Login or Join Now

Upload your photos, chat, win prizes and much more

Username:
Password:
Remember Me

Can't Access your Account?

New to ePHOTOzine? Join ePHOTOzine for free!

Like 0

Canon 7D terrible noise levels! Disappointed.

Join Now

Join ePHOTOzine, the friendliest photography community.

Upload photos, chat with photographers, win prizes and much more for free!

DuncanDisorderly

It's easy to say that I fluffed my focussing on the 60D shot and I'm not comparing like with like.
Maybe...
But please read my comments first!

Sponsored Links
Sponsored Links 
30 Jul 2011 - 1:56 PM

Join ePHOTOzine for free and remove these adverts.

Carabosse
Carabosse e2 Member 1139544 forum postsCarabosse vcard England269 Constructive Critique Points
30 Jul 2011 - 1:59 PM

Noise looks worse on the 5D2 pics?

DuncanDisorderly

Before I slag off the 60D, please bear in mind....
The 60D image included above produced a great A3 print.
So this discussion is a pixel peeping exercise that is largely academic....

Different lenses - I haven't done the 5DII any favours pairing it with the 24-105.
I find my 70-200 gives significantly sharper results. When the 70-200 is used on the 5DII it's possible to zoom in on the back of the camera and wonder why the button stops working; it's because it is already fully zoomed in but it looks like there is still more detail available if I could zoom in further - this doesn't happen with the 24-105. I'm very sure that if I took the same image on the 5DII it would look pin sharp.

Missed focus - I looked at a couple of hundred images taken at that event, all taken at ISO 2500; none of them were any sharper when viewed at 100%. I found some images which allowed me to check if I'm regularly front or back focussing; I'm ever so slightly front focussing but the sharpest point isn't noticeably better than the 100% crop I included above.

5dII is blooming good! Looking at the above 100% crops, I'd estimate that ISO 6400 on the 5DII has similar fine detail to ISO 2500 on the 60D. Which is astonishing considering the 5DII is zoomed in further because it has 21Mpixels and the 60D 'only' has 18Mpixels.

The above statements fits with my experience of producing clean prints without using noise reduction.
60D will print to A3 at ISO 2500
5DII will print to A3+ at ISO 6400

DuncanDisorderly


Quote: Noise looks worse on the 5D2 pics?

You may be right - but compare the amount of fine detail.
The 60D shot looks soft

Besides - none of the noise shows on an A3 print
If I ever need to produce a wall size poster I'll add a dash of noise reduction.
So why worry!

Carabosse
Carabosse e2 Member 1139544 forum postsCarabosse vcard England269 Constructive Critique Points
30 Jul 2011 - 2:39 PM


Quote: You may be right - but compare the amount of fine detail.

Yes, but that's always a bit of a trade-off isn't it? If you have to apply more NR then ultimately you lose some of the fine detail.

So much of what people shoot these days goes straight onto the web, without ever touching paper, that it is how it looks on a screen that has come to matter at least as much as print... if not more so.

DuncanDisorderly

Hmmm - just had a look at why the 5DII image is SO bad at ISO 2500
It isn't !!!!!

I'd sized the crops to fit inside 800px
I didn't realise ePz would downsize them to 600px

There are a lot of resampling artifact in the 5DII noise patterns
The real thing is far smoother

I'll try again - hang on!

DuncanDisorderly

Right - these crops HAVEN'T been re-sampled by ePz!

60D, ISO 2500
20110430-150413--mg-2456-3.jpg

5DII, ISO 2500
20110429-161325-img-5054-3.jpg

5DII, ISO 4000
20110501-124248-img-5730-3.jpg

5DII, ISO 6400
20110501-211822-img-6068-3.jpg

Last Modified By DuncanDisorderly at 30 Jul 2011 - 3:21 PM
DuncanDisorderly

That's better!
There's damn lies, statistics AND re-sampling artefacts!

Remember there are 100% crops.
If printed at this size, you'd need a medium sized hoarding to mount it on!

Last Modified By DuncanDisorderly at 30 Jul 2011 - 3:26 PM
Carabosse
Carabosse e2 Member 1139544 forum postsCarabosse vcard England269 Constructive Critique Points
30 Jul 2011 - 3:30 PM

On ISO 2500 noise looks pretty much the same now for both 60D and 5D2? But it's very hard to be certain as the subject matter is so different.

Test shots using exactly the same subject, in the same lighting, and the same aperture and shutter speed would be the only valid comparison. Even then you have the issue that one image is on FF and the other on 1.6 crop. I guess you would have to crop the 5D2 image to that of the 60D.

Carabosse
Carabosse e2 Member 1139544 forum postsCarabosse vcard England269 Constructive Critique Points
30 Jul 2011 - 3:34 PM


Quote: If printed at this size, you'd need a medium sized hoarding to mount it on!

For the 5D2 at the common or garden 240ppi (as used by some online printing companies) it would be about 24" x 16"... big but not billboard! Wink

DuncanDisorderly


Quote: For the 5D2 at the common or garden 240ppi (as used by some online printing companies) it would be about 24" x 16"... big but not billboard! Wink

Yup - but the screens people are viewing this with are typically rated at 96dpi

Carabosse
Carabosse e2 Member 1139544 forum postsCarabosse vcard England269 Constructive Critique Points
30 Jul 2011 - 3:55 PM

Fair point, so 60" x 40" then. Way beyond what most people would print for the whole image.

But................ if you had to do a lot of cropping, you could easily be using these images at 100%. Your 60D shot is an example of a close-up, where you might not be able to actually get sufficiently close, even with a reasonable sized tele lens.

mikehit
mikehit  56692 forum posts United Kingdom11 Constructive Critique Points
30 Jul 2011 - 10:06 PM


Quote: Test shots using exactly the same subject, in the same lighting, and the same aperture and shutter speed would be the only valid comparison. Even then you have the issue that one image is on FF and the other on 1.6 crop. I guess you would have to crop the 5D2 image to that of the 60D.

Surely if you need to go to those levels then the 60D in absolute terms is as much as most photographers would need (probably more. But in comparative terms the 5DII is certainly better.

csurry
csurry  129230 forum posts92 Constructive Critique Points
30 Jul 2011 - 10:34 PM


Quote: I tend to, out of habit, underexpose by about 1 stop, as it seems to me that Canon retains detail in shadow more than it does in overexposed highlights. So as a habit, rather than clip the highlights, I tend to underexpose by a stop. Unfortunately, the noise seems to be pretty uniformly applied to all areas of the frame, whether in shadow or mid's

Certainly on Nikon's if you underexpose and then bring back in PP you are asking for trouble.

Just try a few test shots in poor light - if you live in the UK any day should be a good day for that Wink

Shoot at a static scene at -1, -0.5, 0, +0.5 and +1 and then process. Label appropriately during processing and compare the results. I would suspect that the underexposed images should be much more noisy than those from 0 to + 1.

HTH

certx
certx  7415 forum posts United States1 Constructive Critique Points
31 Jul 2011 - 4:22 AM

@Duncan sorry.. haven't had a chance to take pics and upload yet. I've had a very busy weekend so far. I'll get it done though. Also haven't been able looked back through the latest responses on the thread yet either.

Add a Comment

You must be a member to leave a comment

Username:
Password:
Remember me:
Un-tick this box if you want to login each time you visit.