fredmiranda.com
Login

Moderated by: Fred Miranda
  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1       2              4       end
  

Archive 2005 · canon 20D and nikon D200 samples

  
 
jeffH70
Offline
• • •
Upload & Sell: Off
p.3 #1 · canon 20D and nikon D200 samples


John,

If you meant the sharpening, it was from Chuck Westfall. However, he was talking about the consumer DSLR raw file (.crw) from 10D and you can almost completely turn off that in the raw converter, almost though to -2 while this option is not in camera as 0 is the minimum. For 20D, it shares the same raw file format with the bigger brother 1 series (.cr2) so I am not too sure if Canon still does the same thing or not.

Jeff



Dec 23, 2005 at 02:15 AM
Arka
Offline
• • • • • •
Upload & Sell: On
p.3 #2 · canon 20D and nikon D200 samples


The Nikon files at higher ISO look just awful compared to the Canon. Look at those window panes in the dark areas... with the 20D, you can see the pane dividers clearly. On the Nikon, they are amorphous blobs of color that lack any real definition. This coming from a camera that has more photosites, and with a lens attached that, by all accounts, should substantially outperform the 17-40 f/4L.

I was hoping that this would be a categorically impressive release, but these tests don't really encourage me in the area of image quality. I look forward to seeing more output from the camera in the coming months.

Arka C.



Dec 23, 2005 at 03:13 AM
21farms
Offline
• • •
Upload & Sell: Off
p.3 #3 · canon 20D and nikon D200 samples


i've added yet another set but these are using 85mm primes at f4.0. the histograms on these are pushed further to the right and i did the conversions in NC and DPP (finally got them both running on the same computer) instead of both in ACR. i'm not sure why--the aperture, the lenses, the better lit scene, the RAW conversion,the tripod--but the results are just the opposite of the other ISO 800 crops (admittedly, i am not a fan of DPP and always thought i got less noise using ACR with my canon files. this is also my first time using NC and, boy, is it s-l-o-w...and i've got a pretty fast P4 3.0 ghz system too). anyway, as before, i'm happy to share my RAW files with anyone who wants them...just drop me an email.

hrow, i shot everything in RAW+JPG fine and have also included the straight-from-the-camera JPGs with the EXIF intact for almost everything i've posted to provide the comparison. initially and to me, at least, the RAW converted files looked much better than the camera JPGs.



Dec 23, 2005 at 04:58 AM
Hrow
Offline
• • • • •
Upload & Sell: Off
p.3 #4 · canon 20D and nikon D200 samples


The only reason that I asked about the JPGs was to try to get the software out of the equation because I thought that it might be coming into play big time but with NC and DPP now running that becomes somewhat moot. The change is pretty dramatic and the newest set(s) is very close in IQ, which is what I would have expected. My guess was that the 20D and D200 would be about the same in IQ with the ergonmic advantages going to Nikon big-time. It will be interesting to see if that guess was valid. Thanks for all of the time you are putting into this as I know that a lot of people are very interested in your results and impressions.


Dec 23, 2005 at 07:58 AM
joesmith
Offline
• •
Upload & Sell: Off
p.3 #5 · canon 20D and nikon D200 samples


Interesting, is the 20D iso800 that much more sensitive than the D200 iso 800?
Looking at the histograms, the D200 expose was about half a stop more to the right.
Does anyone have a link for the 20D showing what is the true iso for each iso?



Dec 23, 2005 at 09:16 AM
Sam Bennett
Offline
• • • •
Upload & Sell: Off
p.3 #6 · canon 20D and nikon D200 samples


21farms wrote:
(admittedly, i am not a fan of DPP and always thought i got less noise using ACR with my canon files. this is also my first time using NC and, boy, is it s-l-o-w...and i've got a pretty fast P4 3.0 ghz system too).


You will see more noise with DPP since it doesn't do any noise reduction. This isn't necessarily a bad thing since the dedicated noise reduction utilities generally do a better job of noise reduction than any of the RAW converters do. This is definitely the case with Nikon Capture, which has terrible NR, imo.

And yes, NC is a truly painful program to try and use. People complain about DPP (I love it, personally), but my god... if my choice of camera came down to the RAW utilities provided by the manufacturer - Nikon would lose instantly. To add insult to injury, NC isn't even free!



