GPS Error/accuracy.. I got long winded here, your warned

[Terence Dodge]

Greetings

Consumer GPS is and cannot be ever more than ( current Technology 2007, early 2007 at that )  3-4 meters, 3-4 minimum satellites ( yes more is better ) fairly open terrain, no urban canyons. If you have a GPS  with WAAS ( Wide Area Augmentation System, http://www.garmin.com/aboutGPS/waas.html  ), and not all currently available and sold due have WAAS, and it is more often than not in urban situations and as it uses ground stations to assist in accuracy/calibration and  if there is a large enough building or hill between you and that tower ( frequently near airports/harbor mouths, the original intended end users as a aide to navigation, as differential D-GPS/LORAN is phased out, yes they have a lobby so your grand children may still be paying for D-GPS/LORAN ).

Anyway unless you have WAAS map in memory ( you remember ) just remember it is sucking additional battery power and if you are out of a WAAS enabled area turning it the WAAS off in a menu will reduce your accuracy to 4 meters with the minimum satellites and foliage density.

 By the way spring foliage is denser that late summer/fall. What? Spring uptake of nutrients to leaf cover, then to seed, flower, fruit ( trees have to reproduce too ), why is this important to GPS for gods sake... microwave signal reception of our GPSr ( we seldom remember the "r" is for receiver) is diffused by plant density and moisture burden/load, takes longer to acquire a decent signal and if your GPSr is looking for a WAAS station that is not in range it will slow you down. You bid the job how? hourly no of course not. Imagine a double or triple canopy jungle with your trust GPSr.

Blue tooth GPS sounds fine and it may be OK for a few applications, setup is via a computer usually ( if you carry one in the field great ), error here is a issue when you first turn on your GPS it may take several to many minutes to acquire satellites, and as you linger in a spot the calculations refine to a point, now I have a D1x and it just relates that it has a signal not the quality of that signal nor the approximate accuracy ( size matters, is it a building, a building entrance, smaller than a Fiat 500, that is a small car for the younger members ), most of the BlueTooth GPSr have 2-3 LED indicating power, signal ( satellite acquired, not accuracy of that signal ),and maybe a  error LED. This why I have a Garmin GPSMap 60Cx mounted on my D1x ( moded a Garmin  bike cradle }, I can check accuracy, enable or disable WAAS with out the laptop, plus I have maps loaded into a memory chip in the thing. I can shoot in the rain, GPS is mounted on me with a long cable I made just for this purpose with the camera in its rain coat ( AquaTech ).

Track points which is the bread crumb metaphor for where you have been, you can store them but unless you are going to use them ( started here with the software comparison method ) do not plan on trying to average them so as to refine your site accuracy, it apparently does not work, kindly search some of the "geeks" GPS FAQ sites.

If someone wants to pay you serious money for accuracy ( meter or dramatic pause sub-meter, multi centimeter...) Leica GeoSystem ( Switzerland, hence the money ) GS20 ( $5000 ), will get you a basic meter capable accuracy unit, They can be rented and their software is windows only and it likely to remain that, do I care? Yes if you want the sub-meter accuracy which will cost you more ( additional hardware, external antenna and storage, batteries, after accuracy acquisition averaging ), plus in bidding the job you still have to spend time to acquire sufficient data, it still takes many minutes to acquire the data  for the refined accuracy calculations to be made either in the field or back in the office. Bidding a building in a city canyon or a gravestone/historic marker out doors ( clear landscape, no heavy tree cover ).

  I know of no one who has used this for photography, there is no cable made for it ( if you use a D1x without the "redhen.com" option one must roll their own to use the old colloquial expression, I had to ), however the specs for this thing ( GS20 ) do indicate that one can get the Nikon protocol needed ( NEMA, 4800 baud ) so the cable is quite doable.

Trimbal ( GPS mfg. for the survey crowd, like Leica, just not quite as expensive, they use windows in many of their GPS units ) just introduced GPS with a VGA camera built in and no it does not do camera raw, perhaps a firmware update.

GPS accuracy, think Sedan deVille, M1 Humvee..

Terence

[David C. Buchan]

GPS accuracy, think Sedan deVille, M1 Humvee..

