7D test images taken today (high ISO)

[FrankLamont]

I am rather disappointed with ISO 1600 - and higher ones, too - on the 7D, but with no NR and on a 1.6x 18MP sensor, it isn't bad, considering. I'd say it's as good if not better than the 450D/500D.

[Ads]

[eighty4]

Noise aside, my friend recently got one of these, his only complaint with it is it tends to overexpose in almost every shot. I read on dpreview.com that it is a pretty well known problem.

For that reason alone it put me off buying one, I've had way too many problems with the Nikon D80's metering

[rufus5150]

Should average them out. Mine seems to underexpose by about 1/2 a stop. :P

[FrankLamont]

How much so? Half or one whole stop is fine for me - I prefer that over underexposed metering.

[JamesMason]

Noise ?

Wow i really need to upgrade my d80

[CyclonePWR]

Cool thanks for taking the photos.

[musicaleCA]

Exposure problems are resultant of the new metering system which takes into account colour and active AF points. Operator error shouldn't be ruled-out. I've had some exposure problems in difficult situations, like hockey rinks. The wash of bright white, coloured jerseys on the active AF points, and dark, dark backgrounds (the seats), throws the meter through a loop. It tends to over-expose the ice a tad (1/3-1 stop), but properly exposes the player in focus. Solution: Know what the meter is doing, and compensate for it. In this case, take manual control and get consistent exposures without bleaching the ice white, then use some fill light in post.

[Sam6644]

Thanks for the write-up!

I'm still planning to get a 7D sometime in the next couple months, and these photos have definitely shown me that this camera will be VERY usable at high ISO for the stuff I do most of the time.

The photo at 6400 would be more than acceptable for news paper printing.

[inTempus]

Keep in mind these are RAW images processed with no noise reduction. I didn't shoot in JPG mode with NR enabled. I imagine the ISO 6400 shot with NR would look pretty darn good, either applied in camera or in post.

The only thing that I saw that rang true with previous reports was the images looked a tad bit soft when compared to the 5D Mark II. The 5D2 images are exceptionally sharp right out of camera compared to the 7D. But with sharpening applied the 7D images look pretty good.

I'm still not sold on the color metering of this camera. I'm glad Canon left it out of the 1D Mark IV. I'm sure the next iteration in the next generation bodies will have more of the kinks worked out.

With that said, I would still like to have one to use for travel. It's still smaller than the 5D2 and I really do like the new features. The view finder and AF system really impresses me. If the 5D Mark III has the same arrangement in it's next iteration, I will be all over it!

[Montana]

Thanks for the sample images Tim. I am personally skipping the 7D, keeping my 5DII and added to the list for the 1DIV. I fully believe that most of the "issues" on the internet about the 7D are user related. Its a whole new ball of wax for Canon users. That and the aftermarket editing software hadn't caught up yet. I think it (the 7D) will pan out to be a homerun for Canon and a true step in the right direction. I would venture to call it the top crop body camera. I have seen some stunning results from it.

[musicaleCA]

Montana: Indeed. In your case it probably makes total sense to just hold onto the FF and get a 1D series camera (okay, okay, so I'm tossing 1.3x in the same boat with 35FF; sue me ).

Here's a good review from Scott Bourne on the 7D. He's a Nikon guy now, but has extensive experience with both Nikon and Canon gear (having switched from Canon to Nikon recently). He notes how much of the negative reviews of the 7D can either be attributed to user error, or overlooked key factors (like RAW processing and sharpening) in the reviews. He also notes that the AF bloody well does work (which a few of us with 7D's have been saying for awhile, strengthening the argument that it's user-error when people run into focusing problems, not the camera) and that the noise is also a different kind of noise, which appears more grain-like, and is less abhorrent to the eye, which I also noted a while ago.

[Sw1tchFX]

Montana: Indeed. In your case it probably makes total sense to just hold onto the FF and get a 1D series camera (okay, okay, so I'm tossing 1.3x in the same boat with 35FF; sue me ).

Here's a good review from Scott Bourne on the 7D. He's a Nikon guy now, but has extensive experience with both Nikon and Canon gear (having switched from Canon to Nikon recently). He notes how much of the negative reviews of the 7D can either be attributed to user error, or overlooked key factors (like RAW processing and sharpening) in the reviews. He also notes that the AF bloody well does work (which a few of us with 7D's have been saying for awhile, strengthening the argument that it's user-error when people run into focusing problems, not the camera) and that the noise is also a different kind of noise, which appears more grain-like, and is less abhorrent to the eye, which I also noted a while ago.

I agree with him about the noise comment. In the 100% crop of the monkey picture last page, noise can be seen, and that's because it's 18MP in an APS-C sensor. The amplifier really has to kick in. Same things happens to point and shoots.


And yes, if canon had kept the MP at 10 or 12, it could possibly be the best APS-C camera you can get. Sometimes marketing though gets in the way of the engineers.

[Montana]

Sometimes marketing though gets in the way of the engineers.

This is very true, ......so very true. However, I must say that most (like 85%) of the prints I sell are 18 inches or larger on the long side. I don't mind the higher megapixels myself, but my sales are not typical of everyone elses I am sure.

[inTempus]

Here's some more comparison crops from the 7D and 5D shot at ISO 100.

7D no sharpening:


7D with 70% unsharp mask in CS3:


Here's the 5D2 crops.

5D2 no sharpening:


5D2 with 70% unsharp mask in CS3:

[rufus5150]

7D with 70% unsharp mask in CS3:

What were the radius and threshold?