Set M8's clock to second?

[MarcRochkind]

Advertisement (gone after registration)

I'm interested in setting the clock to the second so I can shoot with a tracking GPS device while the camera is moving (e.g., from a car) and get an accurate match to the track points.

 

All I've come up with is this: While watching an accurate clock, move the M8's minute to the next minute on the setting screen. When the clock hits that minute, press the M8's set button.

 

This is OK, but it requires waiting up to a minute.

 

Also, it seems that there's no way to check the clock seconds without shooting a picture.

 

Has anyone come up with (1) a better way to set the clock, and (2) a way to view the time to the second without shooting a picture?

 

Nikon D200s can be set from a computer (although I haven't tried it). Do M8s have such a feature? (I never installed the software that came with my M8... maybe there's a feature I don't know about?)

 

--Marc

[scott kirkpatrick]

What's a minute? I set my Seiko watch that way, changing it only when the summer to winter time transition occurs, and it remains accurate to a few seconds over the half year. Remember that there is an unknown delay involved in receiving a time signal from NIST or whatever source you need, especially if you take it from the Internet. Have you tested to ensure that the setting procedure you follow starts you off at hh:mm:00 ? And how accurate is the M8 clock over periods like weeks or months?

 

scott

[rwfreund]

Scott,

Network Time Protocol which can be used to set the time on a pc, normalizes network delay and clock jitter.

It is typically less than 20 msec off.

see http://www.ee.udel.edu/~mills/database/reports/ntp-survey99-minar.pdf for a 1999 paper that surveyed time hosts on the net.

-bob

[rwfreund]

Marc,

It appears that Leica capture does not provide this facility.

Hacking the time the way you mentioned works well enough for my gps uses,

You might try taking a picture of your receiver's clock display and then use a tool such as robogps to adjust the offset.

-bob

[scott kirkpatrick]

Scott,

Network Time Protocol which can be used to set the time on a pc, normalizes network delay and clock jitter.

It is typically less than 20 msec off.

see http://www.ee.udel.edu/~mills/database/reports/ntp-survey99-minar.pdf for a 1999 paper that surveyed time hosts on the net.

-bob

 

That's an interesting link. It's an MIT course paper surveying NTP clocks that concluded that only 28% of the tier 1 time sources available to a low level query gave useable results. I use

Current local time in Jerusalem - Israel here in Jerusalem, and a service hosted at NIST in the US, so I assume that there can be delays of 1-2 seconds before my laptop, which receives the time signal over HTTP through an academic network and then my DSL modem displays a specific time. That's still good enough for me to get to my class on time or predict subway train arrivals, and should work for mapping gps data to exposures. By contrast, a research project that I am part of uses GPS clocks on packet transceiver cards to get 10 nanosec precision doing point to point packet train transmission studies across Europe.

 

scott

[rwfreund]

Scott,

I set up my domain server to use a tier two time server and thus becomes a tier three.

My domain clocks in my home network are very close to accurate, but I have not been able to precisely figure out by how much.

My pc time is close enough to the time calculated by my gps receiver such that I can't see visually any difference between my pc clock and my gps clock.

 

I think that the tier one that I use is using gps time as well.

There are a few products that will allow you to use a gps receiver as a time source, but I have not yet been motivated to explore them since my time base seems to be pretty good.

 

All this is pretty funny since I have no reason at all to be any more precise in my personal schedule than a minute or two at the worst case. Must be my anal german ancestry.

-bob

[MarcRochkind]

Advertisement (gone after registration)

It appears that Scott is right... there is NO way to set the M8's clock to the second. When I bring up the setting menu and position the minute to the next minute and then press the Set button at exactly that time, the Leica's seconds are completely unaffected.

 

In other words, it appears that setting the minute sets only the minute. It has no effect on the seconds.

 

Can someone verify this?

 

Imagine: A German clock that can't be set! ;-)

 

--Marc

[DaveB]

The CBS radio network tone leading into the news broadcast has always been a highly reliable indicator for the top of the hour. Automated broadcast equipment synchronizes to this tone. No concern about internet delay with over the air broadcasting.

 

I cannot vouch for it, but I believe all the BBC top of the hour news broadcast are similarly introduced by a tone.

 

I've always kept my watch synchronized with the CBS clock since my first radio job which required my to talk up until the beginning of the news. Now, many years after leaving radio, I still maintain the habit.

 

This, I'm assured, can now be effectively treated with medication.

