Long time no see....

After using my trusted Meridian platinum for about 6 or 7 years it is time
to update. I might consider the garmin gpsmap 62 st, but only if it is more
accurate. Using official Dutch benchmarks my Meridian never was off by more
than 3 meters using EGNOS. (I can repeat my tests again and again and they
are consistant) The garmin gpsmap 62 st has to equal that or do better if I
want to buy it.

I'm looking for measured (not claimed by garmin) accuracy data of the garmin
gpsmap 62 st. SA Watch or Visual GPS plots would be fine.

I have googled for hours but have found nothing... at... all.

Even gpsinformation.net (which has a nice article about this gpsr) does not
mention anything about accuracy. Nice story about the menu's etc. but no
measurements.

Hmm, am I missing something here? The whole point of a gpsr is to pinpoint
position as accurately as possible? Or am I being naive?

Anyway: does anybody have accuracy measurements/tests of the garmin gpsmap
62 st? Help would be greatly appreciated!

TIA, Vincent

On 2011-06-27 17:58 , Vincent van der Laan wrote:
> Long time no see....
>
> After using my trusted Meridian platinum for about 6 or 7 years it is time
> to update. I might consider the garmin gpsmap 62 st, but only if it is more
> accurate. Using official Dutch benchmarks my Meridian never was off by more
> than 3 meters using EGNOS. (I can repeat my tests again and again and they
> are consistant) The garmin gpsmap 62 st has to equal that or do better if I
> want to buy it.

I can assure that your GPS is only as accurate as local conditions. I
too can get 2 - 3 meter SBAS (WAAS) benchmarks if the sky is clear of
trees and other obstructions....

But leave my GPS in a forested area and let it record for 10 minutes and
the walk can cover 50 meters radius.

> I'm looking for measured (not claimed by garmin) accuracy data of the garmin
> gpsmap 62 st. SA Watch or Visual GPS plots would be fine.
>
> I have googled for hours but have found nothing... at... all.
>
> Even gpsinformation.net (which has a nice article about this gpsr) does not
> mention anything about accuracy. Nice story about the menu's etc. but no
> measurements.
>
> Hmm, am I missing something here? The whole point of a gpsr is to pinpoint
> position as accurately as possible? Or am I being naive?

Reception conditions can have a great effect on accuracy. Don't assume
when walking in the woods, or on a trail on the side of the mountain, or
in the urban canyons of the city that you are getting the same accuracy
as at a "clear sky view" benchmark.


--
gmail originated posts filtered due to spam.

> I can assure that your GPS is only as accurate as local conditions. I
> too can get 2 - 3 meter SBAS (WAAS) benchmarks if the sky is clear of
> trees and other obstructions....
Which GPS receiver do you have?

> Reception conditions can have a great effect on accuracy. Don't assume
> when walking in the woods, or on a trail on the side of the mountain, or
> in the urban canyons of the city that you are getting the same accuracy
> as at a "clear sky view" benchmark.
>
I have been using GPS for 10 years now (and have read all articles from sam
Wormley about multipath, VDOP, HDOP, error sources etc.) so I totally agree!

I would like information on multipath handling as well.
And also: does the garmin do auto averaging when standing still?

I can't find any accuracy related technical specs.
Not even on Garmin's website.

Cheers, Vincent

On Tue, 28 Jun 2011 11:01:06 +0200, Vincent van der Laan wrote:
[....]
> I would like information on multipath handling as well. And also: does
> the garmin do auto averaging when standing still?
>
> I can't find any accuracy related technical specs. Not even on Garmin's
> website.

This may be urban legend, and if not probably requires
installations with a l-o-o-o-n-n-gg *fixed* baseline (originally between
MIT and CalTech iirc; and before geostationary satellites, I think, so
they may have had a third point on the Moon) -- but fwiw the technology
seems to've developed out of an experiment designed to allow measuring
continental drift. In inches per century. That's with the sky the limit
on cost, weight, everything else, of course; but afaik they did measure
it.

It's plausible insofar as NASA once boasted an ability to aim a
beam from Houston into a coffee cup on the Moon. And triangulation is an
exact science .... <grin>



--
Beartooth Implacable, PhD, Neo-Redneck Linux Convert
What do they know of country, who only country know?

