km
12th Jan 2001 trip 9.1km, track 9.2km
3rd May 2001 trip 8.9km, track 8.9km
7th May 2001 trip 3.1km, track 3.3km
10th May 2001 trip 15.0km, track 15.0km (cycle)
12th May 2001 trip 12.0km, track 12.0km (cycle)
16th May 2001 trip 12.0km, track 12.0km (cycle)
20th May 2001 trip 9.7km, track 9.9km
24th June 2001 trip 15.4km, track 15.5km
25th June 2001 trip 13.2km, track 13.2km
9th Oct 2001 trip 4.3km, track 4.5km
17th Oct 2001 trip 7.2k, track 8.1k (Rhossili with Bill Grey)
24th Oct 2001 trip 7.9km, track 8.5km (Craig y Llyn with Bill Grey, Fran
etc.)

Out of 13 instances, the trip and track lengths match perfectly on 6 of
them. It's probably worth noting that three of those that match perfectly
were cycle rides, with the GPS positioned on the handlebars giving good
reception. Also the cycling was done in a good reception location, and the
relatively high speed and non-stop nature of the cycle rides may have
contributed to giving a more accurate result. These prove that the odometer
and track distances can match very accurately under good conditions.

On those that don't, the difference is only 0.1 or 0.2km, with the exception
of the last two, and even then, the difference is still less than a
kilometre, nowhere near 25%. I recall that the Rhossili walk involved a
deep valley, wooded in parts, so that was no doubt the cause of the bad
reception there, while the Craig y Llyn walk was done entirely in forestry
(but no deep valleys). In fact, there were only a couple of tiny breaks in
the last track, so it did surprisingly well there.

Paul
 
D

Dominic Sexton

Guest
In article <[email protected]>, Gordon Burns
<[email protected]> writes
>
>"Dominic Sexton" <{d-sep03}@dscs.demon.co.uk> wrote in message
>news:[email protected]
>>
>> These two will account for some of the discrepancy however the extra ~ 25%
>> you have seen here does seem to be larger than I would have expected....

>
>When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
>route on Memory Map that is about 12 - 14 miles long. When I come back my
>tracklog usually reads about 16 - 17 miles. I put this down to the inability
>to plot every little twist and turn on the map. That difference is not far
>short of 25%!
>


Well I still think that is a high percentage. On average you are talking
about an extra 25 metres per 100 metres on the map. Some of the walking
will be in a pretty straight line (almost no extra distance) so to get
an average of ~25% some of it must be very much more than 25%.

As an example if you were to walk 100m along two sides of a right angled
triangle instead of directly between the points you would walk an extra
59m = 42%.

There are certainly places where many people will zig-zag up or down
steep slopes but I doubt that the extra distances involved (that can't
be roughly plotted in MM or similar) outweigh the straighter parts of
the walk.

Unless you can easily see a lot of places where the planned and recorded
tracks have different lengths when you zoom in in MM I think it is a bit
fishy! I would be interested if you could compare the tracks with
another program to see if that agrees with MM.

It would be ideal if you could download the recorded track directly from
the GPS into another program (such as the freeware GPS TrackMaker
http://www.gpstm.com/ ) to eliminate the possibility that MM is doing
something funny to the data when it downloads it from the GPS.

--

Dominic Sexton
 
D

Dominic Sexton

Guest
In article <[email protected]>, Gordon Burns
<[email protected]> writes
>
>"Dominic Sexton" <{d-sep03}@dscs.demon.co.uk> wrote in message
>news:[email protected]
>>
>> These two will account for some of the discrepancy however the extra ~ 25%
>> you have seen here does seem to be larger than I would have expected....

>
>When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
>route on Memory Map that is about 12 - 14 miles long. When I come back my
>tracklog usually reads about 16 - 17 miles. I put this down to the inability
>to plot every little twist and turn on the map. That difference is not far
>short of 25%!
>


Well I still think that is a high percentage. On average you are talking
about an extra 25 metres per 100 metres on the map. Some of the walking
will be in a pretty straight line (almost no extra distance) so to get
an average of ~25% some of it must be very much more than 25%.

