Here are some 40D sample images with (fewer) 5D images for comparison. These shots were taken pretty quickly, no artistry here. I had the 40D for a very short time, so my test isn't very methodical (in particular, high-ISO noise tests are not well done). See my other galleries for more 5D sample images.
Conclusions: 40D is great value; 5D images still look a little more natural (especially when looking at extra-large images) but the gap has been narrowed compared to 20D vs. 5D; I love the 40D 6.5 fps; LENSES ARE MOST IMPORTANT PURCHASE - DUH!!!!
Lenses: 17-55 2.8 IS (EF-S, so 40D only), 35 1.4L, 85 1.2L II, 70-200 4L IS; lens and aperture are noted in image captions; you can also get EXIF info
40D Settings: RAW, Adobe RGB, EV 1/3, High ISO NR, Highlight Tone Priority (could be affecting shadow noise), Picture Style Neutral, AI Servo, continuous drive, partial metering, Aperture Priority (Av)
5D Settings: RAW, Adobe RGB, EV 1/3, Picture Style Neutral, AI Servo, continuous drive, partial metering, Aperture Priority (Av)
Default CS3 ACR PP: Adobe RGB, 16-bit, Red Saturation -25, Blue Saturation -15, EV 0.00, Recovery 0, Fill 0 (7 for 5D - accidental, but not relevant), Blacks 5 (2 for 5D), Brightness 45, Contrast 30, Clarity 10, Vibrance 10, Auto WB (indoor shots only), no noise reduction (NR); differences from these defaults are noted in image captions
Default CS3 PP: SmartSharpen 70%, 0.6, not "more accurate" (then edit->Fade Luminosity), Unsharp Mask (USM) 2%, 40, 0 (then edit->Fade Luminosity), convert to 8-bit sRGB
Exposure: 40D tended to underexpose by at least one f-stop compared to 5D, particularly indoors; that's annoying
Drive: 6.5 fps on 40D is awesome and missed on 5D; found myself using this all the time; best feature of the camera
AF: 40D AI Servo stayed locked on with kids running/biking toward the camera (HUGE improvement over 20D); 5D AI Servo is no slouch either and stayed locked, too
Noise: 40D underexposed, so my tests didn't come out great; nonetheless, the 5D clearly has less noise at ISO 800 and ISO 1600; note: I will try to get the 40D again and do a better set of tests
Overall IQ: more lens-dependent than body-dependent; 5D images look a little more natural (especially when looking at extra-large images); 40D wide angle performed better than expected (EF-S 17-55 2.8 IS is no slouch, 35L is better); 85LII images look better on 40D than expected and allow one to get in closer (due to 1.6x crop); 70-200 4L IS is a great lens, but I prefer the look from a prime a little more, particularly from a fast prime; 40D images are more contrasty out of the camera for some reason, perhaps to a fault of looking a little fake; 14-bit vs. 12-bit advantage on 40D did not show through in my tests (unless that explains the contrast difference); 40D images did not need more PP - contrary to what has been reported in other reviews - in fact, I needed to tone down the contrast with the 40D PP.
40D, 17-55 @ 20mm f/2.8, ISO 800, no NR;
PP: EV 1.75
40D, 17-55 @ 20mm f/2.8, ISO 800, no NR;
PP: EV 1.75
Camera: Canon (Canon Eos 40d) |
Original size: 1744px x 2616px |
Current: 200px x 300px |