Canon EOS 1000D/Rebel Xs IR – Bayer matrix removal

 I purchased this dead camera to get the hopefully intact CMOS imaging sensor.


How to open the Camera

IMG_5483 IMG_5485 IMG_5487 IMG_5489 IMG_5482

  • Four screws on the bottom from the left: medium, short, short, long
  • Right side: 2x medium
  • Left side: 2x medium
  • Top: medium, long
  • Viewfinder: bottom 2x short, up left little longer

Now release the plastic backside with screen and keys

IMG_5494 IMG_5497

The mainboard is held in place by 4 screws and one at the ground shield. Unplug all 13 connections by pulling at the tiny hole, if available. A color separated part of a connector is a lever which has to be tipped upwards to unclamp the contacts first.

IMG_5498 IMG_5500

There are five screws around the sensor. Don’t forget the connector of the piezo dust off shaker.

IMG_5511 IMG_5512 IMG_5514 IMG_5520

 Bayer matrix removal

The idea was to use ultraviolet radiation which would bleach the colors of the Bayer matrix. This will increase the sensor’s spectrum (band width), brightness, sharpness and signal to noise ratio with the lack of color. Usual glass is opaque for strong UV, so the three glasses in front of it were removed. The firs glass is a dust repellent piezo driven shaker. Underneath is an infrared filter which additionally produces a little blur to hinder moiré effect. (->see Nyquist-Shannon sampling theorem) The last glass is directly bonded to the sensor case as protection and encapsulation.

 IMG_5540_c IMG_5619_s

After the UVC exposure there was no color degradation, but the epoxy resin became brittle and can be wiped away without much force.


With removed Bayer matrix, the sensor lacks color but gains more light.

Back focus compensation

The optical system is incomplete without the three glasses. They virtually shifted the focus plane towards the lens. Plane parallel plate: the geometrical shift can be calculated with the thickness d and the glasses’ refractory constant n:


The thickness d was about 2 mm and n is set to be 1.6 so the sensor should be shifted 0.75 mm to the front.

Another method is to use a focus test chart in the distance g, the lens’ focal length f and measure the focal error e:


e=13cm, f=50mm, g=100cm; a=0,4mm


Sensor filter system

The piezo dust shaker glass appears to be a good optical filter for visual light from 410 to 680 nm. See below the continuous spectrum by an incandescent light bulb in my spectroscope:


 Astronomical purposes

Taking photographs from week light sources like stars, the debayered sensor gathers more photons. Especially with attached color filter for H-alpha it is a win in brightness and resolution. As only every fourth pixel could see the red light, a gain by 4 can be achieved.


(C)Langhans 2015



Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s