As an example if you were to walk 100m along two sides of a right angled
triangle instead of directly between the points you would walk an extra
59m = 42%.

There are certainly places where many people will zig-zag up or down
steep slopes but I doubt that the extra distances involved (that can't
be roughly plotted in MM or similar) outweigh the straighter parts of
the walk.

Unless you can easily see a lot of places where the planned and recorded
tracks have different lengths when you zoom in in MM I think it is a bit
fishy! I would be interested if you could compare the tracks with
another program to see if that agrees with MM.

It would be ideal if you could download the recorded track directly from
the GPS into another program (such as the freeware GPS TrackMaker
http://www.gpstm.com/ ) to eliminate the possibility that MM is doing
something funny to the data when it downloads it from the GPS.

--

Dominic Sexton
 
P

Paul Saunders

Guest
GSV Three Minds in a Can wrote:

> (you often get a really stupid point when
> you turn the GPS on and it gets an initial lock)?


Oh yes, forgot to mention that one. I got into the habit of avoiding that a
long time ago. When you first switch the GPS on, wait a little while until
it gets a good lock, THEN turn track recording on. If you don't, the track
usually leaps around like Zebedee on speed for the first minute or so.

Paul
 
G

GSV Three Minds in a Can

Guest
Bitstring <[email protected]>, from the wonderful person
Dominic Sexton <{d-sep03}@dscs.demon.co.uk> said
>In article <[email protected]>, Gordon Burns
><[email protected]> writes
>>
>>"Dominic Sexton" <{d-sep03}@dscs.demon.co.uk> wrote in message
>>news:[email protected]
>>>
>>> These two will account for some of the discrepancy however the extra ~ 25%
>>> you have seen here does seem to be larger than I would have expected....

>>
>>When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
>>route on Memory Map that is about 12 - 14 miles long. When I come back my
>>tracklog usually reads about 16 - 17 miles. I put this down to the inability
>>to plot every little twist and turn on the map. That difference is not far
>>short of 25%!


I see nowhere near 25% - rarely more than 5% in fact (using downloaded
track in Ozi/Tracklogs vs the GPS handset vs the route planned in
Tracklogs). You have to be WAY off course many times to get to 25%. I
suspect there is a bug in there somewhere .. maybe there is an initial
'way, way off' point which MM is using but which everyone else was smart
enough to throw away (you often get a really stupid point when you turn
the GPS on and it gets an initial lock)?

Yes, I've gone out on '20km walks' and come back having done 25, but in
those cases I can see where I backtracked, re-routed around problems,
etc. And both the GPS handset and the downloaded track (in Ozi) said
'25km' (+/- about 0.4, usually).

--
GSV Three Minds in a Can
Contact recommends the use of Firefox; SC recommends it at gunpoint.
 
G

Gordon Burns

Guest
"Paul Saunders" <[email protected]> wrote in message
news:[email protected]
> Gordon Burns wrote:
>> When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
>> route on Memory Map that is about 12 - 14 miles long. When I come back my
>> tracklog usually reads about 16 - 17 miles.

>
> Usually? If that's normal for you then you're either doing a lot of
> walking in deep valleys or forests, or more likely, you're not placing
> your GPS where it can get and keep a good view of the sky.
>
>> I put this down to the inability
>> to plot every little twist and turn on the map. That difference is not
>> far
>> short of 25%!


Perhaps I didn't clarify myself enough. When I plot a route I don't set out
to plot every little twist and turn. Unless the route I choose is going near
dangerous terrain I simply plot a straight line. eg I went from High Street
to Thornthwaite on Saturday. I plotted this as a straight line, even though
it is anything but. However getting lost on that stretch takes some doing
(although Carol Taylor once did, went north, and lost her life down by
Hayeswater). Then on the Yoke/Ill Bell/Froswick ridge I simply marked the
summits. Again the paths over there are so broad as to make getting lost
very difficult - but not impossible! Hence a shorter distance on the map
than on the ground. However I wouldn't advise a novice hillman or novice GPS
user to be so "casual".
 
