DSLRUsers.com

Bought a GPS System off EBay. It's the Yongnuo GP1 copy, and according to the 'blurb' it is recieving three or more satelites for info, and am not happy with its accuracy, its about 30k's off. Has anyone used one or have one that can tell me if they are accurate.
I have tested it against my car Navman and the car one is accurate in my backyard.
It tags your pics with gps references and then you can use these on Google Maps etc to see where you took the pic.

Hi Jon,

We use the Geo One from Phottix (Poon) and have no issues, but to be hnest, we've not yet tested it against any of the other GPS units that we have.

Out of curiosity, what sort of issues are you seeing?

One trap for young players is that the Geo One takes its power from the camera, and if the camera goes into stand by mode, then the GPS turns off. That means that you need to wait until it re-acquires a signal from the satellites again before it's again ready for use. You're not getting caught with inaccurate readings because of that sort of a situaton are you?

To protect against this sort of situation, adjust your stand by time cut-off in the camera.

J.Davis wrote:...its about 30k's off.

That's a lot. Which standard is it using for co-ordinates? There are a few and they will give you different results - although they shouldn't be 30k different.

J.Davis wrote: Has anyone used one or have one that can tell me if they are accurate.

I use a Holux M-241 which is both cheap and quite accurate. You could also check out GPS Geotagging options.

I use several, and have had at least one since about 1990. Generally they will give you accuracy to within 10 metres, and often better. BUT the accuracy is satellite dependent. One of mine was once out by over 2000km. I think one of the satellites went belly up at the time.

Modern units use 12 or more satellites to get a fix. They will work with only 3 or 4 but accuracy improves dramatically with more satellites. 3 will give a very poor fix. I would be suspicious that yours may be a copy of a very old ( and poor!) design. Possibly the Sony which had a very bad reputation.

Things to note.
Terrain will affect accuracy. heavy tree cover can disrupt accuracy by up to a kilometre as will nearby cliffs. They are almost useless in the concrete canyons of the inner city. I have been near Martin Place (Sydney) and had a modern GPSr tell me I was in Kensington - about 15km away.. Other times, I have been in the same spot & it's had me in the harbour.
I believe the car type ones do some special tricks to get a better signal in the city at the expense of performance elsewhere.

I also use the Holux Andrew has & regard it well, but mostly use the Garmin Oregon these days, though my older Garmin 12XL is often more accurate ( but much less convenient)

I use a Phottix Geo One as well....never tested it against other GPS units for accuracy....maybe on the next minimeet. I have noticed that Altitude can be way off on the Phottix unit, but I guess (and have been told) that altitude info is often compromised for better long/lat info.

YougNuo.....according to Poon, they OEM for Hahnel (Cleon knock-off remotes) and are not as good as Phottix, in my experience they dont look too bad in quality (the Hahnel remotes).

aim54x wrote:I have noticed that Altitude can be way off on the Phottix unit, but I guess (and have been told) that altitude info is often compromised for better long/lat info.

Yes. I believe the GPS altitude accuracy is about double the lat/long. So, if your GPS is telling you you have 6m accuracy, you would be getting about 12m accuracy for altitude.

Some GPSr's have an inbuilt barometer which is more accurate, but not without it's own problems.

Michael

I have adjusted the 'off timer' on the camera so the gps stays on while taking pics. It only indicates that it is seeing three satelites - even in an empty 100 acre paddock.
The reason I went for this one is the advertising - gp1 knockoff, its listed accuracy, and price.
The fact that it was a camera direct unit was a plus as I didn't have to have a seperate handheld GPS.
Could there be a secret in the way the co-ordinates are written to Google Maps? I have been entering them directly into Google Maps.I do not have software that reads the co-ordinates directly, only CS3.
Thanks for the replies so far.

J.Davis wrote:Could there be a secret in the way the co-ordinates are written to Google Maps? I have been entering them directly into Google Maps.

The GPS unit should be writing to the EXIF. There should be some software available that translates them back to Google Maps. I would Google for that. I would expect that Picasa, being Google software, should handle this.

