Examples of photographing clinical X-ray images
This is not an example of obsessive compulsive perfection, but more a real life example of what can be achieved when you put an Xray on an Xray viewing box and use your camera hand held to take the photo within about 20 secs of setting up as would be done in a busy work day by someone not employed as a clinical photographer.
Xray's are difficult to image digitally and retain detail in the shadow areas as well as the highlight areas.
For this one, I chose an infant who presented with cardiac failure due to severe iron deficiency anaemia showing an enlarged globular heart with plethoric lung fields.
Both of these were 10 mpixel jpegs straight from the camera taken with AWB setting and have been slightly cropped and then resized and compressed for the web.
Olympus camera settings were "Natural" mode with "Normal" gradation.
Canon camera settings were "Standard" picture style.
No post-processing otherwise including no sharpening in PS.
The above was taken with the Olympus E510 dSLR with IS on, ISO 400, low NR setting, ZD 50mm macro lens at f/5.6, 1/125th sec.
This one was taken with the much more expensive and heavier, Canon 1D Mark III at ISO 400 with EF 24-105mm L lens at 105mm with IS on at f/5.6 and 1/125th sec.
It would seem that the ISO level on the Canon is a bit more sensitive than that on the Olympus as despite the same exposure settings, the Canon image is more exposed giving less detail in the highlights (eg. behind the heart) but more detail in the shadow areas (eg. lung bases). Of course, another explanation would be the calibration of the apertures on each of the lenses differs.
Unless the image densities were identical, its hard to compare effect of the wider dynamic range of the Canon, but at least in these photos, there is no evidence that one is definitely more superior than the other, so the Olympus has done well.
Another infant CXR, this one with Fallot's tetralogy and the classic boot shaped heart:
Above is the Olympus E510, same settings as above but 1/200th sec
and here is the Canon 1D Mark III with same settings as earlier Canon shot but 1/200th sec.
Again, the Canon is slightly more exposed and thus has slightly more detail in the shadow areas, again I'm not convinced I can see the difference in sensor dynamic range - perhaps you need to shoot in RAW mode to demonstrate this ?
Next task is to repeat this experiment and try to get densities equivalent.
Inhaled foreign body - inspiratory film (left) vs expiratory film (right):
Above is Olympus E510 f/5.6 at 1/500th sec, ISO 400
Above is Canon 1D Mark III f/7.1, 1/400th sec, ISO 400
Now we are talking, image densities are much the same, but I still can't see any better shadow detail in the Canon.
Let's have a look at some 100% crops of these, 1st the Olympus then the Canon, then the Canon with USM applied:
This last image is the Canon with unsharp mask applied in PS CS2 using 100%, 1.2 radius, 0 threshold.
The Olympus image shows a bit of digital noise but the Olympus has also seemingly applied more in-camera sharpening even though it was taken in "Natural" mode which has sharpening, contrast and saturation set to zero.
Maybe that's why many recommend setting the Olympus E510 to minus 2 for sharpening, although the Olympus image does seem to show a touch more detail - see the vertical lucency in the rib 3rd from left, although perhaps this would be more evident in the Canon with a bit of sharpening but I can't see it in the last sharpened image.
If I was really wanting best image quality for enlargements with the Olympus, I would set sharpening to minus 2 and use ISO 100 whenever possible - fortunately, the Olympus E510 has built-in image stabiliser which works really well on the macro lens and allows lower ISO settings in ambient lighting.
For the web, ISO up to 800 is fine for the Olympus.
At these low ISO levels, I am not convinced of any real tangible benefit of the Canon over the Olympus (when you make the levels of sharpening equivalent), but as I alluded to above, perhaps the benefit of extended dynamic range requires RAW capture to demonstrate.
My conclusion:
Given the immense advantage of the smaller size and lighter weight of the Olympus E510, for most clinical purposes, it will be the tool of choice as the difference in image quality of the Canon is not sufficient to make up for its weight and bulk in this environment, not to mention the risk of having it stolen in a clinical setting and at 4x the price, it poses a much greater fiscal risk.