
he above 100% sample crops were taken under controlled conditions using a Westcott
Spiderlite TD5 with 5000k fluorescent bulbs and a softbox. Identical exposures were manually
set - 1/30 and f/8 were the ISO 400 settings. The shutter speed was halved for each 1 stop
increase in ISO setting. Auto white balance and neutral Picture Styles were selected and
sharpness was set to "1" - a light but normal real world setting (noise gets sharpened). All
images were shot as RAW files, no noise reduction was used. The RAW files were processed
to 16-bit TIFF files, imported into Photoshop and output as a JPG using "Save For Web" at a
quality setting of 80.
I usually select a lighter
Color Checker block to use for my noise comparison, but in this case, I
felt the black block best showed the noise differences I'm seeing in real life shots. It is also
one one the noisiest blocks. Noise varies greatly from color/shade to color/shade, so it is
difficult to concisely, visually show the complete differences. Actually, the
Canon 1D Mark III
does not look its best in the above comparison as it is clearly the low-noise leader among this
group. The 40D's noise improvement over XTi/400D ranges from small at ISO 100 up to very
noticeable at ISO 1600. Below is another comparison ...
he above shots (100% crops) were taken with a tripod-mounted Canon EF 180mm f/3.5 L
Macro USM Lens
with available noise reduction turned off. I mention tipod-mounted not
because this is a detail/sharpness test (the above crops were not taken from the focus point
of the shot - and don't confuse noise with sharpness) but to explain the subject size
differences. The size of the subject in each image block is due to the sensor density of the
camera used - The lens was not moved between camera changes. You are looking at the
back window of a BMW - with a headrest inside and another SUV's mirror on the other side.
ll were shot with identical exposures (you can see the slightly lower sensitivity of the 40D
Comentarios a estos manuales