Re: Wrong timestamp (Hour!)
- From: "Bengt Thuree" <bengt thuree com>
- To: f-spot-list gnome org
- Subject: Re: Wrong timestamp (Hour!)
- Date: Tue, 9 Aug 2005 18:20:33 +0900 (JST)
Hej
First of all Larry: Congratulations to your expecting fatherhood :)
Welcome to the club...
Why should I not change the time in the camera to match the local time?
Right now I am in Ireland, and I am usually located in Australia so if I
do not change the time in the camera then it will be a bit weird comparing
the amount of sun light in the picture with the photo time.
I guess I must be missing something rather obvious here...
I guess we need a TimeZone field for every Import session (both from file
as well as from Camera). This should hopefully correct the problem.
Setting a timezone based on a camera is not good. Since the camera might
be traveling around the world a bit.
But, how would you present the time in the following example.
Home : Time Zone 0
Vacation : Time Zone -12 hr
You take some photos at home on the way to the airport, and then you fly
out to your vacation, where you spend two weeks. You take loads of photos
before you fly back home and upload all photos to your local computer
(with the home timezone active).
Will you show all the photos timestamp according to Vacation time, or
according to the Home time? (Sorry, I am a bit confused)
If you change the cameras time to reflect the local time, you have no
problems with the local time in the Photo itself.
If you do not change the camera's time, the photo will reflect 12 hours
wrong. Instead of showing a 13:00 it will show 01:00 (which will not make
sence since you are sunbathing)
I must be missing something fundamental here.
To me, it seems like you have to change the time in the camera, and then
when you import the photos, you either
1) use the local photo time as it is
or
2) use the local photo time together with the timezone you took the
picture in to convert it to a UTC time.
(The third option, which I will start to hack on later this week, is to
later go in and change the timezone for selected pictures)
Sorry if I am making things more confusing or rambling a bit here.
/Bengt
<citat vem="Larry Ewing">
> You would change the time in our camera to match the local time? That
> seems very unlikely to me. Or do you mean you would change the timezone
> of your computer each time and not the camera's time? Those are the
> cases where the timezone would end up being wrong for your own photos.
> The timezone info would clearly be wrong for photos from other people in
> different timezones under the current method. But without any timezone
> info they will always be wrong anyway.
>
> I'd like to be able to locate the time of the photo properly in the
> right timezone. Just using the exif date with no timezone information
> isn't ideal to me because in the long run it makes it more difficult to
> make correlations between events in your calendar and photos.
>
> I wonder if we let you set a per camera (perhaps based on serial number)
> and then stored both the exif time and the suspected timezone we could
> avoid most of the problems. Like I said before it needs some thought to
> be both simple and useful.
>
> --Larry
>
> On Sun, 2005-08-07 at 20:31 +0900, Bengt Thuree wrote:
>> Hej
>>
>> Checking a bit more, and I see that you convert the local time from the
>> EXIF data to UTC time format. This is based upon the timezone in the
>> computer you are using when you run F-Spot. The conversion changes the
>> correct EXIF timestamp to some wrong values.
>>
>> I do not think this is a good idea.
>>
>> Consider the following:
>> 1) You take some photos in Melbourne/Australia (with correct time in
>> camera)
>> 2) Board a plan to Ireland
>> 3) Take some photos in Singapore airport (with correct time in camera)
>> 4) Take some photos in Copenhagen airport (with correct time in camera)
>> 5) Arrive to Ireland, and take more photos (with correct time in camera)
>> 6) download the photos to your laptop wich is running either Melbourne
>> or
>> Irish time.
>>
>> In this case, as long as you do not know the timezone for each and every
>> photo, it is much better of just to take the photo time at face value.
>> Or?
>>
>> /Bengt
>>
>> <citat vem="Larry Ewing">
>> > This is the timezone causing problems. I need to go through the
>> current
>> > logic and figure out what to do.
>> >
>> > --Larry
>> >
>> > On Wed, 2005-08-03 at 21:53 +0900, Bengt Thuree wrote:
>> >> Hej
>> >>
>> >> I discovered that F-Spot has the wrong hour compared to the EXIF
>> data.
>> >> This is easily confirmed by opening a picture in "Image Viewer" and
>> >> checking the EXIF data.
>> >>
>> >> For instance, two of my pictures should have hours 18 and 13, but
>> >> instead
>> >> F-Spot reported the hours as 08 and 02.
>> >>
>> >> I also verified that the photo.Time.ToString ("MM") gives the same
>> >> faulty
>> >> result.
>> >>
>> >> I verified with the MetaDataExtractor assembly from Ferret Renaud,
>> and
>> >> this one reports the
>> >> correct timestamp in the EXIF fields.
>> >>
>> >> I am using the stock 0.0.13 version.
>> >>
>> >> Can this be confirmed by someone else, so I can write a TR, or is it
>> >> something else?
>> >>
>> >> /Bengt
>> >>
>> >> p.s.
>> >> I tried to go through the code to see where the picture gets the time
>> >> stamp from. But could only see that it took it from the unix file's
>> >> creation time, which is obviously not the correct place, since the
>> >> timestamp in F-Spot is the correct one apart from Hour.
>> >>
>> >
>> >
>> >
>>
>>
>
>
>
--
Bengt Thuree bengt thuree com www.thuree.com/bt
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]