Dec 23, 2005 at 09:44 AM
MPerdomo
Offline
• • •
Upload & Sell: Off
p.3 #7 · canon 20D and nikon D200 samples


21farms wrote:
mperdomo, does it really start to show at f8



Well, I know that the D200 is supposed to be somewhere in the range of f/8 where it starts to lose sharpness, then again, im not sure if it's that or the lens. I have heard that the 17-55 doesn't like to be shot past f/8. Maybe you could play around and see if indeed f/8 is the sharpest, or if it biases to being shot closer to wide open



Dec 23, 2005 at 10:04 AM
thepiecesfit
Offline
• • •
Upload & Sell: Off
p.3 #8 · canon 20D and nikon D200 samples


Judging by your samples the 20D shows better detail and sharpness along with lower noise.


Dec 23, 2005 at 10:26 AM
21farms
Offline
• • •
Upload & Sell: Off
p.3 #9 · canon 20D and nikon D200 samples


joesmith wrote:
Interesting, is the 20D iso800 that much more sensitive than the D200 iso 800? Looking at the histograms, the D200 expose was about half a stop more to the right.


yes, it is. i read in one of my photo mags (can't remember which but it's either popular photography, american photo, outdoor photographer, or photo but i'm pretty sure it was pop.photo) that nikons are dead on with their rated ISOs but that canons are understated (more sensitive than rated) and that is what i am seeing as well.

for example, the same exact scene in my image files DSC_0025 and IMG_2934 were shot at f4.0 but at 1/45 for the D200 and 1/80 for the 20D yet exposure is nearly identical. the main reason i chose DSC_0019 & IMG_2925 for my first comparision was because the ISO, shutter speed and aperture were the identical but, in retrospect, that means the D200 shot was slightly underexposed in comparison and hence more shadow noise.

after thinking about it some, i think what i see happening is that the D200 noise looks better than the 20D's when the exposure is to the right of the histogram...that's my working assumption at least but adds to my growing knowledge of the D200's preferences.



Dec 23, 2005 at 10:29 AM
chez
Online
• • • • • • •
Upload & Sell: Off
p.3 #10 · canon 20D and nikon D200 samples


Another test, another viewpoint.

[url=]http://forums.dpreview.com/forums/read.asp?forum=1021&message=16395785[/url]



Dec 23, 2005 at 10:30 AM
jhapeman
Offline
• • • •
Upload & Sell: On
p.3 #11 · canon 20D and nikon D200 samples


Boy, this is interesting. To be honest, having been a Nikon shooter for 20+ years that *very* reluctantly switched to Canon, I was not surprised. Very disappointed, but not surprised. Nikon has really struggled to handle high-ISO noise. The D100 was pretty poor, the D2H was awful when it had been hyped as being excellent, and while people like to rave about the D70s and D50, they still don't come close to the low luminance noise of the Canons--all of them.

For many people, this does not matter. It did for me. I labored over my decision to switch to Canon for more than a year, and then bought a 20D just to see what I would think. I was blown away at the low noise at high ISO's--it was at least 2 stops better than my D100 and D2H. The nail in the coffin were the "early release" images from the D2X, which I had a deposit on, and was guaranteed to get one of the first ones in the US. I called up and switched that order to a 1Ds Mark II and have never looked back.

What do I shoot? A lot of night photography--aurora, the milky way, astrophotography. Wildlife--where the IS on long primes from Canon was a huge bonus (and the lenses cost less), and where high ISO can be necessary. To me, these made the switch to Canon imperative. Add to that my belief that FF was the way to go (Why not? I have yet to see a *real* argument that convincingly points out that a crop factor is really the best thing. It's about price, period. The minute FF costs the same, no one will buy crop cameras anymore), and Canon became compelling. I didn't like the ergonomics of the 1-series, but have gotten used to them. I still prefer Nikon ergonomics. The ergonomics and interface of the 20D and 5D are very good, however.

Other things I learned: Like many, I believed the bunk that Canon's wide offerings were nowhere near as good as Nikon's. It's just a load of junk. I have gone back to images with my vaunted 17-35mm f/2.8 Nikon and have been appalled by the CA and flare problems that lens had compared to the Canon 16-35mm f/2.8 and the 17-40mm f/4 (a real bargain lens). The bull from the Nikon crowd that Canon uses so much noise reduction that their images look "plasticky" and lose detail. Not true at all. It seems that since the D2X, and now with the D200, that trend has switched to Nikon. They clearly are smearing detail now.

