[digiKam-users] Raw metadata files writing in incoming 6.0.0...

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

[digiKam-users] Raw metadata files writing in incoming 6.0.0...

Gilles Caulier-4
Hi all,

I just responded to this entry in bugzilla about RAW metadata writing option :


I vote to drop this dangerous option in incoming 6.0.0 release. See my viewpoints in bugzilla entry...

As it's always easy to cut source code than to add new implementation, i post this message here. Please continue the discussion in bugzilla as well for a better visibility.

Best

Gilles Caulier

Reply | Threaded
Open this post in threaded view
|

Re: Raw metadata files writing in incoming 6.0.0...

Remco Viëtor
On dimanche 1 avril 2018 11:19:48 CEST Gilles Caulier wrote:

> Hi all,
>
> I just responded to this entry in bugzilla about RAW metadata writing
> option :
>
> *https://bugs.kde.org/show_bug.cgi?id=387351#c9
> <https://bugs.kde.org/show_bug.cgi?id=387351#c9>*
>
> I vote to drop this dangerous option in incoming 6.0.0 release. See my
> viewpoints in bugzilla entry...
>
> As it's always easy to cut source code than to add new implementation, i
> post this message here. Please continue the discussion in bugzilla as well
> for a better visibility.

Personally, I don't see any reason to write to raw files:
- it's the original data, that I cannot replace (in most cases).
- most raw formats are very poorly documented, so writing to them is at best a
(very well educated) guess, as not all corner cases are known. This is
especially bad for the "makernote" section. (I use Sony, so .ARW files...)

DNG is a bit of a special case. If it is generated directly in the camera, I
don't know how well documented it is. But the DNG is still the unique original
data, and basically irreplacable. So writing to them is (imo) bad.

If the DNG is produced from another raw format, it can be dealt with as any
other derived file, and writing to it could be allowed (with the "experimental"
note attached, if that's the status in the Exiv2 library).

Tl;dr: I vote to drop writing to RAW files completely, with a possible
exception for DNG files.

Remco
Reply | Threaded
Open this post in threaded view
|

Re: Raw metadata files writing in incoming 6.0.0...

Gilles Caulier-4
Right DNG is RAW file, but it's :

- based on TIFF/EP standard.
- normalized.
- fully documented.

If one RAW can be written as well with metadata, it's this one.

All others are mostly TIFF based with internal customisations, not documented, and few encrypted (as Nikon for ex). There are like patented data + code.

To touch the contents of this kind of files is in fact very dangerous.

This explain also why the Raw demosaicing is a big puzzle. Typically, your camera as the firmware to manage this format and export to a target format as JPEG, and the cost is not negligible. But it's another topic.

Gilles Caulier

2018-04-01 12:17 GMT+02:00 Remco Viëtor <[hidden email]>:
On dimanche 1 avril 2018 11:19:48 CEST Gilles Caulier wrote:
> Hi all,
>
> I just responded to this entry in bugzilla about RAW metadata writing
> option :
>
> *https://bugs.kde.org/show_bug.cgi?id=387351#c9
> <https://bugs.kde.org/show_bug.cgi?id=387351#c9>*
>
> I vote to drop this dangerous option in incoming 6.0.0 release. See my
> viewpoints in bugzilla entry...
>
> As it's always easy to cut source code than to add new implementation, i
> post this message here. Please continue the discussion in bugzilla as well
> for a better visibility.

Personally, I don't see any reason to write to raw files:
- it's the original data, that I cannot replace (in most cases).
- most raw formats are very poorly documented, so writing to them is at best a
(very well educated) guess, as not all corner cases are known. This is
especially bad for the "makernote" section. (I use Sony, so .ARW files...)

DNG is a bit of a special case. If it is generated directly in the camera, I
don't know how well documented it is. But the DNG is still the unique original
data, and basically irreplacable. So writing to them is (imo) bad.

If the DNG is produced from another raw format, it can be dealt with as any
other derived file, and writing to it could be allowed (with the "experimental"
note attached, if that's the status in the Exiv2 library).

Tl;dr: I vote to drop writing to RAW files completely, with a possible
exception for DNG files.

Remco