[marknorton]

Here's a picture, cropped from a much wider shot of the RTC (Real Time Clock) chip they use, made by Ricoh with a 4-wire serial interface apparently to the Intel processor. The gold cylinder is the crystal, probably running at 32.768kHz.

 

The chip has an on board calibration register for fine tuning and allows the second register to be set to a specific value; when it is, the sub-seconds counter is reset so that setting the clock should include resetting the seconds register to 0.

 

So, IMHO, changing the time should reset seconds to zero; if it doesn't (I haven't tried it), it's a firmware bug.

 

[ATTACH]39858[/ATTACH]

[scott kirkpatrick]

Well, I tried it, and it seems to reset the time to the hours and minutes selected, plus 06 seconds. I got that result several times. The experiment is:

 

press set time

press set again so the hours and minutes show

at exactly hh:mm:00 on my watch, press set the final time.

wait 15 seconds, and shoot a picture.

press play, and info. the time shown is hh:mm:21. Perhaps that is the time at which the picture is written out to the SD card.

 

scott

[cbretteville]

The camera records seconds in the EXIF info as pr. the specification. Its an ascii string, so no way of checking what it really reads. Looking at file timestamps on the memory card doesn't really help as FAT has a 2 second resolution as its max (DOS legacy from the early 80s).

 

I agree with Mark that this needs to be reported, but: If I was the project manager assigning priorities to issues, this would get a low rating.

 

- C

[fourfa]

be aware the your GPS track may not reflect correct time either:

 

"While most clocks are synchronized to Coordinated Universal Time (UTC), the Atomic clocks on the satellites are set to GPS time. The difference is that GPS time is not corrected to match the rotation of the Earth, so it does not contain leap seconds or other corrections which are periodically added to UTC. GPS time was set to match Coordinated Universal Time (UTC) in 1980, but has since diverged. The lack of corrections means that GPS time remains synchronized with the International Atomic Time (TAI).

 

The GPS navigation message includes the difference between GPS time and UTC, which as of 2006 is 14 seconds. Receivers subtract this offset from GPS time to calculate UTC and specific timezone values. New GPS units may not show the correct UTC time until after receiving the UTC offset message. The GPS-UTC offset field can accommodate 255 leap seconds (eight bits) which, at the current rate of change of the Earth's rotation, is sufficient to last until the year 2330."

 

(Global Positioning System - Wikipedia, the free encyclopedia)

 

Modern GPS receivers correct for leap seconds, the one I have does not. There have been 14 leap seconds since the satellites launched. Double check your receiver.

 

Camera makers really ought to make this level of precision easy to set. For instance shoot a wedding with two cameras, then in editing sort by date/time (to account for different filename sequences). Unless the cameras are exactly synched, this can get confusing when the action gets quick.

[MarcRochkind]

Since the time can be incorrectly set even on cameras that allow setting (apparently the M8 does not, except to the minute), I've decided not to worry about the M8. These are the two situations that arise in ImageIngesterPro:

 

1, For GPS tagging, there's a simple interface for setting a time adjustment. You pick an image, and its EXIF time is shown. Then you use another control to set the actual time. The app does the subtraction for you.

 

With an M8, or any camera, you just shoot an image of a reasonably accurate clock. I use a so-called "atomic clock" (it's not really atomic, of course), but even my computer clock is accurate to a second. Then I choose that image, set the actual-time control to what the image shows, and I've got the adjustment, which is automatically applied to each image before its (adjusted) time is looked up amongst the GPS track points.

 

2. The other place is the multi-camera-shoot feature of ImageIngesterPro. Here the absolute time doesn't matter, as long as the cameras are synchronized. There's an interface for adjusting each set of images, and then the rest is automated.

 

So, the lack of precision in the M8's time setting isn't a problem, just an annoyance. Still, I do find it ironic that the clock doesn't zero the seconds when you set the minute.

 

--Marc

[scott kirkpatrick]

So, the lack of precision in the M8's time setting isn't a problem, just an annoyance. Still, I do find it ironic that the clock doesn't zero the seconds when you set the minute.

 

--Marc

 

If you read several posts back, I did the obvious experiment, and the result, consistent with Mark Norton's understanding of how the clock chip functions, is that the seconds are zeroed when you set the time.

 

However, the time recorded with each image seems to be the time that it is written, not when it is shot. That could introduce an error in mapping shots to gps time if the gps time is taken from the hot shoe as the picture is shot and then offset, since the delay in writing could depend on the number of shots in the pipeline when you are shooting fast or continuously.

 

scott