What is the point of this? Look at good side-by-side evaluations like this before jumping to conclusions based on specs. Yes, more pixels on the D200--but not enough to make a major resulotion difference if you understand the match. Weather sealing? It's minor, not in the same class as the D2-series, and not going to give you much over the 20D--besides, the bulk of Nikon lenses don't have a weather seal, so what's the point? The water will leak in anyway.

In the end, these are tools, not toys (OK, they are toys, too I will admit it!), and you need the right tool for the job. If you do a lot of low-ISO shooting, the Nikon's are superb. If you have to bump up to higher-ISO shooting, they lose their edge, that is pretty clear.

If I could have my Canon sensors in a Nikon camera with my Canon glass....I would be a very happy person. Oh, except for the Nikon 200-400mm VR and the 200mm macro--I would want those lenses back in my stable, too.

Thanks for the great comparison. At some level, it makes me sad that Nikon can't put a little more pressure on Canon, but glad I didn't wait even longer to switch to a tool that better fits my needs. I am sure we will see more of these, and it will be interesting to see if the trend that I have seen in multiple posts already continues to hold. And let's not forget that the 20D is up for replacement this spring, based on Canon's history--and contrary to what many think, a higher-MP version of the 20D will NOT eat into sales of the 5D, which is a specialty camera for those who have to have FF now at a more reasonable price than the 1Ds Mark II.

Jeff

Edited by jhapeman on Dec 23, 2005 at 02:48 PM GMT



Dec 23, 2005 at 10:36 AM
jmcfadden
Offline
• • • • • • •
Upload & Sell: Off
p.3 #12 · canon 20D and nikon D200 samples


Jeff

I have repeatedly dodged the initial user reports from all cameras from Canon or Nikon. Too many variables skew results too easily and not out of malice in many cases. Norm is a levelheaded guy

here , tell me what you think of these 2 crops from Norm's testing, please tell me which is the canon file and which is the Nikon one.

it should be esay given your apriori , right?

J

http://www.pbase.com/norm/image/53876521/original.jpg



Dec 23, 2005 at 10:43 AM
Sam Bennett
Offline
• • • •
Upload & Sell: Off
p.3 #13 · canon 20D and nikon D200 samples


jmcfadden wrote:
do you have opinion on whether there is always Some sharpening done on Canon cameras even if they are set to 0 and shot in Raw? I remember Askey saying something to that effect but cannot remember the context.


I can't recall Askey saying that. What I do remember him saying is that the Canon algorithm will actually turn down sharpening as the ISO goes up.

In my experience with super-high ISO work, I'd have to say "no", I don't think the Canon's are doing sharpening when sharpening is set to 0. The enhancement you usually see in noise just isn't there.

jmcfadden wrote:
I am also wondering if the D200 is Always doing some NR at higher ISO? Both look pretty good here I think, but the canon files here look sharper but so hard to determine the reason


Well, that's a tough question. The D200 files, even with NR "off" in NC always seem to have less color noise than the Canons. It's hard to know if that's a difference between CCD and CMOS or if Nikon is actually doing CNR on the chip, which seems unlikely. But when you look at the ISO 1600 and 3200 files there sure look like they've been CNR'd - to a point where it makes it impossible to do any further CNR, unfortunately.

As I said in another thread, the one thing the D200 really does provide people at higher ISOs straight out of camera in JPEG files is lower color noise, which indeed does make for a more "film-like" appearance. In my high ISO work, I primarily do CNR on my files to get the same sort of look. Color Noise, IMO is the truly offensive part of digital noise characterstics. So if you want that look right out of the camera, Nikon may be the ticket. Personally, I've done ISO 2400 prints from the 20D at 24"x36" that have only had CNR applied to them - no luminance noise reduction at all - which look utterly fantastic. But I'm only able to do that because of shooting RAW with DPP and the 20D's very fine noise pattern. The D200's noise pattern is simply too chunky to give desirable results at that size - and the same goes for the 20D with converters such as C1 and Raw Shooter.



Dec 23, 2005 at 11:19 AM
jhapeman
Offline
• • • •
Upload & Sell: On
p.3 #14 · canon 20D and nikon D200 samples


jmcfadden wrote:
Jeff

I have repeatedly dodged the initial user reports from all cameras from Canon or Nikon. Too many variables skew results too easily and not out of malice in many cases. Norm is a levelheaded guy

here , tell me what you think of these 2 crops from Norm's testing, please tell me which is the canon file and which is the Nikon one.

it should be esay given your apriori , right?

