Bryce Boe

For the third year, I competed with team Shellphish in the Defcon quals. We pulled through with some amazing points at the end to finish in 8th place. My successful contributions, however, were really only with respect to Forensics 100 and 300. My write up for the following are below:

Forensics 100
The forensics 100 challenge indicated to find the key, and provided a png file that was 19025×1 in resolution. Immediately our team thought we could simply change the resolution to 25×761 and would be on to something. After working with the resulting image for sometime I finally thought about converting it to 761×25. That was our first break through when we read some text along the lines of “ILoveMeSomesheepysheepies” followed by binary that includes capital ‘O’s in place of some of the ’0′s. After no success with different permutations of that message we incorporated an idea the other team members had about the blue offset pixels that occur at regular intervals. Our first attempt at wrapping the image at the blue pixel boundaries (every 450 pixels) resulted in success! The key “thankYouSirPleasemayIhaveAnother” appeared and worked. The following is my simple python solution for Forensics 100:

1
2
3
4
5
6
7
8
9
10
11

#!/usr/bin/env python
import sys, Image
 
def main():
    orig = Image.open('f100.png')
    img = Image.new('RGBA', (450, 43))
    img.putdata(orig.getdata())
    img.show()
 
if __name__ == '__main__':
    sys.exit(main())

Forensics 300
Forensics 300 was quite an interesting challenge. I don’t have the original file, nevertheless, one had to extract the initial file with a password to get a dmg containing a dump from an iphone. I came into the challenge a little late, after one of my teammates had gone through all the images, videos, and audio files looking for Waldo and ‘grep’ing for various relevant keywords. Further more, my teammates had previously used the iPhoneTracker on the consolidated.db file to see where the phone had been, however San Antonio didn’t prove to be very useful.

While the iPhoneTracker app seemed pretty cool, I wanted to programmatically see where the phone had been the most. Thus, after figuring out what was what with respect to the consolidated.db file I wrote a little python script to find the most visited places rounded to less precision to account for some variance. The top three results were the following where the first number represents the number of occurrences in that location, and the two numbers between the parenthesis represent the latitude and longitude respectfully.

• 30 (‘-77.846′, ’166.677′)
• 18 (’0.000′, ’0.000′)
• 10 (’36.106′, ‘-115.173′)

When I did a google search for the coordinates -77.846 166.667 I knew immediately that it was no coincidence that I was centered in a small town in Antarctica. Unfortunately, Google maps doesn’t have a name for this location so I had to revert to Bing (for the first time ever) to figure out that this location is called Ross Island. From that point we simply attempted different “places” listed Ross Island’s wikpiedia page until “McMurdo Station” submitted successfully. Below is the script I used to find the coordinates from the consolodated.db input file:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

#!/usr/bin/env python
import os, sys
 
def main():
    os.system('sqlite3 consolidated.db "select Latitude, Longitude '
              'from CellLocation;" > tmp')
 
    uniq = {}
    for line in open('tmp'):
        pos = tuple('%.3f' % float(x) for x in line.split('|')[:2])
        if pos in uniq:
            uniq[pos] += 1
        else:
            uniq[pos] = 1
 
    for pos, count in sorted(uniq.items(), key=lambda x:x[1]):
        print count, pos
 
if __name__ == '__main__':
    sys.exit(main())

You can find links to solutions to other Defcon 19 Quals challenges at the following locations: Rogunix, negative foo, VNSecurity site.

Related Entries

• UCSB’s International Capture The Flag Competition 2010 Challenge 6: Fear The EAR
• Defcon 18 Quals Forensics 200 Write up
• Bye Bye Facebook: A Guide to Leaving Facebook
• iCTF09: UCSB’s International Capture the Flag Competition
• Moving Images in 3D Space with Pyglet

Tags: hacking, python