The aluminium plate is the actual mounting for the camera that will attach to a rod passing through the base of the integrating sphere.
You can see the red and green network status lights on the rear of the camera. There are also two status lights adjacent to the network socket hole. All these LEDs will be blanked off so that their light does not contaminate the illumination within the sphere.
The white sticky pad on the mount is for attaching the digital thermometer probe. The probe (25mm long silver cylinder on the LHS)and the ORP12 photo resistor (RHS) can be seen on the following image:
The camera mount has been assembled and painted matt white. With the camera body being matt black, all the components within the integration sphere are mono chrome. All the wires are either white, grey or black. This was decided upon so as not to colour cast the light within, allowing equal level responses from the red, green and blue parts of the CMOS sensor. This is one of the attempts to maximise the generated noise /entropy, the other being to reduce the light level to the minimum possible before the camera software clips the image to pure black.
Also visible on the extreme right of the camera mount is the LED for illuminating the integrating sphere. It is a 5mm white clear lens type with a forward voltage of 3.2V. It's orientation is such that there will be at least two reflections inside the sphere prior to it's light rays entering the camera lens. This is a common integrating sphere rule of thumb. Part of the experimentation will be to determine the appropriate light intensity to maximise collected entropy. There is an external dropper resistor that can be varied to adjust the LED current. It is likely that the LED itself will have to be covered somehow to further reduce it's light output.
Paul dreams some day (other than of perfecting the cold hot dog) to actually capture and measure the individual photons emanating from said LED. This would facilitate a much faster (and more random?) random number generator. Such photo counting techniques are typically used nowadays in growed up quantum random number generators producing Gbits /sec of output. Ahhh.