Geotagging question

Started 6 months ago | Discussions
timo Veteran Member • Posts: 5,284
Geotagging question

I have just spent a week wandering around Hanoi (Vietnam) taking photos while tracking my movements with Geotag Pro on my iPhone. The process of loading the data into Lightroom and tagging the photos works fine.

Except that the locations are wildly inaccurate, despite my having carefully synchronised the clock on my cameras with my phone (and I have rechecked that since).

I realise that this time I set geotagging to use GPS and GSM data, rather than GPS alone. Given that phone coverage in Vietnam can be a bit sporadic, might this be the cause of the waywardness? Could GSM give a distorted location?

I'm posting this also on the Landscape forum - those guys must use geotagging all the time.

-- hide signature --

Tim
'If I were you, I wouldn't start from here ... '
http://timauger.smugmug.com
http://timauger.blogspot.sg/

 timo's gear list:timo's gear list
Pentax K-5 Pentax K-30 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GX8
a_c_skinner Veteran Member • Posts: 7,054
Re: Geotagging question

Can you examine the data from the phone to see if that is accurate or is it the tagging of the photos the problem?

-- hide signature --

Andrew Skinner

 a_c_skinner's gear list:a_c_skinner's gear list
Fujifilm X-T2 Fujifilm X-E3 Fujifilm XF 35mm F1.4 R Fujifilm XF 55-200mm F3.5-4.8 R LM OIS Fujifilm XF 27mm F2.8 +3 more
OP timo Veteran Member • Posts: 5,284
Re: Geotagging question

a_c_skinner wrote:

Can you examine the data from the phone to see if that is accurate or is it the tagging of the photos the problem?

The phone produces a track that is here and there not as accurate as it should be, particularly out of town, when inspected on the map on the computer. It's a couple of hundred metres out at times, sometimes a bit more, sometimes less.

Then, when the data is integrated with the metadata of the photographs, most of them are imprecisely located, sometimes quite wrong. IN the right city yes, but often out by several hundred metres. As I said, I have checked that the app and the phone clocks were synchronised, so that is not the problem.

The actual process of tagging goes smoothly. It's the results that are a bit haphazard. I haven't had this problem before, but I was probably using GPS only, before.

Hence my question.

-- hide signature --

Tim
'If I were you, I wouldn't start from here ... '
http://timauger.smugmug.com
http://timauger.blogspot.sg/

 timo's gear list:timo's gear list
Pentax K-5 Pentax K-30 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GX8
Alex Ridgway Regular Member • Posts: 268
Re: Geotagging question

timo wrote:

a_c_skinner wrote:

Can you examine the data from the phone to see if that is accurate or is it the tagging of the photos the problem?

The phone produces a track that is here and there not as accurate as it should be, particularly out of town, when inspected on the map on the computer. It's a couple of hundred metres out at times, sometimes a bit more, sometimes less.

Then, when the data is integrated with the metadata of the photographs, most of them are imprecisely located, sometimes quite wrong. IN the right city yes, but often out by several hundred metres. As I said, I have checked that the app and the phone clocks were synchronised, so that is not the problem.

The actual process of tagging goes smoothly. It's the results that are a bit haphazard. I haven't had this problem before, but I was probably using GPS only, before.

Hence my question.

So generally when set to GSM and GPS the phone will use GSM to get a quick fix (say within hundreds of metres) and then GPS to get a proper fix, if you camera is getting to location from the phone, then it's going to reflect the location that the phone thinks it is, which in turn is going to be down to how frequently the app / phone updates the GPS.  I'd expect this is down to either the app not updating frequently enough or due to the setting which has allowed it to fall back to GSM for its fix.

Also it's worth keeping in mind that the map you're looking at might not match up 100% due to differences in projecting and grid systems.

a_c_skinner Veteran Member • Posts: 7,054
Re: Geotagging question

Can you inspect the file of location data the phone produces?