On Jun 28, 2:01*am, Vincent van der Laan <vinc...@allalin.nl> wrote:
> > I can assure that your GPS is only as accurate as local conditions. *I
> > too can get 2 - 3 meter SBAS (WAAS) benchmarks if the sky is clear of
> > trees and other obstructions....
>
> Which GPS receiver do you have?
>
> > Reception conditions can have a great effect on accuracy. *Don't assume
> > when walking in the woods, or on a trail on the side of the mountain, or
> > in the urban canyons of the city that you are getting the same accuracy
> > as at a "clear sky view" benchmark.
>
> I have been using GPS for 10 years now (and have read all articles from sam
> Wormley about multipath, VDOP, HDOP, error sources etc.) so I totally agree!
>
> I would like information on multipath handling as well.
> And also: does the garmin do auto averaging when standing still?
>
> I can't find any accuracy related technical specs.
> Not even on Garmin's website.
>
> Cheers, Vincent

On every Garmin I own, you have to select averaging. If you dig
through old posts on this group, averaging itself isn't as good as you
think. Now most of these GPSs will do some weighted average of the
satellites that they can see, based on SNRm but that is not like
averaging readings.

To measure accuracy, you really need to find a landmark that has been
measured accurately.

http://www.ngs.noaa.gov/cgi-bin/ds_mm.prl

It has been my experience that a great deal of these landmarks are not
exactly easy to access. Some are in the middle of the street! The
spots that can be seen from satellite are my choice. You can look at
them on google earth to determine access. You also want to make sure
you can linger on the spot if you want to do averaging. On my
gps60csx, using English units, I get readings good to 4ft with
averaging. Not spectacular, though it is sub-meter.

The satellite visible landmarks show up as a big X. Those google earth
bastards trespassed on some undeveloped land I own and set up a
landmark. [Yeah, I was surprised to see a big X show up on google
earth on my property.] The lines of the X appear to be north-south and
east-west. Anyway, if you see an X on google earth, it may not be a
permanent marking. The ones in the USGS database are painted, but the
survey points used by google earth are done with some sort of
biodegradable sheeting. Not as flimsy as paper, but it does decay.

A reviewer shows a track on a satellite photo, compared to the
solution from an iPhone 4, at this site:

http://www.geek.com/articles/gadgets...vice-20100810/

Apparently walking around a New York apartment complex with a lot of
foliage in some areas.

Worth reading the comments. One commenter links to his expanded
remarks here:

http://forums.groundspeak.com/GC/ind...=253992&st=188

Note that all of this is nearly a year old, and there have undoubtedly
been software updates since then.

Sounds like the unit may have implemented a baro bias state in its
filter, which was common in the early days of GPS user equipment. A
lot of papers were written in the early 1980s, and maybe earlier, on
navigating with 3 satellites and a calibrated baro altimeter.

More comments on accuracy and the altimeter input here (again, almost
a year old):

http://gpstracklog.com/2010/06/garmi...announced.html

One web post asserts that the chipset is from ST Microelectronics, not
SiRF:

http://gpstracklog.com/2010/06/garmi...announced.html

STM is a bit stingy with details:

http://www.st.com/internet/automotive/class/1547.jsp

http://www.st.com/internet/com/SALES...rtesio1007.pdf

On Jul 4, 1:10*pm, "Ed M." <pat_n...@yahoo.com> wrote:
> A reviewer shows a track on a satellite photo, compared to the
> solution from an iPhone 4, at this site:
>
> http://www.geek.com/articles/gadgets...ap-62-handheld...
>
> Apparently walking around a New York apartment complex with a lot of
> foliage in some areas.
>
> Worth reading the comments. *One commenter links to his expanded
> remarks here:
>
> http://forums.groundspeak.com/GC/ind...=253992&st=188
>
> Note that all of this is nearly a year old, and there have undoubtedly
> been software updates since then.
>
> Sounds like the unit may have implemented a baro bias state in its
> filter, which was common in the early days of GPS user equipment. *A
> lot of papers were written in the early 1980s, and maybe earlier, on
> navigating with 3 satellites and a calibrated baro altimeter.
>
> More comments on accuracy and the altimeter input here (again, almost
> a year old):
>
> http://gpstracklog.com/2010/06/garmi...-62st-announce...
>
> One web post asserts that the chipset is from ST Microelectronics, not
> SiRF:
>
> http://gpstracklog.com/2010/06/garmi...-62st-announce...
>
> STM is a bit stingy with details:
>
> http://www.st.com/internet/automotive/class/1547.jsp
>
> http://www.st.com/internet/com/SALES...URCES/MARKETIN...