Have you been able to compare the readings from your camera GPS and your normal GPS whilst in the field? Most traditional GPS units have a mode that lets you see the coordinates directly. By switching to that mode and then making an image and reading the coordinates of the image, you should get some good information. I may do some of that today with the D300s and, Geo One, a Navman, and an iPhone.

gstark wrote:The GPS unit should be writing to the EXIF.

If this is a "connect to camera" system then yes it does.

gstark wrote: There should be some software available that translates them back to Google Maps. I would Google for that.

Geosetter is the program that both Andrew (ATJ) and I use for this. Works a treat, though I notice that Andrew seems to have just discovered a bug in its daylight savings time handling.

J.Davis wrote:It only indicates that it is seeing three satelites - even in an empty 100 acre paddock.

J.Davis wrote:its listed accuracy, and price.

It sounds to me that you are suffering from a severe case of "you get what you pay for"
For the 'direct-to-camera' types, i know Cameron has good results from his Photix one, though it does eat the camera battery.
Both Andrew and I have had good results from the Holux one, and we have both checked accuracy against both Google Maps and other GPSrs. This requires post processing using, in our case, Geosetter. Stubbsy has also had good results with another brand, but I can't remember what that was. I also use my Garmin Oregon for the purpose, but it is less portable than the Holux (just slightly larger than a 35mm film canister), and considerably more

expensive

photohiker wrote:Yes. I believe the GPS altitude accuracy is about double the lat/long. So, if your GPS is telling you you have 6m accuracy, you would be getting about 12m accuracy for altitude.

.
This is true for all except dedicated aviation ones. I have noticed on my Garmin 12XL that it consistently reports an altitude about 20m higher than actual. My Oregon uses both barometric and Topo map corrections & is usually spot on. Never bothered to check the Holux.

Just noticed that Andrew mentioned "standards for co-ordinates"
This is a complicated topic. There are hundreds, if not thousands, and yes it can affect accuracy.
Basically, the Earth is an oblate spheroid with bumps. Surveyors and map makers have trouble dealing with this, so they break the whole thing down into a series of flat plates, called a geodetic model. They choose a reference point, and work out from there. Depending on how far you are from the reference point, the accuracy of the model, and how close you are to one of the plate boundaries, your apparent location can change quite considerably. Though never in my experience as far as you describe.

Google Earth and the Holux use a geodetic model called WGS84. This is a very accurate computer model based somewhere in the USA.
Others I have used are Australian Geodetic 84 and AG66. These are both Australian models using Obervatory Hill in Sydney as a reference point. The 84 one is more recent, and is very similar to WGS84. It varies by about 2cm from it. The 66 one will put you about 120m NE. (or is it SW . I can't remember) Canberra uses a different one based on Mt Stromlo. It is VERY accurate in the ACT, but not much use elsewhere.
Both my Garmins can specify which geodetic model to use. This can be important when referencing paper maps as they will be drawn to a specific Geodetic, and you need to match the Geodetics so both the GPS and maps use the same one.

If your GPSr is using an arcane Geodetic (e.g. Zanderij), it could explain the discrepancy, though I doubt it.

Mr Darcy wrote:Never bothered to check the Holux.

Mine is pretty poor for altitude. It has recorded altitudes of 20-80 metres (and even higher) while on dive boats on the open sea.

ATJ wrote:

Mr Darcy wrote:Never bothered to check the Holux.

Mine is pretty poor for altitude. It has recorded altitudes of 20-80 metres (and even higher) while on dive boats on the open sea.

Andrew,

Obviously, you were out on the high seas.

Garry - I have used the Camera and Car GPS's side by side and the readouts are different at the fourth didit.
Will try some sort of software and see the results.
The person that I bought it off is in contact with me so I might just have a dud.

Downloaded Geosetter and all is good, It can tell the difference from the road to the back yard, but I still don't know the difference in readouts between the different GPS Systems,but they all work as advertised.
Initially I must have been inputting the data incorrectly into Google Maps.
Thanks to all for your valued assistance.

