20D jpegs on web

I most like will get this camera (to replace 10D as backup camera, next year) and I have tried to evaluate the image samples objectively.

My initial impressions about 20D image quality is that Canon has primarily sought after low noise and fine detail and so they have

a) lowered contrast (yes slight DR increase but also by internal shadow/hight processing just like shadow/highlight tool in PS)
b) made CMOS hardware noise reduction harder (or more coarse) on low detail areas
b) set up sharpening to touch only finer detail.

The overall look is ok but also has a sligh anomaly feel which can be described perhaps as "like using a slighly bad copy of a lens". Lens resolution vs. 1.6X (small chip trying to pull up lots of data from lens) surely adds to this problem. Or maybe that is the problem?

A visible examples of that

http://www.steves-digicams.com …/samples/gr3b06​21_cr2.jpg
vs
http://www.steves-digicams.com …/20d/samples/im​g_9955.jpg (f/8!)

Same lens used.

In photo http://metasync.com/20​d/pages/IMG_6930.htm you have fine detail here and there, but unevenly - exactly like after using Neat Image or such. And as a sunny day high contrast shot it lacks real contrast and definition.

the difference between 1D MK II and 20D is pretty obvious, MK II produced more DR and much better contrast.

saving up is still a long way to go ...

I'll be playing with a 20D this friday, hopefully i'll post some images

You know, that's the time of thing that worries me. I only took a few pictures so far. But the ones outside in bright sun lot look like that, very low contrast. I mentioned this on the DR forum to see what other say. But you are kind of confirming it for me with the comparison shot.

I thought it maybe just settings I was using P-mode only didn't pay attention to metering.
Can good post processing make make it better? The shot you are posting from my camera is what I considered one of the better ones. I did not even post the ones I thought looked to "white". As I was not sure what to make of them.

Pekka wrote:
I most like will get this camera (to replace 10D as backup camera, next year) and I have tried to evaluate the image samples objectively.

My initial impressions about 20D image quality is that Canon has primarily sought after low noise and fine detail and so they have

a) lowered contrast (yes slight DR increase but also by internal shadow/hight processing just like shadow/highlight tool in PS)
b) made CMOS hardware noise reduction harder (or more coarse) on low detail areas
b) set up sharpening to touch only finer detail.

The overall look is ok but also has a sligh anomaly feel which can be described perhaps as "like using a slighly bad copy of a lens". Lens resolution vs. 1.6X (small chip trying to pull up lots of data from lens) surely adds to this problem. Or maybe that is the problem?

A visible examples of that

http://www.steves-digicams.com …/samples/gr3b06​21_cr2.jpg
vs
http://www.steves-digicams.com …/20d/samples/im​g_9955.jpg (f/8!)

Same lens used.

In photo http://metasync.com/20​d/pages/IMG_6930.htm you have fine detail here and there, but unevenly - exactly like after using Neat Image or such. And as a sunny day high contrast shot it lacks real contrast and definition.

Pekka wrote:
A visible examples of that

http://www.steves-digicams.com …/samples/gr3b06​21_cr2.jpg
vs
http://www.steves-digicams.com …/20d/samples/im​g_9955.jpg (f/8!)

Same lens used.

Are you sure it's the same lens? It doesn't appear to be.

From Steve's Digicams 20D samples:

"We used a custom SET 1 parameter with Sharpening at +1 and used a
Canon EF 28-105mm F3.5/4.5 USM lens and ISO 100 unless noted."

The 20D shot you've referred to isn't otherwise noted, so it's shot with Canon 28-105.

From Steve's Digicam's 1d Mk II samples:

"Pictures taken at ISO 100 unless noted, most with Canon EF 16-35 f2.8L lens."

The shot you've referred to isn't otherwise noted, so it's shot with the Canon 16-35L.

As you know, these are two VERY different lenses, the 16-35L would be expected to be sharper and more contrasty. In which case, not a fair comparison.

Cheers
Martin

What kind of contrast settings does the 20D have?
Also, using RAW and high contrast profile (i.e. film profile in Capture One) may help a lot.

runic wrote:

Pekka wrote:
A visible examples of that

http://www.steves-digicams.com …/samples/gr3b06​21_cr2.jpg
vs
http://www.steves-digicams.com …/20d/samples/im​g_9955.jpg (f/8!)

Same lens used.

Are you sure it's the same lens? It doesn't appear to be.

From Steve's Digicams 20D samples:

"We used a custom SET 1 parameter with Sharpening at +1 and used a
Canon EF 28-105mm F3.5/4.5 USM lens and ISO 100 unless noted."

The 20D shot you've referred to isn't otherwise noted, so it's shot with Canon 28-105.

From Steve's Digicam's 1d Mk II samples:

"Pictures taken at ISO 100 unless noted, most with Canon EF 16-35 f2.8L lens."

The shot you've referred to isn't otherwise noted, so it's shot with the Canon 16-35L.

As you know, these are two VERY different lenses, the 16-35L would be expected to be sharper and more contrasty. In which case, not a fair comparison.

Cheers
Martin

Thanks for correcting, yes they were not with same lens. But still, all lenses are practically equal on sunshine day and f8

I'm sure many of you have been to the site but for those who haven't there's an in depth review on the 20D. After reading it through I ordered mine this morning from Vistex in Toronto they may see them on Wed.

http://dpreview.com/

larryvt wrote:
... there's an in depth review on the 20D ... http://dpreview.com/

I'm just waiting for their "preview" to turn into a "full review".

I'd also like to see a review of the camera with 17-85 lens kit.

W

I just found this posting on dpreview with a shot taken with the EF-S 10-22 lens

http://forums.dpreview​.com …rum=1019&messag​e=10283888

I am used to shooting with 300d where defaults are +1 on contrast,saturation. On the 20d, and 10d, the default is zeros. So I will play with the Parameter 1, which is +1 contrast,satration,sha​rpness. This is from memory...

And sounds like Pekka comparison was with a different lens and different exposure... I hope that's all it is.

More sample photos:

http://www.dragtimes.c​om/canon-20d-ef-100-400.php

Here's a link to a full sized (3504 by 2336, 3804KB) large, fine JPEG right out of my 20D. It's a photo of one of my lightsabers.

It was shot at f/3.5 through a 28-135 f/3.5-5.6 at 28mm. Oh, and at ISO 3200.

http://www.planet-zhadum.com …emp/20D_ISO3200​_Saber.JPG

For ISO 3200, I don't think that the noise is too bad!

I know it's a bit blurred ("neutral" parameters and relatively long hand-held exposure) but I was only interested in testing the noise levels.