[Lowell Goudge]

So you're reading the time right off the GPS?

Thank you
Russell

No actually,, the way it works in geotagger, is you set up the photos you want to tag (in a directory usually), tell it where teh GPX data files can be found, and then go through the set-up for time sync, where you enter the time zone your camera thinks it is in, and any offset correction, assuming the camera is not exactly set to your GPS time (i find it drifts a bit)

The program then shows what your camera time and day should read right now. and you check.

then you use these settings and do a trial picture, and it will put a pinpoint on google map where the photo was taken. If all this is correct, then you tell it to sync all the files in the directory.

You can tell it to either look for exact times, or interpolate between time points because not everyone sets up their gps to record every 1 second.

It took me about 10 minutes to figure out, and from then on, it is no problem, the only issue is that when you cosnider new high MP cameras, to save the data (separate operation) tagging and saving several hundred or thousand shots may take all night.

The idea of taking oa photo of the GPS and the time of day is not as good as taking a shot at a known place, perhaps outside your hotel in the morning, because then there is no confusion. It is all about having the place exact.

[Nimrad]

Is there no need to edit the time in GPS loggers? Does it also get the time settings from the satelites?

[Derridale]

OK - my Phototrakr Mini arrived today, and is charging in the USB port as I type this. I've downloaded and installed the Mac software. Now all I have to do is work out how to match up the GPS data from the Phototrakr and the photos.

As far as I can tell from initial glance at the instructions (such as they are) is that I have to set up an "album" folder, copy the photos into there, then run the Phototrakr software which will match up the GPS data to the photos. Of course, the Phototrakr and the computer must be time-synced to within a couple of seconds.

Once I've actually tried to do it, and to get the photos into Aperture after geotagging, I'll post how well it works.

Of course, to read my RAW files, I need to upgrade the software ($25 I think) to the Pro version. For anyone using JPGs, that isn't necessary. I usually shoot in RAW using DNG files, so after playing with a few JPGs for this exercise, and if it all works, I'll upgrade to the Pro version.

[Russell-Evans]

Is there no need to edit the time in GPS loggers? Does it also get the time settings from the satelites?

The time stamp is from the satellites and is going to to be more accurate than any positional data. I personally think correlating between the time display captured on a photo of a sync time source with the data in the photo EXIF, is the better solution.

Thank you
Russell

[Russell-Evans]

Of course, the Phototrakr and the computer must be time-synced to within a couple of seconds.

Actually this is incorrect. You only need to correlate the times stamps in the tracking log to the camera's time stamp in the EXIF data. You need to use the EXIF data because the camera does not display seconds on any of its screens. At least not the K10D and K20D. I don't have experience with the other bodies so verify with your model.

Thank you
Russell

[Lowell Goudge]

Is there no need to edit the time in GPS loggers? Does it also get the time settings from the satelites?

that is correct. the way a GPS works is that it measures the time difference in receiving the time signals from each of the satelites.

[Lowell Goudge]

Actually this is incorrect. You only need to correlate the times stamps in the tracking log to the camera's time stamp in the EXIF data. You need to use the EXIF data because the camera does not display seconds on any of its screens. At least not the K10D and K20D. I don't have experience with the other bodies so verify with your model.

Thank you
Russell

True and false. You can set the times within a second easily because the seconds are reset to zero when you set time, and you can set the time within a second of the gps moving to the next minute, however it is not relevant to when you take a shot and the need to be that syncronized. Unless you are travelling in a car, where at 60mph you are going 88 feet per second, walking around your speed is measured in 3-4 feet per second, and in a whole minute, you might at a brisk walk, move a whopping 300 feet, but at that pasce it is doubtful that you are taking photos.

Just how accurate do you need to be with your GPS coordinates any way?

[Russell-Evans]

Just how accurate do you need to be with your GPS coordinates any way?

Enough so that I don't end up having to hand tag a group of photos.

Thank you
Russell

[Lowell Goudge]

Enough so that I don't end up having to hand tag a group of photos.

Thank you
Russell

you should not have to hand tag any shots, the whole issue here is a function of the software that matches the GPS coordinates to the photos. GeoSetter allows you to interpolate between points up to programmable time limits. if you don't move for several hours, (let's imagine sitting in a resturaunt for example) your GPS may not record very frequently, so geosetter allows (if you set it) a fairly large window for interpolation between track points, and can figure this out.

It is nothing to do with the GPS but how well thought out the software is.

I can't comment on other software packages, because GeoSetter was free, and has done everything needed, so I never looked elsewhere.

[Entropy]

A photo of a clock set to GPS time is the simple solution to time correlation. You have the GPS time captured and you have the time stamp to the second in the EXIF data. Now you can correlate the two. I'm not sure how you are correlating your data to the camera's clock?

I looked at an iblue tracker since you posted it and it looks like a nice solution. I just found the AGL3080 in my price range when I was looking about a year ago and bought it. I just don't have a reason now to spend again for a solution right now, as I'm happy with the performance.

Thank you
Russell

That's the pain-in-the-butt solution to time correlation as it is a manual operation. The easy solution (assuming someone else has already done the automation step for you, which many people have) is:
Pull photo timestamp from EXIF data
Search in GPS tracklog for that time
If there is a large gap in time AND distance between tracklog points, interpolate the position between those points
Put the position associated with that time into the EXIF data

This can be done because any GPS tracklog I have ever seen is timestamped appropriately.

Unlike taking a photo of the GPS clock frequently, all of the above steps can be automated. Software that includes this capability includes:
GeoSetter
digiKam (for Linux/Unix users unless you're hardcore and running KDE for Windows)
I believe Aperture and Lightroom
gpicsync
Plenty more

My workflow is:
Download pictures from SD to a folder
Download the tracklog from my gps datalogger
Go to that folder in digiKam
Choose the GPS Correlator function from digiKam's menus (I may have the exact name wrong)
Load the GPS
Click "Apply"

The only reason to take a picture of the GPS clock would be for the purposes of having a backup method for ensuring that the camera clock was correct. (Compare EXIF to picture content). If you're doing that, you may as well just check the camera clock.

[Russell-Evans]

Unlike taking a photo of the GPS clock frequently,

snipping ...

The only reason to take a picture of the GPS clock would be for the purposes of having a backup method for ensuring that the camera clock was correct. (Compare EXIF to picture content). If you're doing that, you may as well just check the camera clock.

Is once a day when starting out, frequently? You can't just check the camera clock as the camera only displays to the minute. So is 59 seconds important?

Assuming the average speed of walking for most people is 3 miles per hour, 60 seconds is 264 feet. My GPS has a stated accuracy of about 9 feet in the clear. It actually can be really really good in the clear and it can be within a foot. Yes, I like it when it is down to that precision. I just think it is cool when it happens and it hasn't gotten old yet. When doing the standard tourist stuff, I'm walking, taking a car, a bus, a train, walking, taking a car, bus, train, cab, walking, .... 59 seconds could be anywhere from .5 to 1 mile then.

If you get the results you are looking for, great. For me I like having a photo once a day of the clock to be sure that I can correlate the time between the camera and the GPS. It really isn't that big a deal people.

Thank you
Russell