I’ll leave all camera body details and the list of features for the magazines and get right to the question: Which camera should I get?
(Note: I have already owned the following DSLRs: Canon XTI, Canon 5D and as of three weeks ago – a Canon 40D.)

Around Christmas, Aden Camera Toronto started selling Canon’s excellent 40D body for $809.99 CAD. As of last week, they reduced the price of the brand new Canon 50D camera body by $250 and are selling it for $1119.99 CAD. Not bad.

The latest price cut triggered the question: Is the new 50D worth the extra $400 and should I exchange/upgrade to it? Some of the new features that attracted me to 50D were: 15MP which is 50% greater megapixels, micro adjustments – referring to the feature borrowed from the Pro camera, better LCD screen, Digic4 chip, etc. Well, I made my move and purchased the 50D, however, for comparison purposes I kept my less than a month old Canon 40D as well. The reason was that I wanted to make an educated purchase e.g. did not get caught in the camera release hype…

Round 1: Tripod mounted shots

Reading the Canon 50D specs I was pretty amazed. Therefore my initial intention was to prove to myself that the new 15MP camera body offers much better picture quality, better focusing and greater details. To minimize any personal bias in my intended testing I created a “semipro shooting set”: I arranged a small table and placed some suitable items on it including the free Resolution Card from Digital Camera magazine (actual card size is a half letter size).
See pic.1:

http://img58.imageshack.us/img58/9607/unprocessed40auislq2.jpg

I placed two tripods 4’ across from the table and started shooting. At first I began attaching various fixed and zoom lenses to the two tripod mounted bodies. Such lenses were the Canon 50mm f1.8 II, Canon EFS 17-85mm f4-5.6 IS, Sigma 24-60mm f2.8 EX DG and the Canon 75-300mm f3.5-5.6 II. Well, as you can see all is (joking) “semipro” e.g. not overly expensive optics. Soon I realized that due to the variables added by each lens it would be really hard to compare the data. Therefore, all pics shown will be from the Canon EFS 17-85mm IS USM lens.

Some very important notes:

- Images are straight from the camera and no manipulation whatsoever was done to the pics
- All focusing is right at the middle of the setup picture shown above, where the magazine cover shows a tripod. I always focused on the square logo on the tripod head shown. Mind you, due to the depth of field, some 3D objects I displayed may be a bit off focus.
- Some more info is given in the title bar of the pictures shown: 40 or 50 refers to Canon 40D or 50D, AU means autofocus, IS means the 17-85mm IS was used and all at Av (Aperture priority the lowest number for the lens).
- When reviewing the images I made sure I select a picture representable for a whole group of images and it is a quality shot. That’s why the results I show represent not a lucky or bad shot but predominantly consistent performance.

After the first 20 or so shots were done, I reviewed them and then something started to change: Both cameras produced outstanding pictures (see Pic.2: 40D is on the left and 50D on the right side),

http://img294.imageshack.us/img294/6899/unprocessed40dleftand50hh4.jpg

but they were exactly the same. Except the slightly bigger size of the 50D images, I was not able to see the benefit of the added 50% more megapixels.

The only difference in resolution I was ever able to see was from the black and white bar code tests I did, where two parallel lines are just a microscopic part away. Although it depends on the distance the picture is taken from, the 50D showed a more pronounced white gap in between. In all other shooting situations when we have colours, shapes and shades it would be impossible to observe that. Here is the proof, where virtually, no detectable resolution differences can be found on the 1:1 crop from the right end of the set table where the Resolution Card is. See Pic.3: 50D in on top and 40D on the bottom.

http://img172.imageshack.us/img172/8474/focusingcanon40dbottom5ti0.jpg

The picture shows that from this shooting distance both cameras start to blur below the number 7 line on the left side of the card.

Round 2: Handheld shots

Most people in the “semipro market segment” including myself would be shooting from hand. Although I sometimes carry a travel tripod, typically I’m caught without one. That’s why I did bunch hand held shots using the IS lens I have. Once again The picture review showed that both cameras produce very good images. However in my tests 40D provides better quality images including colour saturation and dynapics, contrast and resolution. See Pic.4 showing four wide angle images: two 40D pictures are on the left side and two 50D on the right.

http://img107.imageshack.us/img107/4858/brickhouses40dleft50drizs6.jpg

Pic.5: four zoom images at 85mm and again two 40D on the left and two 50D on the right. Check the focus achieved on the glass… Unless the case is that 50D needs a focus adjustments the pics show 40D doing a better job.

http://img107.imageshack.us/img107/1139/brickhouseszoom40dleft5or8.jpg

Pic. 6. One more time, I wanted 50D to shine and did another brick focusing test. Although the focusing was not as easy as I thought the wall test showed 40D focuses better.

http://img103.imageshack.us/img103/3590/walltest40dleftand50drich5.jpg

Pic.7: My favorite test that I always do: The INDY test. For all 4 pictures shown the focus is between both NNs. 40D is on the left side, 50D on the right. One can see the DOF changing but altogether-amazing quality from about 3 feet.

http://img292.imageshack.us/img292/4659/indydetailcanon40dleft5vz1.jpg

Results, based on my expectations

Focusing:
From the nearly 800 shots taken, just couple of times the focusing of the 50D was a little better than the 40D. Possibly due to required adjustment issues. However for the vast majority of pics both focused really well and consistently.

Sharpness:
Rather than shots of barcodes no increased sharpness was noticed on the pics. Actually lost of detail was consistent on the pictures from the 50D (discussed further down).

