50D a sleeper camera

[Peter Gregg]

I've been playing with a Canon 50D and it is a sleeper camera. The files pretty much compare to the 40D and are actually a little bit better at the high ISO settings.

The big story though is not in the noise levels when looking at the files at 100 percent, the story is what happens when you print them on paper. The big wow happens here and the camera actually produces an excellent large print compared to the prints from the 40D.

This gives an advantage I see very little written about on the internet. Mostly, people are disappointed because they didn't get what they thought they should get. In fact they are getting more. The actual prints at 12x18 and 16x24 are bordering on medium format quality level.

The weakness of the camera is in the focusing abilities compared to similar Nikon's. On the other hand, if you buy a few extra cards and fire more pictures than normal, and up your skill level on focus techniques your chances of getting "wow" images become pretty high and that makes the camera a great buy.

Peter

[MarkAlsop]

That is interesting…

I was looking at the 50D as a backup/knock-around body but the DP Review of the 50D indicated that vis-à-vis the 40D the difference in actual resolution/detail was just about nil; they said that “…if you scale the 40D's output up to match the 50D the results are almost indistinguishable.”

It sounds like you are saying that more pixels are producing better images on paper regardless of what the images look like on the screen.

[DougAxford]

I have ALWAYS said that people should not be comparing camera output on the screen, it must be on the output you are normally using and that is rarely on a screen.

I can't comment on the 40D vs 50D except to say that I was anxious to upgrade, then saw the many reports of non-existent image differences with some reporting lower quality at high ISO.

I assume that Peter tested both cameras with high ISO and on print output - right?. The one thing that really peed me off in reviews of 5DMII is that on high ISO tests it seemed to underexpose slightly and all reviewers left it at that setting. Everyone knows that slightly underexposed high ISO images are much noisier. You've got to compare properly exposed images. Yes, I understand that the 5DMII should have exposed better but at least show BOTH results.

DougA

[Noel_Carboni]

The 1.6x Canon "prosumer" model cameras have always been a great and functional add-on/alternative to the more expensive pro models, and the current model is certainly no slouch. Personally I haven't been compelled to upgrade from the 40D to the 50D as even the 40D is capable of quite stellar results.

Don't believe everything you read about focus capability being lacking, either. I think those reports must be written by really, REALLY picky people. In real use I haven't lost a shot due to focus problems in recent history. I do prefer to use the center focus point only, and in this mode focus is fast and virtually flawless. I used both ONE SHOT and AI SERVO focus on my 40D just day before yesterday at a track meet and in both cases I was able to capture competitors running, handing off batons, crossing the finish line, etc. without fail with my 100-400 lens aperture set wide open.



-Noel

[mcolby]

pixel density is a wonderful thing.

I had a hard time understanding all the reviews that insisted that the 50D's had little benefit from 50% more pixels than the 40D. I could understand slightly more noise than the 40D at a given ISO, given the noise signature of 50% more photosites on the same size sensor. I just didn't understand how so many reviewer's analyses concluded that the there was not image benefit to 50% more pixels.

A large part of getting a crisp stunning large print is having enough pixel density to distinguish the details in the print.

[Noel_Carboni]

I think what they're saying - and I don't mean to dis the 50D - is that the antialiasing filter and/or lenses are such that the extra pixels are not really carrying that much more detail simply because the image isn't focused sharply upon them.

What higher pixel density generally means, with a Bayer-patterned imager, is that the color information will be more accurate, because the demosaic operation has a higher density of varying color filters to work with.

However, the reviewers are right - because of factors other than the megapixel count alone, the end result coming out of the 50D camera isn't as much better than the camera's predecessor as you might think.

-Noel

[Peter Gregg]

Originally Posted by DougAxford

I assume that Peter tested both cameras with high ISO and on print output - right?. The one thing that really peed me off in reviews of 5DMII is that on high ISO tests it seemed to underexpose slightly and all reviewers left it at that setting. Everyone knows that slightly underexposed high ISO images are much noisier. You've got to compare properly exposed images. Yes, I understand that the 5DMII should have exposed better but at least show BOTH results.

DougA

The print is where it's at, and yes I am speaking of testing on print output. I have an Epson 9880 which gives me a very different view of the world than does the computer screen.

If all a person is going to do is make images for the computer screen, the TV screen, and 4x6 prints occasionally then all the arguing on the internet is totally irrelevant. Only only ONLY when you take an image and print it at 12x18, at 16x20, at 30x40 does ANY of the conversation online about noise really apply. Of course I am speaking of mostly 2007 and newer cameras - heck, the Olympus E3 would be one of the top buys on the market if 8x10 prints or smaller is where you are at.

Cropping (hello Nikon D3X) and large prints make the cameras perform and expose what they really are about.

On the 5D Mark II darker images issue - you know, I agree with your thinking 100 percent, but yet when it came time to put up comparison shots I felt safer and more comfortable showing what the camera really does even though I know better. Go figure.

My 5D Mark II, my good buddy's 2 5D Mark II cameras, and a "guest" 5D Mark II camera all underexpose the same exact way. So it's either the photographer (me) or it is the camera - lol. Seeing how my D3, 40D, XTi, and D3X do not exhibit that behavior, nor did my past 1D Mark III nor the 1Ds Mark III that is sitting here at the moment do that. So it may be an unconscious decision to expose the fallacy of the camera, or it may be me protecting my butt from the folks who cry foul when one camera is treated differently than another in a test situation by my posting the underexposed shots.

By the 5D Mark II underexposing it's images it actually causes it to exhibit LESS noise, not more noise. However - and this is a BIG however, when you go to correct the image and you give it a 1 stop gain in exposure to make things right, you are in effect adding 1 stop of noise to your image which IS causing more noise to be in the image.

On the testing I did for the 1600 Club images at 20x30, the cameras WERE exposed correctly. I believe this is the correct way to test a camera's performance and so that is the way it was done. An underexposed image is not acceptable for tests on sensor performance.

Canon's biggest biggest biggest weakness is it's focus system. It is just plain sloppy. I have had soft images from my photo work all the way back to the D30 until today. The 5D Mark II is LESS sloppy, but the sloppyness is NOT gone, just a lot better. When I added Nikon to my bag it was a revelation in the higher percentage of sharp images I was able to bring home.

The Canon CAN be tamed, but one REALLY needs to know the focus system and how to help it at times when it CAN be helped. The flashes AF Assist light is one of the players regardless of an image needing or using flash or not. i will be writing an article on that in the near future.

Peter