-- hide signature --

Andrew Skinner

 a_c_skinner's gear list:a_c_skinner's gear list
Fujifilm X-T2 Fujifilm X-E3 Fujifilm XF 35mm F1.4 R Fujifilm XF 55-200mm F3.5-4.8 R LM OIS Fujifilm XF 27mm F2.8 +3 more
OP timo Veteran Member • Posts: 5,284
Re: Geotagging question

a_c_skinner wrote:

Can you inspect the file of location data the phone produces?

Yes, so far as I can tell it looks the same as it looks on the map on the computer. What you can't plot on the phone is where it thinks individual photos were taken - that's only visible once you have married them up in Lightroom.

-- hide signature --

Tim
'If I were you, I wouldn't start from here ... '
http://timauger.smugmug.com
http://timauger.blogspot.sg/

 timo's gear list:timo's gear list
Pentax K-5 Pentax K-30 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GX8
OP timo Veteran Member • Posts: 5,284
Re: Geotagging question

Alex Ridgway wrote:

timo wrote:

a_c_skinner wrote:

Can you examine the data from the phone to see if that is accurate or is it the tagging of the photos the problem?

The phone produces a track that is here and there not as accurate as it should be, particularly out of town, when inspected on the map on the computer. It's a couple of hundred metres out at times, sometimes a bit more, sometimes less.

Then, when the data is integrated with the metadata of the photographs, most of them are imprecisely located, sometimes quite wrong. IN the right city yes, but often out by several hundred metres. As I said, I have checked that the app and the phone clocks were synchronised, so that is not the problem.

The actual process of tagging goes smoothly. It's the results that are a bit haphazard. I haven't had this problem before, but I was probably using GPS only, before.

Hence my question.

So generally when set to GSM and GPS the phone will use GSM to get a quick fix (say within hundreds of metres) and then GPS to get a proper fix, if you camera is getting to location from the phone, then it's going to reflect the location that the phone thinks it is, which in turn is going to be down to how frequently the app / phone updates the GPS. I'd expect this is down to either the app not updating frequently enough or due to the setting which has allowed it to fall back to GSM for its fix.

Also it's worth keeping in mind that the map you're looking at might not match up 100% due to differences in projecting and grid systems.

What you say makes perfect sense and is what I suspected - that GSM positioning is a bit rough and ready: presumably it is based on whichever transmitting tower the phone is talking to, and that certainly won't be all that precise (although CSI makes it look miraculous on the TV ... ). Forcing the phone to depend on GPS might be better. But I can't work out whether there is a timing issue. I find that, say, two photos taken only a minute or two apart, are showing up in locations several hundred metres apart geographically. This could simply be the phone jumping around between different GSM transmitters I suppose.

-- hide signature --

Tim
'If I were you, I wouldn't start from here ... '
http://timauger.smugmug.com
http://timauger.blogspot.sg/

 timo's gear list:timo's gear list
Pentax K-5 Pentax K-30 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GX8
a_c_skinner Veteran Member • Posts: 7,054
Re: Geotagging question

If that is the reason then the track you see on the phone should be wrong, shouldn't it?

How often does the phone record a location? That may be the source of the error if you are moving and the location used is out of date.

-- hide signature --

Andrew Skinner

 a_c_skinner's gear list:a_c_skinner's gear list
Fujifilm X-T2 Fujifilm X-E3 Fujifilm XF 35mm F1.4 R Fujifilm XF 55-200mm F3.5-4.8 R LM OIS Fujifilm XF 27mm F2.8 +3 more
khunpapa
khunpapa Senior Member • Posts: 1,888
Re: Geotagging question

timo wrote:

I have just spent a week wandering around Hanoi (Vietnam) taking photos while tracking my movements with Geotag Pro on my iPhone. The process of loading the data into Lightroom and tagging the photos works fine.

Except that the locations are wildly inaccurate, despite my having carefully synchronised the clock on my cameras with my phone (and I have rechecked that since).

I realise that this time I set geotagging to use GPS and GSM data, rather than GPS alone. Given that phone coverage in Vietnam can be a bit sporadic, might this be the cause of the waywardness? Could GSM give a distorted location?

I'm posting this also on the Landscape forum - those guys must use geotagging all the time.