I don't expect any phone to be as good as a dedicated GPS. Smartphones
are Swiss Army knives. You cram a lot of electronics in them and put
them on a power budget too. I can tell you the GPS 60 compass sucks. I
set my up to use it rarely. You can set the break point between
between virtual compass (difference the GPS reading) and the flux gate
compass. Before I took it out in the back country, I did a few tests
and discovered there was really no way to make the compass be
accurate, certainly not when compared to a lensatic compass. I hope
the 3d compass is better, but still you don't get a "sight" with a
GPS, unlike the lensatic. If I need a vector, I use a real compass. I
have to laugh at those satellite pointing apps using the iphone. We
know the iphone GPS isn't great, and the compass is really really crap
You need at least two degree accuracy to set up a satellite dish
quickly. Those lensatic compasses are military quality at consumer
prices.

I see the 62 is more sensitive than the 60. Hey, time marches on. I'm
not exactly ready for an upgrade. I've run the 60 through some canopy
and it seems to do well. (Not so for my old emap). I like everything
about the 60 series except for the display. Black and white is far
better as a reflective display than color. I got a jacket for the
GPS60, but it is tough to read with plastic over the LCD. What worked
best for my emap was the yellow booty and a screen protector.

Supposedly there are iphone apps that contain map databases in memory.
I don't know anyone who has tried them. I can tell you as a GPS, the
iphone 3 is really crap. Maybe the iphone 4 sucks less, but as a
product, the iphone 4 is pretty crappy, well other than a gaming
platform. The iphone is way to flimsy for any serious backcounty
hiking. The stupid glass screen breaks all the time. I think all
phones should have screen protectors, but on a typical phone, you just
get a scratch on the screen. With the iphone, it shatters. The last I
looked, they want $200 for the Apple "genius" to replace the glass.

On 2011-07-04 16:10 , Ed M. wrote:
> A reviewer shows a track on a satellite photo, compared to the
> solution from an iPhone 4, at this site:
>
> http://www.geek.com/articles/gadgets...vice-20100810/
>
> Apparently walking around a New York apartment complex with a lot of
> foliage in some areas.

I did a similar comparison some time ago between an AMOD 3080 (track
recorder) and my iPhone 4.

Some areas had good sky view and others a lot of foliage.

The iPhone track was 10 - 20 metres off (or more) for most of the track
(a 30 minute walk).

When I would go to an easily recognized point (Google Earth) and mark it
with the AMODS 3080, the error would be about 1 - 2 metres. (The AMOD
3080 is a WAAS receiver, so that performance is no surprise). The AMOD
could be off by 10 or more metres when I had heavy foliage to the south
of me.

That of the iPhone 4 could be over 20 meters.

From the recording I get the impression that the iPhone updates are 0.5
Hz or less. But that could have been the application I was using that
was under sampling. No way to tell.

All that said I think the iPhone 4 GPS is fine for what it is - it gets
the urban/sub-urban/extra-urban user to where he needs to go. It also
tracks quickly with a first est. position within 50 metres or so and
then that improves quickly to 10 - 20 metres.

For photography (recording position to tag photos), the AMOD 3080 is
great. (And I have a few available for purchase at a great price if you
need one. Bought some for a project and still have 3 new in their boxes).

--
gmail originated posts filtered due to spam.

On Jul 6, 2:50*pm, Alan Browne <alan.bro...@FreelunchVideotron.ca>
wrote:
> On 2011-07-04 16:10 , Ed M. wrote:
>
> > A reviewer shows a track on a satellite photo, compared to the
> > solution from an iPhone 4, at this site:
>
> >http://www.geek.com/articles/gadgets...ap-62-handheld...
>
> > Apparently walking around a New York apartment complex with a lot of
> > foliage in some areas.
>
> I did a similar comparison some time ago between an AMOD 3080 (track
> recorder) and my iPhone 4.
>
> Some areas had good sky view and others a lot of foliage.
>
> The iPhone track was 10 - 20 metres off (or more) for most of the track
> (a 30 minute walk).
>
> When I would go to an easily recognized point (Google Earth) and mark it
> with the AMODS 3080, the error would be about 1 - 2 metres. * *(The AMOD
> 3080 is a WAAS receiver, so that performance is no surprise). *The AMOD
> could be off by 10 or more metres when I had heavy foliage to the south
> of me.
>
> That of the iPhone 4 could be over 20 meters.
>
> *From the recording I get the impression that the iPhone updates are 0.5
> Hz or less. * But that could have been the application I was using that
> was under sampling. *No way to tell.
>
> All that said I think the iPhone 4 GPS is fine for what it is - it gets
> the urban/sub-urban/extra-urban user to where he needs to go. *It also
> tracks quickly with a first est. position within 50 metres or so and
> then that improves quickly to 10 - 20 metres.
>
> For photography (recording position to tag photos), the AMOD 3080 is
> great. *(And I have a few available for purchase at a great price if you
> need one. *Bought some for a project and still have 3 new in their boxes).
>
> --
> gmail originated posts filtered due to spam.

