40D is a pixel-peepers nightmare.

Ok, so I upgraded from a 350D to a 40D.

As a self-confessed pixel-peeper, I often click on '100%' and use this as a judge between exposures for 'keeperiness'. You know, look at edge sharpness, colour aberrations, exposure etc.

However, I had taken a shot at f/2.8 17mm, and noticed some pretty harsh purple fringing - white pants, dark foliage, you know the drill. It shocked me, because I had never seen it as bad before. I did a search for shots with the same lens, aperture and focal length, and looked for similar shots- lo and behold, some of my older shots from the 350D were probably similar, at least with respect to how much the image had colour aberration, and what I was seeing was probably the same degree, but magnified with more pixels.

I know I shouldn't be concerned - I'm never going to print out that large to see - but is this the end of my pixel-peeping ways? I think I just have to get used to the 40D and the way it renders - you know how you become familiar with a camera. I believe you get familiar with stuff like judging sharpness, exposure etc from the LCD screen (your mind compensates - like when you magnify, but learn to spot what is sharp, even on a small screen). I have to relearn all that with the 40D. Do others find that happens when making a significant upgrade for a body?

While I'm here, from the same shoot, I was outdoors in the shade, and someone was wearing a very bright red shirt - it appear in my shots to have reasonably severe 'clipping' at the edges. Is it possible to 'over-expose' colours? You know, like how whites can be blown - can you do it with reds? Or is it just my monitor/colourspace (sRGB)/difficult lighting?

Can you post an example. I have a 5D, but before that a 20D and 350D....and I like to crop and view and 100% on occasion to check sharpness and don't recall having seen anything to get into a twist over. What you have to remeber is that even modern cameras are not flawless. 5D is the first DSLR I've been very happy with, however even with the lastest large formats cameras you could no dount still pick faults with it. Could it be your monitor, do you calibrate it regularly ?

40D:
ghosting at 100%?:

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by TINYPIC

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by TINYPIC

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by TINYPIC

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by TINYPIC

HTTP response: NOT FOUND | MIME changed to 'image/gif' | Redirected to error image by TINYPIC

Not an expert on this but in my opinion, since it is happening on the same lens, maybe its the lens at fault.

On your last image, you mentioned about the reds being overexposed. The background is also underexposed and there is a few stops difference between background and foreground. I think your expectation is too much from the camera.

According to this article, at 12mp, APS-C sensor has reached a limit and streches your lens one step too far.
http://www.kenrockwell​.com/tech/full-frame-advantage.htm

Here is the bombshell for people with no time to read it:
$100 zoom + FF is sharper than $3k lens + best APS-C.

I wonder why the net hasn't boiled with arguments about this claim yet.

post!
(sry, this is just so i can read that link when i get home)

The lens determines the degree of chromatic aberrations and the heavy purple fringing, not the camera. The camera merely records the image. Yes, it can look more severe on higher megapixel cameras because there are more pixels to record the image.

Yes, individual colors can be overexposed and Canon dSLRs are known for overexposing reds. You can have a normal brightness histogram but if you check the RGB histogram, reds can be overexposed. I always use the RGB histogram, myself. If you have overexposed reds in an image you can recover some lost detail by desaturating the reds a little bit in post processing the image. That will allow more details recorded by the blue and green channels to show up a bit more.

fe3836 wrote in post #3979997
Here is the bombshell for people with no time to read it:
$100 zoom + FF is sharper than $3k lens + best APS-C.

Not exactly.

Canon zoom + Canon FF camera is sharper than Nikon APC-C and Nikon prime.
No big news for me here

And ... you are slightly exaggerating.

Neither is the old Canon 70-200/4 a '100$ zoom' (that would be the kit lens),
nor is the Nikon 85/2 AIS a $3K lens. It is certainly a good lens (on par with the Canon 85/1.8 ), but it is a $ 300 lens.

Best regards,
Andy

Hi

IMHO with every new camera or lens comes a new learning experience the settings that were used on your xt will be different than what the 40d needs for the same job.

The CA in the pics posted are usually from the lens.