Terence

Ah! Just like me taking a good photo. Extremely elusive.

This was a good introduction to the foibles of GPS Terence. And not too long winded.

David

[Dierk]

Not so much the long-windedness - the issue merits it - but the elusive style made me read the post mor[t]e than once.

Now, what's the actual contents? Are you saying, Terence, that non-military GPS is not accurate enough for our needs? Or are you merely plugging Leica's system? I am really interested, so don't see this as belligerent trolling, I just want to know what to argue.

[David C. Buchan]

Not so much the long-windedness - the issue merits it - but the elusive style made me read the post mor[t]e than once.

Now, what's the actual contents? Are you saying, Terence, that non-military GPS is not accurate enough for our needs? Or are you merely plugging Leica's system? I am really interested, so don't see this as belligerent trolling, I just want to know what to argue.

I simply found it interesting to know how inaccurate non-military systems are. I'm not planning on purchasing a GPS myself as I don't think I'd be a high enough user to warrant even a basic system (having borrowed my father-in-law's and found it cumbersome to use).

David

[johnbeardy]

I can't recall how, but didn't the Chinese embassy get hit in Belgrade?

Seriously, maybe I'm being really ignorant of the desire for greater accuracy, but isn't one major problem recording where one happens to be, not the direction you're pointing, or the angle of view? But a more important problem seems to be how to record such information so it's portable with the image, as it's difficult and/or undesirable to write to the camera-authored exif. So isn't there already an IPTC-XMP field for this GPS locations? I do see the need though - on this continent, new countries have popped up every few years and one re-united, so  the old Country/State/City/Location concept is too US-centric and too rigid over archival time.

John

[Dierk]

After Nikon other brands follow, even put the GPS receiver into their cameras; Ricoh is the first. Any receiver with an electronic compass gives us the position and direction. Longitude, latitude and bearing are, IIRC, official EXIF fields. Apart from biologists or architectures I can't think of a group needing more accuracy than 3-4 metres. Even the military would just adjust the load of their ammo* ...

For me the most interesting aspect is a quick mapping of my photos on a map/satellite view, particularly for future use, when one cannot remember where one shot a specific picture. I have several of those in my collection, including those from my father and grandparents. currntly this has to be done manually, but it is faster and easier with GPS data in the file.

A second feature not yet available is the automatic translation from a long/lat reading to the Location-Region-City-Country fields of IPTC/XMP. Just import your GPS-tagged photos into MediaPro/Expression Media or similar programs, they read the GPS data, check back with Google Earth/Map, write City, Region, Country into the appropriate fields. Depending on future amendments to IPTC/XMP the Location field could show the nearest road, for instance.




*A "clean" war may be thought of as a nice idea [personally I don't think so] but it is an illusion - unless one uses the Cold War tactic of hired assassin.

[johnbeardy]

For me the most interesting aspect is a quick mapping of my photos on a map/satellite view.

In Lightroom, there's a little arrow next to the GPS value - click it, and it goes to Google Earth.

John

[Terence Dodge]

Dierk wrote:

"Not so much the long-windedness - the issue merits it - but the elusive style made me read the post mor[t]e than once.

Now, what's the actual contents? Are you saying, Terence, that non-military GPS is not accurate enough for our needs? Or are you merely plugging Leica's system? I am really interested, so don't see this as belligerent trolling, I just want to know what to argue."

Greetings Dierk

my Elusive style derives from not writing much and being mostly a verbal, image then writing sort ( my wife, who writes well, shudders and likes to "vet" almost anything I commit to print, electronic or paper ).

It is not that "non-military GPS is not accurate enough for our needs" as those needs may very quite a bit. Client expectations, what they pay for and expect ( education component ) what they may get is where this is aimed ( guided missile versus rocket ). If you want the door or historic marker accurately inform the client about the known error ( presumes consumer GPS ), if it is a building or stadium, acceptable error is probably not worth mentioning.

