24 HOURS OF REAL WORLD ILLUMINATION

By the time I got home early on Thursday morning I had not slept in 41 hours. By then I was in a sort of dreamlike state, going over the events of the previous day in my head, hoping I had collected enough data for my project and wondering if the experience had been worth while. All in all the experience itself was extremely interesting, particularly the time between 4:30-6:00 a.m when I was the only person in the junction. Then to see how busy it became during the afternoon made me realise how relevant it was to use the phrase ‘it’s like Piccadilly Circus’, which is commonly used in the UK to refer to a place or situation which is extremely busy with people. I was asked at one point by a curious man whether I worked for Google Earth, at least it’s good to know my actions appeared genuinely professional.

Looking over the photographs I’ve noticed a couple of variables that I should consider if the results are to remain consistent. The first is a slight unbalancing of the camera during each exposure which if ignored will create a blurred and jittered HDRI which is not what I want. Also I was very conscious of the sphere getting dirty over the 24 hours with me constantly moving it around. The results of the last few hours are a little murky, however this wasn’t as bad as I had predicted and shouldn’t affect the illumination of my scene in 3D.

I’m using a program called Photomatix to convert the images into HDRI (also referred to as radiance maps). I had intended to do this in Photoshop but there is no function that allows me to ‘unwrap’ the mirrorball into a flat 360 degree panoramic image. After photographing the sphere from the side closest to the neon signs I did the same for the other side, facing the memorial. I thought this would allow for a fairly decent 360 degree view of Piccadilly Circus when both sides were unwrapped and stitched together. However, with the final HDRI next to each other the stitching process became a little tricky as both images became warped and felt out of place when aligned. Ultimately I was pretty happy with how both the images had unwrapped from the mirrorball in Photomatixs. The ‘pinching’ of the corners wasn’t as severe as I expected, leading me to the conclusion that I only need to use the images from one side, and not both. In the end I have decided to use the side closet to the neon screens as this plays such an important role in the illumination of my scene during the night time.

I positioned the camera exactly 1 meter from the sphere, photographing exposures of 1s, 1/8, 1/250, 1/2000 and 1/4000. After that I used the Digital Lux Meter to record the surrounding environmental illumination. My results can be seen in the table below.

LUMINANCE VALUES IN PICCADILLY CIRCUS DURING THE 24 HOURS OF WEDNESDAY 19TH MAY 2010

TIME – LUX
00:00 – 25-50
01:00 – 25-50
02:00 – 25-50
03:00 – 25-50
04:00 – 25-50
05:00 – 180
06:00 – 4000
07:00 – 19000
08:00 – 44000
09:00 – 58000
10:00 – 68000
11:00 – 86000
12:00 – 97000
13:00 – 118000
14:00 – 19000
15:00 – 12000
16:00 – 9600
17:00 – 7600
18:00 – 10300
19:00 – 4000
20:00 – 3000
21:00 – 180
22:00 – 25-50
23:00 – 25-50
24:00 – 25-50