On 12/31/2012 02:47 PM, Tobias Ellinghaus wrote:
Hi!
[...]
>> 2) Metadata of derivative images arising from "developing" or processing
>> of original RAW images. For myself, these are typically JPEG or TIFF
>> files. These are the images often shared with others. Having these
>> images in a Sidecar file is unproductive, because only Dt can truly
>> understand the content of the Sidecar files that it wrote.
I think this is not quite correct. At least to my knowledge the
descriptive metadata that dt writes are quite consistent to what other
tools write. So keywords e.g. should be imported properly.
>> Even Digikam
>> has a problem with understanding the Sidecar files generated by Dt. A
>> better solution would be to incorporate the Metadata (of which the image
>> description is the most important; e.g. "Extent of burned grassland at
>> Exeter at the end of the 2009 dry season") into the JPG or TIFF image
>> itself. However, the proviso is that it should be done in a way that
>> almost any imaging software can display this information. At the moment
>> I put this information in the file name but it is a highly inconvenient
>> and inefficient approach.
>> I appears that there are strong limitations imposed by the EXIF
>> specification. For instance, it appears that a "Title" or "Caption"
>> identifier is not normally allowed. Moving to IPTC is an obvious
>> possibility, but few imaging programs support this.
>>
>> Is there an obvious way out that I have overlooked?
>
> What exactly are you trying to say? We already embed all the metadata we have
> into the exported images (provided the format supports that).
I think the OP's main problem seems to be a missunderstanding of what
metadata is what (descriptive vs. technical) together with the problem
that you can't apply to much *descriptive* metadata within dt. All exif
(this is mainly technical metadata) is there as far as I can see in all
exports. For the rest DT supports a dublin core simple embedded in XMP.
I think this could be a hint to the OPs problem. I'm not really sure,
but my feeling is that IPTC is better understood by 3rd party tools
compared to XMP (it is there and standardized for quite a while).
So I think not filling in IPTC properly could very well lead to the
feeling of a "loss" of metadata applied in dt in certain scenarios. I
experience that as well, that's why I hope for full IPTC in dt and
currently do all indexing in mapivi and not in dt.
E.g. I know some tools than handle IPTC.title properly, but can't handle
XMPs dc.title at all. (Most likely they don't evaluate the XMP block at
all.) Same for keywords, description (aka IPTC abstract/caption) In this
scenario it looks like you applied all data in dt but nothing went over
to your 3rd party tool, though everything is there in the sidecars. But
those sidecars get happily ignored by those more simple minded tools
anyway as they just expect the metadata to live in the TIFF or JPG and
thus don't even look at the XMP. This at least matches what was
described by the OP.
The last issue seems to be pointed to by the OPs comment that he
obviously tries to read descriptive metadata (sounds like IPTC.caption)
from the sidecare files. I think that decent tools handle the XMP files
from dt correctly, at least as far as I can see all XML looks pretty
good and I was able to read them with the few tools I use. However, I'm
not sure that 3rd party tools expect the xmp and/or can link the xmp
with TIFF or JPG files properly. Still using exiftool should even write
them from teh XMP to the TIFF/JPG/<whathaveyou> in case they wouldn't be
there already.
> The "only" problem I currently see is that we lack a convenient way of editing
> all the data,
But you'd not want to allow to edit the f-stop and friends, right?
> but that is on the agenda already. That, and a way to decide
> what metadata to *exclude* from being embedded into the exported file.
Sound both great, I still hope that there will be a way to add all
missing fields from IPTC, like location data and so on. The mentioned
description would be IPTC as well. Probably one could think of linking
IPTC.abstract with EXIF.comment and dc.description. For compatiblity
with less potent tools it might be a pretty good idea to keep dc-ecoded
XMP and IPTC in sync for fields that exist in both schemes anyway.
(dc.title == iptc.title e.g.)
--
Kind regards, / War is Peace.
| Freedom is Slavery.
Alexander Wagner | Ignorance is Strength.
|
| Theory : G. Orwell, "1984"
/ In practice: USA, since 2001
------------------------------------------------------------------------------
Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
MVPs and experts. SALE $99.99 this month only -- learn more at:
http://p.sf.net/sfu/learnmore_122412
_______________________________________________
Darktable-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/darktable-users