Leica versus Trimabl. Most of these are large as in hang from shoulder or carry in pack on your back depending on antenna issues. Cost wise the Leica is generally more, if you check on Trimbal they have lots of models and if you equip/accessorize a Trimbal to compare base Leica the cost climbs be similar to the Leica ( computer metaphor: basic windows box versus a apple stock box, After you match hardware specs they often are close cost wise, TCO  "total cost of ownership" ). If the subject is physically dangerous to the photographer ( fire, animal ( includes human ), or just difficult to approach ( I use a wideangle lens so as to claim  "close accuracy" ), laser ranging binoculars that feed the distance from the object/subject to the GPS from your position and in post processing of the data from both devices can then be used to obtain accurate location of said object/subject, suspect this is mostly Military and or FEMA ( Federal Emergency Managment Agency, for those non USA forum members), I do not have that money.

As no one that I have read ( Yes I have not read all the posts on this list/forum ) on this forum has mentioned archaeology, forensic applications ( from limited conversations that I have been in, I know these two broad groups can get very picky about where something is or was photographically ). They may be lurking and do not have much to say, perhaps they have this DAM down to a science...

JohnBeardy Wrote:

"desire for greater accuracy" depends on you or your clients need versus expectations, People may want to find or go back to someplace they have seen a image of and depending on the size of that object ( Karl Marx grave marker, London, about 2 x 2 meters on a side. by 3 meters tall ), my point here is a lot of GPS users whom may look at photos for references to find things in the photo. Direction ( compass wise ) in EXIF data is only supported in the latest D2Xs I recollect, maybe wrong here it is not a tool I have, the D2Xs.

"So isn't there already an IPTC-XMP field for this GPS locations?" Yes there is and if you are careful on export the EXIF data does go with the file, I made this mistake earlier in a post because I did not understand the need for the JPEG 2000 plugin needed to be used on saving from CS2/Photoshop 9. I run a DNG into PS9 for a layer to display the GPS data and save out for uploading as a JPEG.

Our explanation for bombing the Chinese embassy ( at the time ) was that as we did not have anyone on the ground in Belgrade, so we used a old map that had government buildings noted, when is a smart bomb not smart, when we need to use it manually, kindly google Philadelphia Police helicopter bomb attack row house, a sterling example. We are not special here google "El Salvador vs Honduras, 1969: The 100-Hour War" 55 gallon barrels rolled out the door of a airplane. I digress.

Being able to get the data back into the IPTC/XMP after a visit to google earth would very helpful, particularly if I can then add that data to the keywords for referencing.

Terence

[DanZemke]

In Lightroom, there's a little arrow next to the GPS value ...

John - I have Lightroom V1 and can't find a field for GPS coordinates - looked in Metadata browser, Copy window and help.  Do you have a version beyond 1.0?

Dan

[Dierk]

If your photo file has the GPS data in your EXIF files, Lightroom will show them in the Library module, Metadata browser [right panel]. Select the tagged photo, scroll down in the panel until you hit upon the GPS field:

[DanZemke]

Dierk,

Thanks for letting me know that Lightroom doesn't display the GPS field unless there is something encoded in it.  That seems like a reasonable design decision to remove clutter, but I actually expected to be able to type in GPS info.  I also expected to be able to do this in Bridge and CS2 but found I couldn't.  I was surprised how long it took me to find a way to edit the GPS info fields.

Anyway, working now and the link to Google Maps is cool.  Thanks for the info.

Dan

[Dierk]

Since GPS is part of EXIF, it is a read-only field. As you found out there are - unfortunately - ways to edit EXIF data after taing the picture. When changing them one should bear in mnd that EXIF is intended as a protection against certain kinds of fraud.

[DanZemke]

What I did and what you do are nearly identical from a fraud prevention perspective - neither of our methods are trustworthy. Your Garmin unit doesn't output a copy-protected datastream.  So as far as your camera knows, it could just be a buffered keyboard that outputs a NMEA data. 

However, I do take your point about the intent of EXIF to be read only.  Since many would like to be able to easily add GPS location to their images, an IPTC field would be a better implementation than EXIF for "non-trusted" GPS stamping. 

Dan

[peterkrogh]

I agree that an IPTC, or maybe an XMP field would be a better choice.
Peter

[Jean-François Schmitz]

After having played around with different softwares to experiment geotagging, I came back to this thread, and found the last posts about EXIFs vs XMP vs IPTC very instructive! Thanks guys