J.Davis wrote:Garry - I have used the Camera and Car GPS's side by side and the readouts are different at the fourth didit.

Fourth digit? I'm not sure I understand.

The values are given in degrees, then minutes, then seconds, and then ... some part of a second. I'm sure that Andrew or Greg will offer their assistance here.

To my mond, and for most purposes, if the degrees and minutes are consistent, then I think that you're well within the boundaries of what's reasonable for many purposes. And probably for many porpoises as well.

From there, the seconds and subsequent values should be fairly close, and if they're not exactly the same but close, I would be happy with that sort of an outcome.

Now ....

I've just returned from a cappuccino and muffin at Bondi Beach. Actually, I'm not too sure it was Bondi; the humidity today made it feel more like Singapore. For company, I had an iPhone, a Nokia Navigator 6110, a Nikon D300, a Nikon D300s, and 2 Geo One GPS units. Each of these units returned very similar GPS readings, and were consistent enough for me to be satisfied with the accuracy of the Geo One units.

For instance ...

iPhone : 33deg 53'26'S, 151deg 16'36"E
Nokia : 33deg 53'26"63 S, 151deg 16'35'83E Altitude: 29.5m
GeoOne #1: 33deg 53.436'S 151deg 16.595'E Alt: 5-13m
GeoOne #2: 33deg 53.436'S 151deg 16.595'E Alt: 8-13m

Note that the readings on the camera are expressed differently to those from the iPhone and Nokia. Would somebody like to do a conversion of these for us? Jon, could that be a part of the issues that you're seeing: you've been entering a set of figures that are expressed in a different manner?

On the altitude readings, I wasn't able to get one from the iPhone. The basic coordinates are displayed on the Compass app. On the Geo One/D300, I've listed the range of altitudes that was saved with the images that I made. In the main, they were in the realm of 8-9 meters. I would have thought that where I was - a cafe just west of Queen Elizabeth Drive, outside the pavilion, that I would have been about 6 meters above sea level. The water's edge was within sight, and maybe 80 or so meters away. How do tides etc affect these reading? What is the definition of "sea level" within this context?

Clearly the Nokia's altitude value is not correct, but I think that everything else is consistent, reasonable, and usable for these purposes.

In doing this, I will also say that this was my very first play with the D300s, and as I was playing with some settings, I couldn't help but notice some interesting changes in the menus. I need to explore this a little more, but my first impressions are that this is more than just a bit of cosmetic surgery.

gstark wrote:iPhone : 33deg 53'26'S, 151deg 16'36"E
Nokia : 33deg 53'26"63 S, 151deg 16'35'83E Altitude: 29.5m
GeoOne #1: 33deg 53.436'S 151deg 16.595'E Alt: 5-13m
GeoOne #2: 33deg 53.436'S 151deg 16.595'E Alt: 8-13m

Note that the readings on the camera are expressed differently to those from the iPhone and Nokia. Would somebody like to do a conversion of these for us?

Ignoring altitude:

Code: Select all

iPhone   : 33º 53' 26" S    151º 16' 36" E
Nokia    : 33º 53' 26.63" S 151º 16' 35.83" E
GeoOne #1: 33º 53' 26.16" S 151º 16' 35.7" E
GeoOne #2: 33º 53' 26.16" S 151º 16' 35.7" E

gstark wrote:iPhone : 33deg 53'26'S, 151deg 16'36"E
Nokia : 33deg 53'26"63 S, 151deg 16'35'83E Altitude: 29.5m
GeoOne #1: 33deg 53.436'S 151deg 16.595'E Alt: 5-13m
GeoOne #2: 33deg 53.436'S 151deg 16.595'E Alt: 8-13m

Note that the readings on the camera are expressed differently to those from the iPhone and Nokia. Would somebody like to do a conversion of these for us? Jon, could that be a part of the issues that you're seeing: you've been entering a set of figures that are expressed in a different manner?

GPS Longitude and Latitude Converter One of many found on google..

Michael

And on the map: Gary's GPS Devices

Now I wonder which is most accurate according to Mr Stark.