Not all phones are created equal.

I have 8 phones(wife, my offsprings, my mom, my sisters, etc). The one that's most accurate is only one, the Samsung Galaxy Tab A, limited edition. It can catches the GPS signal fastest (less tah 5 seconds) and most  accurate (within 10m). Three of SS galaxy phones need 30-60 seconds to catch the GPS, and 100m - 300m accuracy.

..

The Tab A went along with me from Hanoi to Hagiang  with perfect location even in the remote valleys that could barely see people in 10 km range.

-- hide signature --

Flashes of my Memory.

 khunpapa's gear list:khunpapa's gear list
Sigma DP2 Sigma DP1 Merrill Sigma DP3 Merrill Olympus E-1 Olympus E-M1 +23 more
mdmarqphoto Regular Member • Posts: 411
Re: Geotagging question

Could GSM give a distorted location?

Yes, that could easily cause that.  To get the location via GSM the phone uses triangulation, and the accuracy will be based on how many towers there are near by and how far away they are.  Any time I want the highest accuracy with my phone, I switch to GPS only (even though my phone says both GPS/Networks is more accurate).

OP timo Veteran Member • Posts: 5,284
Re: Geotagging question

a_c_skinner wrote:

If that is the reason then the track you see on the phone should be wrong, shouldn't it?

How often does the phone record a location? That may be the source of the error if you are moving and the location used is out of date.

The recording interval set for the phone was two minutes. The errors represent a greater discrepancy than that. The geographical track recorded is broadly correct but imprecise, and a bit 'off' in places. More so than I have experience before. Varies between just a few metres to maybe half a kilometre. This one might attribute to the use of gsm location rather than gps. What is more puzzling is that there also seems to be a timing discrepancy. Some of the photos are accurately placed; the inaccurate ones are wrong by minutes (it varies), maybe an hour or so in some cases, certainly not by days. And it is not consistent. Therefore I don't think the problem is wrong time setting in the phone vs the clock setting in the camera. I can only assume that my phone was only connecting with its coordinates intermittently (despite the setting of the geotagging programme), and the photos are being tagged with the last known coordinates, rather than an up-to-date one.

I shall have to experiment with the same phone over the weekend.

-- hide signature --

Tim
'If I were you, I wouldn't start from here ... '
http://timauger.smugmug.com
http://timauger.blogspot.sg/

 timo's gear list:timo's gear list
Pentax K-5 Pentax K-30 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GX8
a_c_skinner Veteran Member • Posts: 7,054
Re: Geotagging question

That seems a reasonable hypothesis and would explain why the route seems correct but the tagging isn't.

Having just spend a small fortune on a new camera (and indeed on a couple of other occasions in the last decade or so) I'm disappointed that GPS tagging isn't built in as a given.

-- hide signature --

Andrew Skinner

 a_c_skinner's gear list:a_c_skinner's gear list
Fujifilm X-T2 Fujifilm X-E3 Fujifilm XF 35mm F1.4 R Fujifilm XF 55-200mm F3.5-4.8 R LM OIS Fujifilm XF 27mm F2.8 +3 more
bullet1 Veteran Member • Posts: 7,092
Re: Geotagging question

a_c_skinner wrote:

That seems a reasonable hypothesis and would explain why the route seems correct but the tagging isn't.

Having just spend a small fortune on a new camera (and indeed on a couple of other occasions in the last decade or so) I'm disappointed that GPS tagging isn't built in as a given.

It seems few cameras these days have the GPS built-in. I happen to own one of them, the Canon 6D. In the Canon world, there are a few DSLRs with built-in GPS, such as the Canon 6D/6D Mark II, 5D Mark IV, 1DX Mark II and 7D Mark II. Many Canon DSLRs without a built-in GPS, are compatible with the add-on Canon GPS-E2 unit which can be attached to the hot-shoe to provide the GPS function.

Since Panasonic, Sony and other Canon cameras I own do not have a built-in GPS, I purchased the Canon GPS unit, GP-E2, to provide the standalone GPS logging during trips. Later the ExiftoolGUI is used to add the GPS info to the photos (RAW or JPEG) taken by cameras without the built-in GPS.

