The Tao of tagging in Lightroom

[dreed]

I happened to watch a video clip of Seth (and MIchael?) on youtube where tagging in Lightroom is discussed.

Seth is voluminous in his tagging but for good reason - he keeps all of his images findable and categories easy to browse.

In some areas it is clear that his tag tree is just providing structure (e.g. landscape > category, portrait > category) but others, it has value due keywords being trees on the tree of tags.

How do others approach the tagging decisions?

[howardm]

what video are you referring to?

I try to make use of a good amount of hierarchy so that even if my keyword search is approximate, the returned results have a good probability of being in there. On a recent trip to Montreal (first time after using LR), I might have Places->Canada->Quebec->Montreal  and Things->Buildings->Churches and Notre Dame so that hierarchy doesn't get confused w/ France->Paris or at least Notre Dame would return both churches.

I think it's also important to structure your hierarchy the way *YOU* think and not necessarily how someone else thinks so that when it comes time to do a search, it'll be what you 'expect' vs. struggling w/ someone else's controlled vocabulary.

[Rory]

I think it's also important to structure your hierarchy the way *YOU* think and not necessarily how someone else thinks so that when it comes time to do a search, it'll be what you 'expect' vs. struggling w/ someone else's controlled vocabulary.

Good advice, except I'm not sure how controlled my thought processes are.   

[Tony Jay]

The video in question will be an episode of the "Where the @#$% are My Pictures" series on digital asset management available on this website.

Keywording and keywording strategies are among the most controversial and misunderstood aspects of digital asset management.
Most individuals who do not have a good system of keyword and other metadata capture have no idea how easy it can be to find any image at any time.
The "folder system" dies hard it seems.

Dreed seems somewhat bemused about how Seth chooses to structure his keyword hierarchies but how else is one to organise several thousand keywords.
My keyword hierarchy has 44 000+ keywords.
It would not be possible to manage this if it were just a single long list.
(In fact that number might exceed the maximum allowable top level keywords in Lightroom.)
So sure, I use many different hierarchies that offer a logical way to construct them such as locational keywords or biological taxonomies.

I do agree with Howard's comment about tailoring a keyword hierarchy to suit one's needs.
There are plenty of individuals offering keyword collections for sale but the chances of any one of those collections mirroring one's needs appropriately are small.
It is likely that the granularity (or detail) is not exactly what is wanted.
Part of the sales pitch is that one can subsequently adapt the collection to suit ones needs.
That will seldom happen unless one is capable of doing the job from scratch on the first place - in which case one would be better off starting from scratch.

As time goes all the aspects of digital asset management will be seen to become progressively more important as the already staggering volume of digital images in the world continues to rise and their management and organisation becomes more of a burden.
Most of us have image collections ranging from a few thousand into the hundreds of thousands. One may be just able to cope using folders and renamed images and sketchy IPTC metadata acquisition to find images when required with a couple of thousand images. This approach does not scale well when the collection grows to 500 000 images.
At that point all the images will be there on a hard drive somewhere but they are as good as flushed down the toilet because they will be almost completely inaccessible without excellent and appropriate metadata capture (keywords are a subset of IPTC metadata).

Seth Resnick's approach is entirely practical and, in principle, should not be thought of as something "just for him".
My goal with metadata capture is exactly the same as Seth's - I want to be able to easily find my images as and when necessary.

FWIW I make extensive use of non-exportable keywords to organise my keyword hierarchies. Each hierarchy such as "ANIMALS" is headed by a capitalised non-exportable keyword.
Inside, depending on how complicated the hierarchy is I will make use of other non-exportable (and also capitalised) keywords  to assist in organisation.

However this approach may not suit everybody.
For example: If one shoots an image that has a tree as the main subject, is keywording that image with just 'Tree' suitable or would one want to keyword it according to genus and species as well as its common name.
The difference in the amount of effort expended on one's keyword collection between these two extremes is prodigious.

Between those two extremes there are probably a couple of practically useful middle roads.
Perhaps tree is too generic, but, genus and species (with a full taxonomy above) is far too detailed.
Maybe just to have broad families, as children of "Tree", such as "Acacia", "Conifer", and "Eucalypt" might work with the occasional common name such as "Camelthorn" (that lives under "Acacia" in this case).

The bottom line is that a fair bit of research needs to be done to get a good grasp on exactly what one's needs are with regards to keywording. In addition, getting a really good grasp of IPTC metadata will give one the option of doing a lot of metadata capture (such as locational metadata) that may not need to be duplicated in the keywording.

I hope John Beardsworth spots this thread and chimes in with his views since he is very well versed in digital asset management and will be able to offer plenty of options with regard to keywording.

Tony Jay

[luxborealis]

A lot has to do with the variety of shooting one does. Increased variety of place, style, colour, subject and use means an increase in the number and value of keywords to find images efficiently.

Someone like me with a comparatively small library (~50,000) of primarily landscape and nature shots made at discrete and periodic times through any given year, is less in need of an extensive, hierarchical keyword list. I can find any of my photos from over 10 years of digital shooting on three continents, 15 countries and about 30 states/provinces/counties with two or three keywords and three or four clicks, max. And that's with a keyword library of just over 2,500. When keywording, I try to be as descriptive as possible with regards to subject, place, time, colours, conditions and details.

If I was to have a keyword library as extensive as Seth's, I would be spending all my LR time key wording! There comes a tipping point, which varies from person to person according to style, variety, number of jobs and how important it is to your business to find groups of photos on specific themes. You want the ability to efficiently find them across numerous "folders" without spending too much time entering keywords after each shoot. Thank goodness for autocomplete, otherwise the task would be more monotonous than it is.

[dreed]

The video in question will be an episode of the "Where the @#$% are My Pictures" series on digital asset management available on this website.

I've seen it somewhere else too recently... or words about it... or maybe it was just the short promo...

Seth Resnick's approach is entirely practical and, in principle, should not be thought of as something "just for him".
My goal with metadata capture is exactly the same as Seth's - I want to be able to easily find my images as and when necessary.

Yes, this is key criteria for me.

FWIW I make extensive use of non-exportable keywords to organise my keyword hierarchies. Each hierarchy such as "ANIMALS" is headed by a capitalised non-exportable keyword.
...

I don't think I understood the non-exportable thing until recently and by understood, understood how it could be used for my own benefit.

One of the problems that I've been trying to work out is how do I distinguish meta data for being "at the eiffel tower" vs "of the eiffel tower." To this end I think I'm going to do something like "LOCATION > france > paris > eiffel tower" and "SUBJECT > MAN MADE > BUILDING > eiffel tower". The part I've never been able to figure out before is whether to have "LOCATION" as well as "france" inside of "LOCATION" ticked but I now think that I understand that. Maybe.

I've been trying to get *some* tags in for every photo that I take (e.g. sunrise, sunset, waterfall, etc) but I was starting to wonder if I could be doing it better.