Menu
▾
▴
libexif-devel
|
From: Lutz <lu...@us...> - 2003-10-27 22:43:44
|
Hello! I merged libmnote into libexif. Canon maker notes now should be loaded and saved correctly (see test/test-mnote). That is, you should no longer experience any data loss when using libexif on your canon EXIF data. I have yet to make exif aware of the new functionality.=20 I haven't fixed the pentax and olympus maker notes yet.=20 Regards --=20 Lutz M=FCller <lu...@us...> |
|
From: Roberto C. <rob...@en...> - 2003-10-30 21:17:53
|
Lutz Müller wrote: > > Hello! > > I merged libmnote into libexif. Canon maker notes now should be loaded > and saved correctly (see test/test-mnote). That is, you should no longer > experience any data loss when using libexif on your canon EXIF data. > > I have yet to make exif aware of the new functionality. > > I haven't fixed the pentax and olympus maker notes yet. > > Regards > -- > Lutz Müller <lu...@us...> Hello, I'm very glad to receive such a good news and to see that the mailing list is now back to normal, thanks for your work! I haven't had yet enough time to look at it closely as I would like to, in the meantime a couple of questions naturally come into my mind: What happens to a picture with a maker note not recognized by libexif? Since the format of maker notes is usually not disclosed by camera manifacturers and not completely known, which are the assumptions libexif currently does about it in order to prevent loss of data? Cheers, Roberto |
|
From: Lutz <lu...@us...> - 2003-10-30 21:30:14
|
On Thu, 2003-10-30 at 22:14, Roberto Costa wrote: > I haven't had yet enough time to look at it closely as I would like to, > in the meantime a couple of questions naturally come into my mind: > What happens to a picture with a maker note not recognized by libexif? libexif keeps the data in the MakerNote. If any data is stored somewhere else, it will be lost. By the way, if you save MakerNotes that are recognized by libexif, all data is now stored in the MakerNote tag (as it should have been from the very beginning). --=20 Lutz M=FCller <lu...@us...> |
|
From: Roberto C. <rob...@en...> - 2003-11-03 20:07:57
|
Lutz Müller wrote: > > On Thu, 2003-10-30 at 22:14, Roberto Costa wrote: > > I haven't had yet enough time to look at it closely as I would like to, > > in the meantime a couple of questions naturally come into my mind: > > What happens to a picture with a maker note not recognized by libexif? > > libexif keeps the data in the MakerNote. If any data is stored somewhere > else, it will be lost. I'm not convinced that is an acceptable behaviour the fact that a libexif user that processes his/her images with the library may lose information because of lack of knowledge about his/her camera maker notes. My idea is that libexif should always be conservative and do its best not to corrupt any user exif info. When it's not the case, I think at least the user should be clearly warned of the risks derived from using libexif. > By the way, if you save MakerNotes that are recognized by libexif, all > data is now stored in the MakerNote tag (as it should have been from the > very beginning). I agree that it is a big mistake, nevertheless exif specs don't say that MakerNote tag must be self-contained and that it can't refer to data allocated outside its boundaries. I think it's a bug in the specs we should deal with. Cheers, Roberto |
|
From: Lutz <lu...@us...> - 2003-11-03 20:25:36
|
On Mon, 2003-11-03 at 21:04, Roberto Costa wrote: > Lutz M=FCller wrote: > I'm not convinced that is an acceptable behaviour the fact that a > libexif user that processes his/her images with the library may lose > information because of lack of knowledge about his/her camera maker > notes. > My idea is that libexif should always be conservative and do its best > not to corrupt any user exif info. Agreed. So, if we encounter MakerNotes other than Canon, Olympus and Pentax, we'd like to preserve data between IFDs at its original offsets. The easiest way (I guess) to accomplish that would be to preserve the original EXIF data, to append our new/modified EXIF data at the end and simply point in the EXIF header to our new IFD0... --=20 Lutz M=FCller <lu...@us...> |
|
From: Roberto C. <rob...@en...> - 2003-11-04 20:56:18
|
Lutz M=FCller wrote: >=20 > On Mon, 2003-11-03 at 21:04, Roberto Costa wrote: > > Lutz M=FCller wrote: > > I'm not convinced that is an acceptable behaviour the fact that a > > libexif user that processes his/her images with the library may lose > > information because of lack of knowledge about his/her camera maker > > notes. > > My idea is that libexif should always be conservative and do its best > > not to corrupt any user exif info. >=20 > Agreed. So, if we encounter MakerNotes other than Canon, Olympus and > Pentax, we'd like to preserve data between IFDs at its original offsets= =2E > The easiest way (I guess) to accomplish that would be to preserve the > original EXIF data, to append our new/modified EXIF data at the end and > simply point in the EXIF header to our new IFD0... Yes, I fully agree. Perhaps it should be possible for a library client to request one of the two behaviours explicitly, unless specified the exif data compaction would be enabled for known maker notes and disabled for the others. Have you in mind an interface to do it? I guess it would be sufficient to add a new function to save exif data, the signature would be identical to the existing except for an additional parameter that specifies the desired behaviour. Cheers, Roberto |
|
From: Lutz <lu...@us...> - 2003-11-04 22:06:55
|
On Tue, 2003-11-04 at 21:52, Roberto Costa wrote: > Perhaps it should be possible for a library client to request one of th= e > two behaviours explicitly, unless specified the exif data compaction > would be enabled for known maker notes and disabled for the others. >=20 > Have you in mind an interface to do it? I don't think anyone cares for the possibility to switch between storing the original EXIF data in addition to our new one and not doing so. It should just work. If someone really needs the feature, we can add it easily later. --=20 Lutz M=FCller <lu...@us...> |
|
From: Roberto C. <rob...@en...> - 2003-11-05 22:00:46
|
Lutz Müller wrote: > > On Tue, 2003-11-04 at 21:52, Roberto Costa wrote: > > Perhaps it should be possible for a library client to request one of the > > two behaviours explicitly, unless specified the exif data compaction > > would be enabled for known maker notes and disabled for the others. > > > > Have you in mind an interface to do it? > > I don't think anyone cares for the possibility to switch between storing > the original EXIF data in addition to our new one and not doing so. It > should just work. If someone really needs the feature, we can add it > easily later. Well, I do care, this is why I was suggesting it ;-) More in general, since libexif is a public library, which happens to know how to do a particular job (saving exif data in our case) in two different ways, I think the user should be able to choose which way he wants to do things. This makes especially sense to me if I consider that there is a trade off, one method is more conservative, the other produces smaller and more compact exif data. Cheers, Roberto |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X