-- hide signature --

Nelson Chen
http://NelsonChenPhotography.com/
100% RAW shooter with Capture One Pro V11

 bullet1's gear list:bullet1's gear list
Panasonic FZ1000 Canon PowerShot G5 X Canon EOS 6D Sony a6000 Canon EOS M5 +21 more
khunpapa
khunpapa Senior Member • Posts: 1,888
Re: Geotagging question

a_c_skinner wrote:

That seems a reasonable hypothesis and would explain why the route seems correct but the tagging isn't.

Having just spend a small fortune on a new camera (and indeed on a couple of other occasions in the last decade or so) I'm disappointed that GPS tagging isn't built in as a given.

I have 2 cameras that have GPS. But decide to shut down this functionality forever. Not accurate & always consumes precious battery power.

The good smartphone of today has more accurate geoposition; also fast and consume less power.

Olympus has released the mobile software (app) that would automagicly synchronize the geoposition to  the pictures taken by their camera. Very accurate.

-- hide signature --

If one think something someone says is not right, why don't explain the right one directly to the OP with right knowledge instead of flaming the other? Don't use "OP knows nothing, can not separate right or wrong" as the launching pad for attacking people. Rotten sub-ego should be kept to one own mind.
--
Flashes of my Memory.

 khunpapa's gear list:khunpapa's gear list
Sigma DP2 Sigma DP1 Merrill Sigma DP3 Merrill Olympus E-1 Olympus E-M1 +23 more
a_c_skinner Veteran Member • Posts: 7,054
Re: Geotagging question

Exactly, we should be able to do better in devices that cost £1000.  I have got a phone app, but honestly it should just work.

-- hide signature --

Andrew Skinner

 a_c_skinner's gear list:a_c_skinner's gear list
Fujifilm X-T2 Fujifilm X-E3 Fujifilm XF 35mm F1.4 R Fujifilm XF 55-200mm F3.5-4.8 R LM OIS Fujifilm XF 27mm F2.8 +3 more
OP timo Veteran Member • Posts: 5,284
Re: Geotagging question - a little follow-up test

Just for interest I did a little test yesterday, which demonstrates how imprecise geotagging with the phone actually is.

In this shot you will see two (very faint, sorry!) green highlighted areas. All the shots were taken inside and outside the buildings within those two ovals - enough were outside to get a decent fix, one would have thought. At least 20 of the shots were taken in the upper area, not in the lower area. Not one of them shows up in the proper place, although the track does go to the upper area. There is obviously something awry with the way the app matches the times of the shots with the timing of the track.

All that zigzag stuff across the road is completely misleading, and my walking route off to the bus stop was in a straight line, not wandering from one side of the road to the other as the track leading off to the right. So if I need precision it seems that my phone is probably not the best way to achieve it. It was set to GPS only, logging one-minute intervals.

-- hide signature --

Tim
'If I were you, I wouldn't start from here ... '
http://timauger.smugmug.com
http://timauger.blogspot.sg/

 timo's gear list:timo's gear list
Pentax K-5 Pentax K-30 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GX8
mdmarqphoto Regular Member • Posts: 411
Re: Geotagging question - a little follow-up test
1

Inside buildings and among tall buildings outside is always tough for location tracking, even when just using your network.  I'm not surprised with your results at all.  Turning on GPS may have helped outside, but the taller buildings would likely have caused some issues as well.

Wayne Larmon Forum Pro • Posts: 10,062
Dedicated GPS logger just plain works
1

a_c_skinner wrote:

Exactly, we should be able to do better in devices that cost £1000. I have got a phone app, but honestly it should just work.

But they don't.  One thing that does work is a dedicated GPS logger.  It is designed for a single purpose.  It is dead easy to use (turn the power switch on when you want logging.  Turn it off when you are done.)  It has a long battery life (around 40 hours when logging.)

