I am Cruel to Lightroom

See http://picasaweb.google.com/the.bangkok.photographer/20090802

I went for a great city walk this afternoon. I walked all the way from Chit Lom Skytrain Station to Phetchaburi Subway Station via Phetchaburi Road and Kamphaeng Phet 7 Road.

I stopped a few time for coffee, water and snacks but mostly I was taking pictures the whole way.

My route from PicasaWeb

Thus when I got home I had over 300 photos on the Nikon Coolpix P6000.

I had a shower and a cup of tea while Lightroom was importing them and backing up the originals. Then I set to work rating and keywording the pictures. All but a few were geocoded already so that part was easy.

The way I work is that I make a first pass through the pictures in Grid view adding keywords for WHERE the picture was taken. I can do that very quickly. I have keywords set up for all the popular locations in Bangkok, organised by the administrative district (khet).

I started in Khet Pathum Wan on Ploen Chit Road but once I got to the Rachaprasong Intersection I moved into Khet Ratchathewi.

I have a separate set of keywords for major thoroughfares that cross several khets. For example Thanon (Road) Sukhumvit, Rama IV, Rama IX and, in today case Phetchaburi. I spent a lot of time on Phetchaburi Road so location keywording was easy.

I took a detour onto Khamphaeng Phet 7 Road to check the progress on the new Airport Express Station at Makkasan. It looks like they have a lot of work to do.

On the way I tried to look at Bangkok from the perspective of a stock photographer. So I tried taking some pictures that I thought may be commercial. I took care not to include people (since I didn’t have a load of bi-lingual releases in my back pocket) and potential copyright violations like trademarks.

I found an open air store that sells spirit-houses. That was a great photographic opportunity. At every intersection I took a picture of the street sign (Phetchaburi Soi …). That enables me to check the accuracy of the P6000’s GPS location – and I also like the design of the signs.

Once I completed that – about 15 minutes – I started the detail work of adding WHAT keywords for all the pictures. I am mostly interested in road vehicles so about 200 of the 300 pictures were some sort of vehicle.

Private cars and the like are easy since I have just about every vehicle sold in Thailand in the hierarchy already. The Lightroom keyword suggestions are great at this. If I type “taxi” Lightroom automatically suggests “Corolla” since that’s the most common Bangkok taxi and I usually add “Corolla” after taxi. Once or twice I fooled it and typed “Tiida” or “Sunny”.

Since the keywords are in a hierarchy Lightroom will export the parents – “Toyota” or “Nissan”. I also added my machine tags “veh:man=Toyota”, “veh:model=Corolla” and “veh:type=taxi” as synonyms so they get exported too.

The export is important as that is the set of keywords that end up on Flickr or PicasaWeb.

Bangkok buses are also easy. If I type the BMTA route number as a keyword Lightroom always suggests the correct model (and thus make) of bus as it’s seen tham all before. Often BMTA operates two types of bus on a route: air-con and non air-con. Lightroom suggests both and I have to pick the right one. (It can’t distinguish a red bus from a blue one and decide for itself.)

I have the bus route (e.g. Victory Monument-Min Buri) as synonyms for the route so that gets added to the export set.

I have MISC keywords for things like spirit-houses and signs. They’re easy to add.

When I crossed the Khlong San Sabe Lightroom immediately suggested “ferry” as a keyword. It’s clever like that.

In parallel I made a quick decision on whioch pistures I would upload to Flickr. I flagged those Yellow and usually made some develop adjustments and crops. I also added a Title and Caption. That breaks up the otherwise mind-numbingly repetitive keywording quite nicely.

The keywording took about 90 minutes. At that point I paused for dinner. While eating I did another backup of the database.

The next step was geocoding. I only had a few to geocode manually so I was able to do that using the Interactive mode of Jeff Friedl’s plugin. The un-geocoded pictures were all at the start of the walk when I got off the Skytrain and the P6000 had not found its bearings.

The final step before I declared victory was reverse geocoding. That is, adding location information in the IPTC fields of the database based on the Lat / Long of the geocode. That duplicates the information I have in the keywords and is not strictly necessary.

Jeff’s plugin makes reverse geocoding easy so there is no harm in doing it. Plus Lightroom makes it easy to filter pictures based on those IPTC fields.

However …

When I selected all the pictures the Lightroom database gave a huge gulp. It had to accumulate the metadata for all 300 pictures and present it. That was a waste of time since I wasnt going to fiddle with the metadata – and most of the fields said anyway. But the merged keyword list for export was huge. Here’s a screenshot of part of it:

Keywords with an asterisk apply to only some of the images. The only ones common to all are “Bangkok” and “Thailand”.

This is incredibly powerful but in this case a waste of time. Poor old Lightroom must have run out of memory, virtual and real, as it ground to a halt and I had to terminate it with extreme prejudice. That always worries me with a database. But I knew I had a recent backup and Lightroom had been writing all the changes to the pictures as XMP so if all else failed I could restore from that.

I never bother telling Microsoft about third-party software errors. I wonder what they do with it? Perhaps they send competitors a gloating summary: “Your software crashed for 63,426 of your customers. Ha, ha!”.

I restarted my computer and optimised the Lightroom database. All was fine after that and I was able to reverse geocode the lot.

Jeff gets his reverse geocode information from Google Maps and it isn’t very good with Bangkok. It put many pictures in Makkasan sub-district (khwaeng) but didn’t include the name of the khet – Ratchathewi. Still it is a lot better than nothing and as long as I am consistent it makes sorting and searching easier.

Tags: Bangkok