Location keywords and Metadata Browser

After reading many posts here in the forum over the last months I was wondering how many LR users go for the Metadata Browser approach to add location information instead of using keywords to describe the location the photo was taken.



Even in the brilliant LR Video Tutorial with Michael Reichmann and Jeff Schewe, Michael is referring to location information as something to be added as a keyword and doesn't mention the Metadata Browser at all in this regard.



Background: It is obvious that location information is very valuable information which should always be added into the metadata section of a photo. One way of doing this is to add the location as a keyword like ''Grand Canyon'' or ''LA Downtown''. An alternative is to use the IPTC fields Location, City, State / Province, Country, Planet, Galaxy. The last two are not yet there but that's only a matter of time I guess. It is a very neat feature of LR to automatically build a location hierarchy using the mentioned IPTC fields



So I made an experiment on the last weekend and got rid of all keywords describing location and used Location, City, State, Country consistently filled for all my RAW photos (roughly 4000) instead. So where is the advantage? For me it is the fact that my locations are no longer flat keywords but hierarchically organized and easily accessible via the Metadata Browser.



I am wondering about the experience of other LR users. Did you also consider to switch location data? Are there disadvantages in this approach besides the initial effort it takes?



regards

Chris
Location keywords and Metadata Browser
Keywords have 2 important advantages over other IPTC-Metadata:



1. They are flexible, i.e. I can build any hierarchical structure I want, not just ''Country/Province/City/Location'' which often just makes no sense, especially for landscape images outside of any city. Why does a location need to be a part of a city? According to the fixed location-structure, you can only use Country and Province for images outside a city - no park, no region or any other useful info to further specify the location.



2. Location-Keywords are automatically added to the IPTC-Keywords field and that is where many search engines look for metadata, not in Country/Province/City/Location. I can also automatically include/exclude parent keywords and synonyms. Very useful!



IPTC-Location should be added just to have the metadata complete, but IPTC-Keywords are more relevant to me.
Location keywords and Metadata Browser
Thanks for your comment Matteo,



Basically I see the four fields ''Country/Province/City/Location'' in a more generic way providing me with a four level hierarchy of location information regardless of the names of the fields.

But interestingly I had no problem to fill them even for outdoor shots in parks.



Before LR I was not aware of the advantage to have all four fields filled but now I find myself using the location hierarchy in LR extensively.



To address your second point. LR and many other apps able to search metadata can find a keyword or a text fragment regardless of the field name. I just used the LR find function and entered ''Italy'' and got all pictures with Italy in the IPTC country field.



Do you have an example of apps which are restricted to the IPTC keyword section for the search function?

Matteo,

Of course it's your decission how will you organize and use your (meta)data. But, I'm not sure if that's the right way in long term.



IPTC location tags (Country, State, City) are dedicated tags and any serious application will recognise them as such.

Keywords tag is ''out-of-category'' tag -can be usefull for further descriptions, but if not used in ''organized'' way, you can get a mess. For example: if you have ''Lake'' (you know.. water) in one image and ''Salt Lake City'' (no water on photo) in another image -quite hard to filter out what you search for.



I had similar thoughts about ''City'' tag as you have... and decided: for me, City mean ''place'' (which can be city, village,..). And after all, even deserts have their names.



Right now, you see many benefits of using any kind of info in Keywords. But, many tools rely on Location tags... and it can happen you won't use LR forever.



Speaking for me, location tags are very well used in LR.. so why not use them and be future compatible? But.. I'm just suggesting :-)



Bogdan

1. LR can search in all metadata fields, but I'm not sure if all agency software does that too. E.g. I know that Alamy looks in Caption/Description and keywords. Not sure if they look into location fields



2. (Mis)using fields like city or location for any other location-related info is not clean and can have unwanted sideeffects, e.g. someone looking for city images will find all other images (mis)using that field.



3. Don't know if it's Adobe's interpretation of the location-field, but it would be more useful if it were more generic instead of just a city-location.



The trouble with all these fields and different search approaches is that I'm now ending up filling the location in three different fields:



Caption, Keywords and Location!

Hi all,



From the IPTC guidelines for the Photoshop metadata panels:



Location [Location]

Enter the name of a location shown in the photograph. This location name could be the name of a

specific area within a city (Manhattan) or the name of a well known location (Pyramids of Giza) or

(natural) monument outside a city (Grand Canyon). Location is the most specific term, at the fourth

level of a top-down geographical hierarchy.



City [City]

Enter the name of the city that is pictured in the photograph. If there is no city, you can use the

location field alone to specify where the photograph was taken. City is at the third level of a topdown

geographical hierarchy.



I am sure there is still a lot of room for interpretation in many cases, but these guidelines are a good start I think!



All the details at:

http://www.iptc.org/IPTC4XMP/



David.

