Canon GP-E2 GPS unit

Is the manual for this device available as a PDF somewhere?

Manual isn't even listed on Canon's website. However the thread at http://blog.willshootp​hotography.com …mages-from-log-files.html can provide some useful information. Note many camera models need the latest firmware for the device to interface properly.

Any idea if I can set the time manually both on the GPS unit and 5D3 without having to synchronize with the satellite?

I'm planning to use it as standalone and tag the photo later. I read from one of the review site that you need to mount the GPS and synchronize with the camera everytime before you use it.

http://blog.willshootp​hotography.com …mages-from-log-files.html

Any GPS will set it's own clock each time it establishes its position. That's automatic and mandatory. It just can't work without knowing the time very accurately.

When used for logging, the timestamp for a logged position is compared to the timestamp for a taken photo. If there are logged positions for every second, then you have a good resolution there. If a certain picture was taken at 15:12:48, you know precisely where you were at that time.
If the log resolution is one minute, the logging software would have to interpolate between the GPS readings done at 15:12:00 and 15:13:00 to find the position, so it may be off by a bit, depending upon how you moved.
Likewise, if the clock in the camera is off by one minute, the picture taken at 15:12:48 will get the position recorded at 15:13:48, which may or may not be the correct position, depending upon your movements.

Since clocks today are pretty accurate, it will normally be perfectly sufficient if you set the camera's clock correctly once, then just use the camera for a while. If you see the camera's clock is drifting, you can just repeat the procedure as needed.

Something to try would be to really "dig" into the EXIF data and look for the tag "GPSTimeStamp" and then the camera time. The tag GPSTimeStamp is derived from the GPS satellite and of course the camera time is derived from the camera clock with the GPSTimeStamp as the accurate reference. The GPD related tags are in Universal Cooridinated Time and thus ignore time zone whereas the camera clock includes time zone if you keep it current. The camera clock tag is in the EXIF file as DateTimeOriginal .

Several software programs exist for "adjusting" the times...one at http://jpgtime.learsy.​com/ for example.

apersson850 wrote in post #14855870
Any GPS will set it's own clock each time it establishes its position. That's automatic and mandatory. It just can't work without knowing the time very accurately.

My understanding is that the whole GPS system works on sending timestamps backwards and forwards between the device and satellites and calculating position based on the the time lapse between send and receive for each satellite the device is referencing.

Well the GPS satellites constantly broadcast a code containing the time at which it was sent and the position in the satellite's orbit from which it was sent. GPS receivers receive this signal from potentially dozens of satellites in the GPS constellation and compare the differences in time along with the satellites' broadcasted positions to calculate exactly where they are within the shell created by the satellites' orbits. I find the concept fascinating

Correct. There is no information sent from the GPS navigation device to the satellites, but only the other way.

When a GPS starts operating, it's internal clock is usually not accurate. Thus when it computes the distance from a certain satellite, and compare that to the distances computed from other satellites, it will find that these spheres don't coincide in one single point. Instead their intersections can make up a virtual spherical pyramide, or some other difficult to describe geometrical figure.
The GPS receiver will then adjust its own internal clock until this figure gets minimized. In theory it would become a single point, but there's always some noise and tolerances. Anyway, when the volume of this figure is as small as possible, the clock is correct, and by that we mean down to a few nanoseconds.

This is why four satellites are needed to properly determine your position. For X, Y and Z you'd need only three, but in this case time is the fourth unknown.

From the FAA private pilot's examination...

What is the minimum number of Global Positioning System (GPS) satellites that are observable by a user anywhere on earth? Answer = 5.

How many Global Positioning System (GPS) satellites are required to yield a three dimensional position (latitude, longitude, and altitude) and time solution? Answer = 4