Just a FYI, google earth is often off a bit. That is why I suggested
using the USGS reference points. I suppose if you aren't in the US,
they aren't too handy.

In the scheme of things, the average phone user rather than better
phone reception than better GPS reception. Especially true for any of
the iphones, which universally have poor reception. Hey, there is
always iphone 5. Apple is bound to get it right sometime. When antenna-
gate hit, they advertised for antenna/RF engineers. A bit late, but at
least they are attempting to fix the problem.

On 2011-07-07 23:25 , (E-Mail Removed) wrote:
> On Jul 6, 2:50 pm, Alan Browne<alan.bro...@FreelunchVideotron.ca>
> wrote:
>> On 2011-07-04 16:10 , Ed M. wrote:
>>
>>> A reviewer shows a track on a satellite photo, compared to the
>>> solution from an iPhone 4, at this site:
>>
>>> http://www.geek.com/articles/gadgets...ap-62-handheld...
>>
>>> Apparently walking around a New York apartment complex with a lot of
>>> foliage in some areas.
>>
>> I did a similar comparison some time ago between an AMOD 3080 (track
>> recorder) and my iPhone 4.
>>
>> Some areas had good sky view and others a lot of foliage.
>>
>> The iPhone track was 10 - 20 metres off (or more) for most of the track
>> (a 30 minute walk).
>>
>> When I would go to an easily recognized point (Google Earth) and mark it
>> with the AMODS 3080, the error would be about 1 - 2 metres. (The AMOD
>> 3080 is a WAAS receiver, so that performance is no surprise). The AMOD
>> could be off by 10 or more metres when I had heavy foliage to the south
>> of me.
>>
>> That of the iPhone 4 could be over 20 meters.
>>
>> From the recording I get the impression that the iPhone updates are 0.5
>> Hz or less. But that could have been the application I was using that
>> was under sampling. No way to tell.
>>
>> All that said I think the iPhone 4 GPS is fine for what it is - it gets
>> the urban/sub-urban/extra-urban user to where he needs to go. It also
>> tracks quickly with a first est. position within 50 metres or so and
>> then that improves quickly to 10 - 20 metres.
>>
>> For photography (recording position to tag photos), the AMOD 3080 is
>> great. (And I have a few available for purchase at a great price if you
>> need one. Bought some for a project and still have 3 new in their boxes).
>>
>> --
>> gmail originated posts filtered due to spam.
>
> Just a FYI, google earth is often off a bit. That is why I suggested
> using the USGS reference points. I suppose if you aren't in the US,
> they aren't too handy.

I know. But when you test against a number of easy to identify points
at different locations you tend to get a feel for where there are
obvious errors and where there are not. I can also get Canadian ref
point data if needed - but then I have to translate them from older
spheroids to WGS-84. PITA. (There are online converters, of course).

> In the scheme of things, the average phone user rather than better
> phone reception than better GPS reception. Especially true for any of
> the iphones, which universally have poor reception. Hey, there is
> always iphone 5. Apple is bound to get it right sometime. When antenna-
> gate hit, they advertised for antenna/RF engineers. A bit late, but at
> least they are attempting to fix the problem.

The iPhone 4 GPS implementation is thin - even if "better" than in
earlier iPhones. In order to reduce battery consumption they use a
'weak' correlator and from what I observe in the data they don't sample
and process as often as a dedicated GPS. And of course it still uses
cell tower aiding (you can see this when you start a map program -
initial error circle is very wide (100 - 200m). Then once it latches on
to a few satellites the error circle shrinks quickly).

Given how well it performs for general navigation (where am I and where
is the restaurant) the compromise they've made for power seems
appropriate. In the next iPhone (which I won't buy) I hope they don't
compromise power to improve GPS - it's not worth it.

I do wish there was an App to allow me to drill down into the channel
data and log that, however.

--
gmail originated posts filtered due to spam.