Having 'upgraded' to a Hutech modified Canon 40D Digital SLR, I was keen to see how it compared to my existing Hutech 20D and what the optimum settings were for ISO and sub-exposure length. The images and results here for the 20D and 40D should be comparable with a 350D and 450D respectively since they use the same sensor.

Cropped images as displayed are 50% resolution, click for full size composite.
Full set also displayed separately in the Gallery.

Test Details:
Canon EF 200mm f2.8 lens at f4.
Both cameras are Hutech modified Canon SLRs, shot with an LPS-P2 front filter.
Each image has a total exposure time of 12 minutes. ISO setting and sub-exposure length vary.
Temperature approx 10 degC (50F).
RAW Conversion and Image Alignment performed in ImagesPlus3. (You'll have to take my word for it that the lack of dark frame calibration does not affect the comparison).
Digital Development was applied using ImagesPlus to the Red/Green/Blue channels separately. The Background weight was set to .50 and then the break point automatically selected by ImagesPlus for each channel. Identical levels and saturation adjustments were then applied to each image in Photoshop.

Findings:
Looking at these images and the wider cropped files in the Gallery, the first observation is how similar all these images look. Although taken with different cameras and widely varying settings, the end result at first glance is very similar.

• The 40D delivers higher resolution than the 20D, but you have to look very closely.
• The higher bit depth of the 40D verse the 20D is evident in bright areas, but using a sub-set of shorter exposures with the 20D and masking that over bright areas could easily overcome this slight difference. Remarkably, the 40D held almost full detail in the bright central portion of the Eta Carinae nebula even using ISO1600 for 4 minute sub-exposures at f4.
• There is little difference between ISO1600 and ISO800 with the 40D, with 1600 perhaps having a slight edge. Noise increases (only slightly!) at ISO400 and ISO200 without delivering any benefit in saturation or bright area detail.
• Even very short sub-exposures can deliver a great result. Image quality only improves slightly with sub-exposures longer than 60 seconds. The main advantage of longer sub-exposures is reduced demand on hard drive space. However, longer sub-exposures may become more important with slower optical systems.
• The 20D (350D) is still an awesome camera. In this test, the 20D recorded as much nebulosity as the 40D. However, the 40D did reach fainter stars, which is perhaps a result of higher resolution rather than 'sensitivity'?