Tumblr Pr0n

In the Data Dumps and Storage Wars : Narrating Digital Archives class, AJ LeVine and I collaborated on a project to archive the NSFW gif culture on tumblr. We were particularly drawn to the rise of low-bandwidth gif porn in the age of hyper-HD, and decided that we should archive the gifs in the OG low-bandwidth format of ascii.

Prototype of Archive

http://pr0n.works

Why Tumblr?
  • Risk of censorship
  • Interest in youth consumption trends
  • Risk of obsolescence (think geocities)
Scraping the Images

Tumblr filters out NSFW content in it’s search results by default. Users who are logged in can modify their settings to show these results.

porn_gif_search_Tumblr

porn_gif_search-unlocked-sfw_Tumblr

Tumblr’s Policy on NSFW content

https://www.tumblr.com/docs/en/nsfw

Example of NSWF filters in action:

API Query Posts Tagged “gif”:
http://api.tumblr.com/v2/tagged?tag=gif

API Query Posts Tagged “gif+porn”

http://api.tumblr.com/v2/tagged?tag=gif+porn

This turned out to be a challenge that was very relevant to the Critical APIs class I am currently in.

I experimented with node, wget and import.io.

We ended up using wget to scrape about 30GB, but it was an inefficient process because I had to:

  1. select sites to scrape, as opposed to searching all of tumblr for images of a particular tag.
  2. and in doing so, omit scraping sites with mixed sfw and nsfw content
Current Project Status
  • Prototype built
  • Test data scrape
Upcoming milestones
  •  scrape by tag / image analysis
  • include meta data of gif source in scrape
  • improve presentation

Multiple Exposures as Data Visualization

Assignment: You will be creating three images based on multiple exposure for this assignment. You will work from a set of multiple exposure photos taken by you or sourced from elsewhere. Think about the kinds of stories you hope to reveal before selecting from the list of techniques below.

Technique 1: Strips from different exposures of the same scene at different times. (Landscape example)
Technique 2: Smooth transition between exposures of the same scene at different times. (Boardwalk example)
Technique 3: Overlaid exposures of the parts of same scene at different times. (Airport example or motocross example)
Technique 4: Average exposures of same scene at different times. (Portrait example)
Technique 5: Overlaid color separated images. (Portrait example)


 

For this assignment, I combined 3 self-portraits into the same image. I wanted to play with multiple exposures in 3D, so first, I used a kinect to create three 3D scans of myself, over 3 days.

After experimenting with different techniques, 2D & 3D, to combine the images into a single “exposure”, I found the following composites to most compelling.

1. Compositing the 3 exposures into the same 3D space

00_pat_soild-multi-expo-comped

 

2. Compositing the texture files from the 3 exposures into one, then mapping it onto a single 3D model

00_pat_UV_comped-1000px

 

UV-maps_explained

Long Exposure as Data Visualization

Assignment: Creating four long exposure images.

  • Image 1: Photo of your subject without light trails.
  • Image 2: Photo of your subject using light trails.
  • Image 3: Photo of your subject that is extrapolated from two or more discrete photos.
  • Image 4: A  diagram explaining the motion captured in the long exposures.

 

For this assignment, I created a series of long exposure images that would capture the movement of a Makerbot Replicator in action.

Tools used:
Timer Remote, Canon 6D, 17-40mm Lens,  0.6 ND Filter, 3V Coin Battery, Red LED, Gaffer Tape, Skull Model, Makerbot
Timer Remote, Canon 6D, 17-40mm Lens, 0.6 ND Filter, 3V Coin Battery, Red LED, Gaffer Tape, Skull Model, Makerbot

Timelapse Gif of the Makerbot in Action

makerbot-6D-comp1-702p-v2

 

1 : Image without light trails

example_single

Camera Settings:
  • 60s exposures
  • F22 + 0.6 ND
  • iso 100
  • 1s between exposures,running for about 40minutes

2 : Images with Light Trails, Overlay Comp

makerbot-6D-comp1-cam-static

3 : Images with Light Trails, Extrapolated Comp

makerbot-R2-6D-720p-comp2-light-paint-build

4 : Diagram of Makerbot Voodoo

movement_diagram

Printed Object

Makerbot Print Skull

Time Lapse Video from MakerBot On-Board Camera

Notes:

This shoot was attempted with two models from thingiverse:

Comp from print attempt 1 » 

Starting at 1 Pixel

Single NeoPixels
Team:

Tommy Payne, Natalia Cabrera & Pat Shiu

Assignment:

In the first class, we learnt the basics of setting up neopixels. For the assignment, we are to work in groups of 3 and create 3 projects that each use only a single “pixel” of light to meet the following criteria.

  1. Use a pixel of light to tell a story, convey information.
  2. Use a pixel of light to create a sculptural object.
  3. Use a pixel of light to convey movement.

LIGHT & CONTENTtime_at_LED_lab

LIGHT AS OBJECTIMG_8039-e1428105504868

 

LIGHT & MOVEMENTpixelasobject21

 

The Final Round Up

Over the past 10 weeks, our project has evolved so much. Last Thursday, we presented our project in it’s latest permutation. For the sake of the winter show, we have also finally named our project — PULSE.

10831887_766013156802514_644870000_n

IMG_9387-1024x682
Key design changes since the user tests:
  • The bed / seat has now been replaced with round inflatable lounge chairs.
  • The visuals have been completely re-worked. In response to feedback that the visuals of the smoke rings where hard to co-relate to the pulses, we redesigned the visuals to show the waveforms of both users’ pulse readings.

PULSE
FINAL PROJECT DOCUMENTATION

Two persons lay down and experience visual and sound representation of each other’s heartbeats. It’s a quiet and intimate exchange of our most vital sign.

Description

The experience requires two participants. Each of them get a pulse sensor attached to their index finger and a set of headphones, then lay down on a chaise lounge. White paper domes lower until they hover over the participant’s upper body, shielding out visual distractions from the environment and submerging them into the experience. Headphones provide the sound based on the heartbeat of each other, domes show a projection of visuals generated with heartbeat of each other, participants don’t know if they are experiencing their own or another vital sign. With time, sound gets louder, until it becomes the central point of focus and heartbeat rate tends to slow down. When one of the participants remove the pulse reader the domes come up and the experience is over.

Personal Statement

We are Namira Abdulgani, Natalia Cabrera and Pat Shiu, first-year graduate students at New York University’s Interactive Telecommunications Program (ITP). We came together over a shared interest in meditative practices and technology. This project began with questions that we most desired to explore — how can we design experiences that heighten an audience’s sense of awareness? How do we immerse ourselves in another’s biorhythms? Is the desire to be in sync — physically, energetically and spiritually — an intrinsic human instinct that exists in every one of us?

More details on the PULSE can be found on my partner Namira’s blog »  
Including:
  • Interaction System Diagram
  • Links to notes from our user studies
  • Final BOM
  • Final timeline

 

Don't Sweat the Technique