Gary,
There are quite a few way of representing Latitude and Longitude.
The one everyone learns about at school is ddd mm ss.s, but the most common one I see in use these days is
dd mm.mmm, and dd.dddddd is also fairly common. (Decimal precision is arbitrary)
Anyone who uses paper maps will know the UTM method which is a grid reference, but directly translatable to Lat/Lon.
This looks like (for example) 56 H 0340634 6248637 (This is your Coffee stop a la the Geo One)
There are lots of others as well. These are the ones I have had personal dealings with.

Sea Level is Mean sea level, and ignores tides. It also ignores the actual shoreline too & is based on those geodetic plates I mentioned earlier.

A change of +1 minute in both Lat and Lon from Gary's coffee location will move you 41 metres (roughly) This distance will change as you move toward or away form the equator. How many porpoise lengths is that?

aim54x wrote:Now I wonder which is most accurate according to Mr Stark.

They are all within about five meters, which I think is pretty good, and certainly, I would not argue too much with them for the purposes of this discussion.

But one of them was pretty close to spot on.

Guesses, good people?

Thank you Andrew, Michael, and Greg for your conversions and explanations etc. The information you have provided is most useful, and helps us all to understand this information that is normally hidden from us.

Mr Darcy wrote:Sea Level is Mean sea level, and ignores tides. It also ignores the actual shoreline too & is based on those geodetic plates I mentioned earlier.

So the 8 meters or so may well be pretty close.

A change of +1 minute in both Lat and Lon from Gary's coffee location will move you 41 metres (roughly)

Not 42?

Seriously, is that both Lat and Lon to achieve that distance, or could it be either? I guess I'm asking is 1 minute lat 41 meters that way, and then 1 minute lon is 41 meters perpendicular to the first movement, or would that be 41 meters diagonally?

But in looking at the numbers from the units I was playing with this morning, they're all showing, by extrapolating from what you;re saying here, just a few meters variance in their locations, and that has been borne out by what we are seeing on the map.

This is quite fun!

gstark wrote:Not 42?

You have to move a little further north (each year ) for that.

gstark wrote:Seriously, is that both Lat and Lon to achieve that distance, or could it be either? I guess I'm asking is 1 minute lat 41 meters that way, and then 1 minute lon is 41 meters perpendicular to the first movement, or would that be 41 meters diagonally?

Diagon Alley and in that direction. It may be different in the opposite direction
I'll plug some numbers in later ( Am busy watching Star Trek ATM) and give you figures for Lat shift and Long shift separately.
The numbers change as the lines of longitude spread apart as you get closer to the equator. There is a point near the SOuth pole where you can walk one kilometre north, One Kilometre East then One Kilometre South & end up back where you started (On step 2 you circumnavigate the globe.)

gstark wrote:

aim54x wrote:Now I wonder which is most accurate according to Mr Stark.

They are all within about five meters, which I think is pretty good, and certainly, I would not argue too much with them for the purposes of this discussion.

But one of them was pretty close to spot on.

Guesses, good people? :)

No takers?

gstark wrote:

gstark wrote:

aim54x wrote:Now I wonder which is most accurate according to Mr Stark.

They are all within about five meters, which I think is pretty good, and certainly, I would not argue too much with them for the purposes of this discussion.

But one of them was pretty close to spot on.

Guesses, good people? :)

No takers?

The GeoOne? Only because it's closest to somewhere you might sit and sip a cappuccino...

Michael

photohiker wrote:The GeoOne? Only because it's closest to somewhere you might sit and sip a cappuccino...

Michael,

Two of the readings are mapped as being within the cafe's boundaries.


But yes, the GeoOne was pretty much spot on with where I was seated.

The iPhone mapped a few meters to the north, just within the outdoor seating area of the cafe.

The Nokia mapped east of where I was seated, onto the walkway in front of the cafe.

For many photographic purposes, I think that any of these might have sufficient accuracy, but the GeoOne seems, on the basis if this very simple test, to be very good.

I should probably perform similar tests at other locations to see just how reliable these initial results are. Coffee, anyone?