As the MPs go up, the demand on the lens increases.

The red shirt is over exposed and maybe over saturated. Try the faithful setting.

I am wondering if the anti-aliasing filter in the 40d is weaker due to the higher MPs?
which would make lens issues more apparent

Kevin

What lens are you using? The 17-55 IS. I have seen similar CA on that lens at 17mm. A good camera will bring out the CA a bit more in my experience. I have heard many 5D users complain about CA with lenses where they never had the issue before getting a 5D. The 5D might just be exposing what was always there and could not be seen.

for the record, I'm neither complaining nor expecting much. Just curious.

I realise chromatic aberration has nothing to do with the body - but as resolution goes up, you would be recording the same aberration, lwhich was once, let's say, 1 pixel, now recorded onto two pixels. When zoomed at 100%, of course it would look worse on a higher resolution zoomed to 100%.

I was just musing about the how pixel peepers are becoming even less relevant, especially when coming from a lot lower resolution equipment (and yes, I'm not speaking of the 350D here- I realise that 8-10MP is negligible).

As for my red issue - Iw as just wondering if it was possible to over expose reds (or any other colour for that matter). As airshaq20 explained to me, perhaps the background under exposure vs subject overexposure is simply within the tolerances of the dynamic range, and characteristic of these harsh lighting conditions (which is why I mentioned my shooting conditions). I was just wondering if it was to do with picture styles - in that the 'standard' mode setting may be over saturating over oversharpening my images also, exacerbating my issues.

Just curious to learn more about my equipment - not complaining.

mrkgoo wrote in post #3980179
As for my red issue - Iw as just wondering if it was possible to over expose reds (or any other colour for that matter).

Sure. And you've done it.

I just had a look with the eyedropper in PSP - I'm seeing RGB values like 255, 78, 114 all along the top of the shoulder.

-js

fe3836 wrote in post #3979997
According to this article, at 12mp, APS-C sensor has reached a limit and streches your lens one step too far.
http://www.kenrockwell​.com/tech/full-frame-advantage.htm

Here is the bombshell for people with no time to read it:
$100 zoom + FF is sharper than $3k lens + best APS-C.

I wonder why the net hasn't boiled with arguments about this claim yet.

In case you didn't know, Ken Rockwell is not held in high regard here. There are threads dedicated to him and some of his boneheaded quotes. FYI.

Andy_T wrote in post #3980149
Not exactly.

Canon zoom + Canon FF camera is sharper than Nikon APC-C and Nikon prime.
No big news for me here

And ... you are slightly exaggerating.

Neither is the old Canon 70-200/4 a '100$ zoom' (that would be the kit lens),
nor is the Nikon 85/2 AIS a $3K lens. It is certainly a good lens (on par with the Canon 85/1.8 ), but it is a $ 300 lens.

Best regards,
Andy

Here's what Ken Rockwell has to say about Canon vs Nikon:
http://www.kenrockwell​.com/tech/nikon-vs-canon.htm
If anything, I feel he's biased toward Nikon. It'd be odd if his intention is to trash them. But who knows, maybe Nikon didn't give him a free D3 and he got pissed ? Anyway brand battles belong to different threads, let's not go there .

About the Canon lens, OK, I exaggerated. Following the link in the article onto B&H, I find it used (it's a 20 years old model) in E+ condition for $174, not $100. Sorry about that .

Now a real appology for the Nikon lens. It's not $3k, he merely said it was "my very best lens in this range", and he claims he owns every lens under the sun.

But here is the brand-agnostic quote that is relevent to OP's question
"I get better results on full-frame with crummy lenses than I do with my very best lenses on DX."

gooble wrote in post #3980224
In case you didn't know, Ken Rockwell is not held in high regard here. There are threads dedicated to him and some of his boneheaded quotes. FYI.

Damn, I should've known. He's a Nikon guy !

What I'm really intersted in though, is a serious rebuttle of his findings. How did he distort the facts? How did he manipulate the A/B comparison images he posted? Don't we all learn something when a myth is debunked ?