The tradeoff is expense (around $100) and some time spent getting everything working.  But this time spent is when you aren't shooting.    When you are shooting it is easy to use and is robust.

I described how I use mine in detail on another thread.

Wayne

Joseph V. Morris
Joseph V. Morris Contributing Member • Posts: 701
Not All GPS units are Created Equal

Perhaps this has already been brought up, but if not . . . Many years ago as an avid backpacker, I became fascinated about the prospects of usable GPS in a Smartphone.  I quickly found out that the quality of GPS implementations varied wildly and not just with the hardware but also the intended use.

There were a variety of sensors used in different smartphones and the orientation and sensitivity of the antennas also varied markedly.  The situation was further complicated by the fact that the reviews and literature on the various smartphones seldom if ever provided any usable information to decide on which one was best for my needs.  After a great deal of effort, I selected the Garmin-ASUS GARMINphone (and now you know how long ago I did this!).  Other phones did not work well (if at all) under even low densities of foliage, never mind indoors (residential).  The GARMINphone was still accurate indoors to the extent that I could even reliably tell whether I was in the front or the back of my townhouse when I took a reading.  It was almost as good as using a dedicated Garmin GPS unit.  So, I found a solution to my requirements and quite worrying about it -- until it came time to replace that old GARMINphone.  At that point, I discovered that there has been little apparent improvement in generic GPS implementations -- did not work well indoors or under foliage.

With the newer units, I discovered that, for mind-numbing entertainment, I could always look up the GPS history from a prior week to see just what countries I had mysteriously been transported to by an undocumented feature of my 'newer' GPS implementation.  Very handy when waiting in a doctor's office.

Now, I don't get out bushwhacking these days like I did in the past, so I've sort of lost touch with all of this, but is there some source that provides comparative information today on various GPS implementations?

-- hide signature --

Regards,
Joseph V. Morris

 Joseph V. Morris's gear list:Joseph V. Morris's gear list
Nikon D70 Nikon D7000 Nikon D7200 Nikon D500 Nikon AF-S DX Nikkor 16-85mm f/3.5-5.6G ED VR +16 more
OP timo Veteran Member • Posts: 5,284
Re: Not All GPS units are Created Equal

Joseph V. Morris wrote:

Perhaps this has already been brought up, but if not . . . Many years ago as an avid backpacker, I became fascinated about the prospects of usable GPS in a Smartphone. I quickly found out that the quality of GPS implementations varied wildly and not just with the hardware but also the intended use.

There were a variety of sensors used in different smartphones and the orientation and sensitivity of the antennas also varied markedly. The situation was further complicated by the fact that the reviews and literature on the various smartphones seldom if ever provided any usable information to decide on which one was best for my needs. After a great deal of effort, I selected the Garmin-ASUS GARMINphone (and now you know how long ago I did this!). Other phones did not work well (if at all) under even low densities of foliage, never mind indoors (residential). The GARMINphone was still accurate indoors to the extent that I could even reliably tell whether I was in the front or the back of my townhouse when I took a reading. It was almost as good as using a dedicated Garmin GPS unit. So, I found a solution to my requirements and quite worrying about it -- until it came time to replace that old GARMINphone. At that point, I discovered that there has been little apparent improvement in generic GPS implementations -- did not work well indoors or under foliage.

With the newer units, I discovered that, for mind-numbing entertainment, I could always look up the GPS history from a prior week to see just what countries I had mysteriously been transported to by an undocumented feature of my 'newer' GPS implementation. Very handy when waiting in a doctor's office.

Now, I don't get out bushwhacking these days like I did in the past, so I've sort of lost touch with all of this, but is there some source that provides comparative information today on various GPS implementations?

Very good question. I’ve spent most of this afternoon looking for one and not finding it. Actually there does not seem to be a huge range available, which is surprising.

-- hide signature --

Tim
'If I were you, I wouldn't start from here ... '
http://timauger.smugmug.com
http://timauger.blogspot.sg/

 timo's gear list:timo's gear list
Pentax K-5 Pentax K-30 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GX8
Keyboard shortcuts:
FForum MMy threads