Thanks for the quote:

These definitions make more sense than Adobe's interpretation. Considering this defintion, LR has a bug: Location must show next to city, not below city!

I think LR is correct: the guidelines say 'City' is the 3rd level of description and 'Location' the 4th. But it is fine to leave the 'City' field blank if it is not applicable and put something in 'Location'.



David.

I'm sure we all know what City,Country... mean, but sometimes it's not so easy to decide which tag to use.

I've read some docs on the site David suggested (User guidelines - Documentation PDF). Well, there's black on white (in short): if there's no city (on image or where photo was taken, I suppose), leave that tag blank and use Location instead. But...



..if you look at examples bellow that document (the man working on the field), you can see there's City defined (Watseka) -even image doesn't contain city nor photo was taken in the city. As I don't know Watseka, I assume it's city/town/village nearby that field (on photo).

So I assume it's still OK if we populate City tag in such cases -that's what I meant using City tag as ''a place''. IMO that's still better than defining location names (city, village, etc.) in Keywords tag.



Bogdan

I used to use a collections-like heirarchy to do this job when I was using iView. But then I realized that its version of ''metadata-browser'' made keeping that info in collections or keywords redundant. So I thought about this some when I started migrating to Lightroom.



Obviously, if one is marketing his photos to an agency or business that requires following IPTC standards, the Country, State, City, Location (CSCL) standards created by the IPTC organization (http://www.iptc.org) need to be followed fairly rigorously.



But I expect that most of us do not need to follow those standards and can adapt our use of the CSCL fields however works best for us.



I use the City field in a way that works for me. For me ''Grand Canyon'' is a ''City'' and the location field is used for ''North Rim'', ''South Rim'', ''Inner Canyon'', etc. Other ''Cities'' I have include ''Yosemite'', ''North Cascades'', ''Yellowstone'' and several more. I may also use City for large State Parks or National Forests.



The main **Advantage** of using the CSCL fields is that they ARE heirarchical and the heirarchy resolves duplications. There are 8 cities named ''Denver'' in the United States. CSCL makes sure I get the one I want. (I'm sure SOMEONE is more interested in Denver, Pennsylvania.)



Lightroom allows us to keep keywords in a heirarchical order. But this is in the LR database; the keywords in the image metadata are not stored heirarchically and the relationship between Denver and Colorado would be lost if the image was moved to another program that does not have access to the LR database.

Hi Narsil you wrote:

''Lightroom allows us to keep keywords in a heirarchical order. But this is in the LR database; the keywords in the image metadata are not stored heirarchically and the relationship between Denver and Colorado would be lost if the image was moved to another program that does not have access to the LR database.''



I think due to the fact that the CSCL fields are standardized by IPTC and stored in the XMP container every program can interpret them with the hierarchical semantic defined by IPTC. With other words if you write the XMP metadata to the files the relationship between Denver and Colorado is also available outside LR for every application which interprets IPTC metadata correctly.



Chris

Chris,

Perhaps I should have highlighted it some way but what I said was the KEYWORDS in the image metadata were not heirarchical. Yes, the CSCL are heirarchical, but the KEYWORDS (outside of Lightroom) are not. That's one of the main reaons I use CSCL instead of keywords for location information.



Sorry I was not more clear.

Hi narsil,

Then we have the same understanding of that topic. I hope that this thread helps to point out the advantage of using the CSCL fields for location information rather than keywords.

Yes, your're right: City/Location are tagged as 3rd and 4th level. Only thing that's ugly is that locations without a city will then show under 'unknown city'.

Indeed... And in a country where there is no 'State', that's even worse ;-)

Narsil wrote ''I use the City field in a way that works for me. For me ''Grand Canyon'' is a ''City'' and the location field is used for ''North Rim'', ''South Rim'', ''Inner Canyon'', etc. Other ''Cities'' I have include ''Yosemite'', ''North Cascades'', ''Yellowstone'' and several more. I may also use City for large State Parks or National Forests. ''



I wish I'd thought of that... I don't like the ''unknown city'' and Narsil's solution provides the possibility of another level in the hierarchy for, say, Yellowstone National Park.



However, it doesn't solve the lack of a state for countries that don't have them or I don't care of I include the province or whatever or not.



One reason I went to Lightroom after a couple of weeks with Elements is I don't want location in keywords.



My use of these fields is only for my own use so I needn't worry about a stock catalog.



Judy

Judy,



I noticed that if I don't include a ''state'' the Meadata Browser doesn't find the cities. (I don't travel outside the USA often, so this only applies to my Israel slides.)



What I did was simply insert a hyphen, ''-'', for the state entry for all the Israel images. Then the different cities show up.

Narsil,



Weird, this is not my experience. I do get 'Unknown State/Province', and then the cities appear below that level... I wonder why the difference?



David.