G

Gordon Burns

Guest
"Paul Saunders" <[email protected]> wrote in message
news:[email protected]
> Gordon Burns wrote:
>> When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
>> route on Memory Map that is about 12 - 14 miles long. When I come back my
>> tracklog usually reads about 16 - 17 miles.

>
> Usually? If that's normal for you then you're either doing a lot of
> walking in deep valleys or forests, or more likely, you're not placing
> your GPS where it can get and keep a good view of the sky.
>
>> I put this down to the inability
>> to plot every little twist and turn on the map. That difference is not
>> far
>> short of 25%!


Perhaps I didn't clarify myself enough. When I plot a route I don't set out
to plot every little twist and turn. Unless the route I choose is going near
dangerous terrain I simply plot a straight line. eg I went from High Street
to Thornthwaite on Saturday. I plotted this as a straight line, even though
it is anything but. However getting lost on that stretch takes some doing
(although Carol Taylor once did, went north, and lost her life down by
Hayeswater). Then on the Yoke/Ill Bell/Froswick ridge I simply marked the
summits. Again the paths over there are so broad as to make getting lost
very difficult - but not impossible! Hence a shorter distance on the map
than on the ground. However I wouldn't advise a novice hillman or novice GPS
user to be so "casual".
 
P

Paul Saunders

Guest
Gordon Burns wrote:

> Perhaps I didn't clarify myself enough. When I plot a route I don't
> set out to plot every little twist and turn. Unless the route I
> choose is going near dangerous terrain I simply plot a straight line.


Oh I see, sorry, I misunderstood.

I'm the same with routes. I rarely use them but when I do I just mark a few
key locations, maybe only half a dozen for the whole walk. In most
circumstances navigation isn't difficult so just a few key points will do.
I've never been into planning a complex route that follows every twist and
turn. I've tried it, just for the sake of it, but I don't like that kind of
detailed GPS arrow following. It has it's place of course, at night and in
thick fog, but otherwise it's just too restrictive.

Paul
 
G

Gordon

Guest
Dominic Sexton <{d-sep03}@dscs.demon.co.uk> wrote
>In article <[email protected]>, Gordon Burns
><[email protected]> writes
>>
>>"Dominic Sexton" <{d-sep03}@dscs.demon.co.uk> wrote in message
>>news:[email protected]
>>>
>>> These two will account for some of the discrepancy however the extra ~ 25%
>>> you have seen here does seem to be larger than I would have expected....

>>
>>When I plot a route over the hills of Lakeland/Snowdonia I aim to get a
>>route on Memory Map that is about 12 - 14 miles long. When I come back my
>>tracklog usually reads about 16 - 17 miles. I put this down to the inability
>>to plot every little twist and turn on the map. That difference is not far
>>short of 25%!
>>

>
>Well I still think that is a high percentage. On average you are
>talking about an extra 25 metres per 100 metres on the map. Some of the
>walking will be in a pretty straight line (almost no extra distance) so
>to get an average of ~25% some of it must be very much more than 25%.
>

I think it is larger than any I've experienced, but I always examine the
track in Tracklogs, and move any obvious wild excursions into a more
sensible position by dragging the trackpoint, or else deleting it.
--
Gordon Harris
 
B

Baz

Guest

>
> It would be ideal if you could download the recorded track directly from
> the GPS into another program (such as the freeware GPS TrackMaker
> http://www.gpstm.com/ ) to eliminate the possibility that MM is doing
> something funny to the data when it downloads it from the GPS.
>


I got gps trackmaker and downloaded the track and route from my geko,
exactly the same distances in trackmaker as in MM, but something else
showed up which was not shown im MM, a 1.9 mile section of track at the
beginning of the track traveled at 308 mph, I think this must be caused
by the geko getting a false sat fix, and then having to correct itself?