A timelapse of the Golden Gate, shot from Kirby Cove in the Marin Headlands and then from Grand View Park in San Francisco.

South from Golden Gate Heights Park, Sutro Tower from Twin Peaks, south from Golden Gate Heights again with more clouds and west from Grand View Park. Photo is a panorama of Golden Gate Park from Grand View.

I'm switching from the more ambitious specific weeks to a timelapse of the week (TLOTW) when I have one.

(Previously: Timelapse, Week of Jan 31 2022)

Sun halo over Lake Merced, a container ship entering the bay, the SkyStar ferris wheel in the Golden Gate Park, Farallon Islands, and some nice clouds over the Pacific. Photo is from Lake Merced.

(Previously: Timelapse, Week of Jan 24 2022)

The Golden Gate from Grand View Park, and then the Outer Sunset from Golden Gate Heights Park.

(Previously: Timelapse, Week of Jan 17 2022)

Light cloud cover streams by as the sun sets over the Pacific, shot from San Francisco. 4K 60fps timelapse.

(Previously: Sunset #7)

Three different perspectives of downtown San Francisco. Shot from Twin Peaks (with a great view all the way down Market Street), Mount Davidson (the highest point in the city) and of course Grand View Park.

(Previously: Timelapse, Week of Jan 10)

An experiment in creating a timelapse from a week of walking around San Francisco. Not sure how long I'll keep this up for and only managed four sequences this week: three different views from Grand View park (I go there a lot) and one of the penguin sculpture at Lake Merced.

Timelapse around the shoreline of San Francisco:

This could be arbitrarily long of course, but I only had a day to shoot and picked six spots: the lookout at Fort Funston, the Camera Obscura next to the Cliff House, Eagles Point on the Lands End trail, the Marina Yacht Harbor, Pier 7 and the Giants Promenade Pier.

Shot on a Sony A7C with the 20mm 1.8 G and a Ronin SC. This is the first time I've used the Ronin SC for timelapse and it sucks. Works great for video, can't pan 180 degrees in timelapse mode without introducing shake though. Processed using LTTimelapse, Lightroom, DaVinci Resolve (with a lot of stabilization to fix the Ronin issues) and Filmstro Pro. The finished product is 4K, 60fps. Filmed December 30, 2021 (I usually do something on New Year's Eve but the clouds looked more promising the day before this year).

(Previously: San Francisco New Year's Eve Timelapse 2020)

A timelapse of Lake Merced in San Francisco, California on a stormy December day.

(Previously: Merced)

San Francisco is apparently going to hell with criminals free to do as they please with no fear of consequences. I decided to take a look at the data.

The video above shows a timelapse of SFPD incidents from 2003 through yesterday. Each frame is a day and shows incidents from the previous seven days. The top left corner of the video shows the date and the seven day count of incidents.

I grouped the reported categories into a few colors. Red is used for murder and rape. Orange for arson and kidnapping. Yellow for thefts and assaults. Purple for sex and drugs. Grey for anything else. I excluded some categories from the data (recovered vehicle, traffic collision, case closure and non-criminal).

SFPD reports the location of incidents as the closest intersection. To keep everything visible I move the location randomly within a tenth of a mile where there is a specific location reported. For crimes without a location I use a random spot within half a mile of the center of the police district (or the center of San Francisco if the district is missing - this is unusual).

The volume of incidents changes a bit during the ~18 years shown in the video, but the only real outlier is the dip following the start of the coronavirus pandemic in March 2000. Crime picks back up after this but so far hasn't returned to the level it was at before the pandemic.

Police incidents come from two datasets: Police Department Incident Reports: 2018 to Present and Police Department Incident Reports: Historical 2003 to May 2018. San Francisco is plotted using Elevation Contours. The pre and post 2018 data sets use different categories but I coded both to the set of colors outlined above. There is a drop in incidents right at the end of the video which I expect is caused by incomplete data rather than any change in crime rate.