Colour saturation:
The colour saturation of 40D is (not by much) better than that 50D. I tried some post processing and saturated the slightly “bleached” colours of 50D, but it got me into dealing with each colour individually and I abandoned the approach. Most of the time 40D provided much better shading range resulting in more details on the pictures. Again, 50D showed lesser dynamic range.

LCD display:
The new screen is obviously cleaner. However, soon after I started using it, I personally got used to it. Switching back to the 40D’s screen did not cause even the slightest discomfort.

Micro adjustments:
One of the selling points of the camera was the in camera micro adjustments. When I used it I found that my lenses did not really benefit from it and never tried it again.

Megapixels:
That is probably the marketing tool Canon uses the most. Well for me it produced nearly twice as big JPEG file size, but did not get me anywhere near the expected picture improvement. The 50 percent increase in megapixels did not show even 1 percent increase in quality.

Review conclusion:
I can say that feature wise Canon 50D is the better “feature packed” camera. Offers more refined use and flexibility. However image wise both cameras performed exactly the same. Since I love Canon cameras I do not want to go to extremes and say that 40D is the better camera. Many settings like sharpness, colour saturation and contrast can be adjusted by the user but definitely in my test pics I noticed that 40D provides cleaner, colour rich surfaces and detail just everywhere you look.
Therefore personally I cannot justify using bigger file size images but still get same focus, resolution, saturation and detail. If both cameras were priced the same, I would’ve thought which one to keep, but not at 50 percent ($400) higher price. Based on my tests I decided to stay with the 40D and use the money and go on a photo trip, thanks for the idea Rob …

Recommended:
Yes

Amount Paid (US$): 663
This Camera is a Good Choice if You Want Something… Solid Enough for a Professional

When you have more pixels overall, there are some interesting things that happen that can compensate for the reduction in size. For one, you have more samples in a given area to extract color information. The individual photon sites do not see in color. There are tiny red, green and blue filters mounted over them in a known pattern. Software embedded in the camera’s electronics constructs color by analyzing the relationship of the colors it measures in each site relative to the neighboring sites. This is known as a Bayer pattern, or Bayer interpolation. I would assert that generally when you have more pixels in a given area, you’ll get a better color fidelity overall. Wikipedia has a good page on this – search on ‘Bayer filter’.

Many of the lenses with a ‘DG’ or ‘designed for digital’ designation are also built to allow the incident light to strike the sensor at a steeper angle, i.e., closer to the lens axis. The idea there is that film is unaffected by the incident angle, but a digital sensor actually has millions of photon wells with microscopic filters over them. So you’d get better sensitivity and color gamut from a digital lens. Personally, I think this is all a bit too much hype. I believe that using a full-frame lens on an APS-C camera just gives you more of the sweet spot of the image circle. But you can’t go the other way and a Canon EF-S lens, for example, will not physically mount on a full-frame camera.

Circular polarizers are made to be rotated. If you buy one, you’ll either get one that rotates in its’ housing, or attaches to a rotating filter holder from Lee, Cokin, etc. With a polarizer, you generally focus and compose the shot and then rotate the polarizer for impact. If the lens racks in and out autofocusing after this point, and you are focusing on the same point, the polarizer should wind up in the same place. If you change the focus point (specifically, the distance between the film plane and the focus point) that won’t hold true and you’ll have to adjust the polarizer. The other issue with the autofocus turning the focus ring on the barrel of the lens during auto-focus is that your hand is going to be on the lens :^). There’s a simple solution for this: turn off the auto-focus. Sometimes I’ll put the camera on a tripod, autofocus, and then turn off the autofocus (although I don’t know if those lenses allow it) for the rest of the shooting.

We’re talking about very deliberate, tripod mounted shots for this application. The polarizer issue is much more acute when you are outside on a sunny day with a bunch of kids in a pool or something, and you keep moving the camera from portrait to landscape orientation.

What may be a bigger issue is that it can be very hard to rotate a polarizer when you have a lens hood on, although this is less of an issue with a wide angle lens because the hood has a much broader angle. Having spent a day eating my own dog food, i.e., shooting an LCD monitor with a polarizer, I’m a bit less sold on using it for contrast control, but I was doing a close-up of the screen only and needed to see a good rendering of fine detail on the screen.

As a photographer, I always purchase polarized sunglasses, which is a great way to get a simple preview of what’s going to happen. That’s why people will sometimes see me looking up at the sky moving my head from side to side. I’m checking out how the sky looks through different rotations of the sunglasses. On a super-clear day when you are looking at a right angle to the sun, the sky can turn almost black from the polarizer.

Great input! It does bring up a few additional questions (a newbie, with ‘real’ digital photography).

When I followed your link to a discussion on APS-C sensors, the author said he had both APS-C lens and full frame lenses. He did not indicate that he used the full frame lens on the APS-C camera, but it seemed to be implied. Can you do this? What would happen if you use a lens designed for a full sized sensor on an APS-C camera. I would think that a portion of the image would be lost as it would miss, or fall outside the edges of the smaller APS-C sensor, but I am am having trouble visualizing the full effect of doing this. The image that is received on the APS-C sensor would be cominng more from the center of the lens, which I think would be a good thing. You would not be getting the full benefit of all the incoming light that would be available, so effectively resulting in a smaller apeture openning than where the lens is set. And finally, would there be issues with some of the ‘excess’ light buncing around and causing image problems?

Also, if I go with the Sigma 10 – 22, it does not show IF in its model name, so I believe this means that the outer barrel of the lens would rotate as the camera focuses. Do the circular polarizing filters have an independently rotating portion that allows for rotation of the filter once it has been mounted to the lens? I have seen comments in the lens descriptions for those with internal focusing touting that filter users will appreciate that the outside barrel does not rotate during focussing.

Thank you,

John