J

http://www.pbase.com/norm/image/53876521/original.jpg


I avoid them too...but real-world shooting is what this is all about, and Norm's test is pretty good.

This is easy--the Canon is on the right. It shows higher chroma noise, a characteristic of the Canons--but also shows more detail in the shadowed part of the image. Given the pixel pitch of the Nikon, it should show more detail, and indeed it does in the well-lit portions, and in the ISO 100 images. But my points still stand--the Nikon method of noise reduction is killing fine detail in areas of higher noise (e.g., shadows, etc.).

I think when you see the final graphs come out (and there will be many versions, so nothing is really final), the results are going to be interesting...but my prediction is that Nikon is going to show low chroma noise, and higher luminance noise than the Canons. At the end of the day, this translates to more detail on the Canon side.

Astrophotographers have known this for years, and we shoot quad-color Luminance+RGB images for our work, where you have great potential for shadow noise or rather, dark-area noise. By shooting a very long luminance exposure that has low noise and high detail and then layering on much noisier color data, you can get a final image that is superbly detailed and relatively noise free--the overall appearance of IQ is much higher than straight RGB image of the same total exposure time.

The comparison I make is not 100% valid, but it is relevant, and I think we see that here. I have seen tons of D200 images so far, and have downloaded and processed some myself, and I still haven't seen anything that objectively shows me anything different than this: The D200 still can't touch the high-ISO noise of the Canon cameras, and only makes a smooth image by very aggressively smearing out the details in high noise areas like shadows, etc.

Show me some images that don't prove this point and I will be glad to eat humble pie, and maybe even switch back . Like I said, I was a long-time Nikon user and have no preference one way or another--I just need the tool that can do my job--but I continue to see Nikon lack in their ability to handle high-ISO noise.

Jeff


Edited by jhapeman on Dec 23, 2005 at 02:39 PM GMT (Reason: Typo: Thanks John!)



Dec 23, 2005 at 11:50 AM
21farms
Offline
• • •
Upload & Sell: Off
p.3 #15 · canon 20D and nikon D200 samples


jmcfadden wrote:
...I remember Askey saying something to that effect but cannot remember the context.


john, i recall that too but exactly the opposite of the way you remember it in phil's D2X review, he seemed to intimate that the D2X's "zero" sharpening was actually closer to canon's "1" sharpening: http://www.dpreview.com/reviews/nikond2x/page24.asp was that what you were thinking of?

I am also wondering if the D200 is Always doing some NR at higher ISO?

according to the D200 manual, i think the answer is yes based on page 34 where it says:

"Note that although high ISO noise reduction is always in effect at ISO sensitivities over ISO 800, turning High ISO NR on will increase the amount of noise reduction performed."

i had my D200 set with high ISO NR on but set to 'low' instead of the default 'normal'. in my NC conversions, i turned it completely off but i guess there is still some going on at ISO 800+.

jhapeman wrote:
...we shoot quad-color Luminance+RGB images for our work, where you have great potential for shadow noise or rather, dark-area noise. By shooting a very long luminance exposure that has low noise and high detail and then layering on much noisier color data...


jeff, my head is spinning



Dec 23, 2005 at 12:12 PM
AdrianRogers
Offline
• •
Upload & Sell: Off
p.3 #16 · canon 20D and nikon D200 samples


Sarsfield wrote:
Having looked closely at these small samples, I'm glad I didn't decide to switch to Nikon...



Not that i ever intended to, but likewise! It's certainly not the galactic leap some described it to be. I'm sure Nikon users will be happy upgrading from d70s though, which is what matters i guess.



Dec 23, 2005 at 12:29 PM
Tom_W
Offline
• • • • •
Upload & Sell: Off
p.3 #17 · canon 20D and nikon D200 samples


mcarr wrote:
The color differences are interesting as is the texture differences in the grain. The 20D is also noticeably sharper. Is that a function of the lens diference? I don't know anything about the Nikon.


Nikon's noise reduction at high-ISO settings trades away detail for the appearance of low noise. You can see it in the low-key areas of some of the ISO-800 shots in the earlier test.

Also, it seems the the highly-touted 17-55 lens doesn't have any advantage over the 17-40L Canon lens.



Dec 23, 2005 at 12:35 PM
joesmith
Offline
• •
Upload & Sell: Off
p.3 #18 · canon 20D and nikon D200 samples


AdrianRogers wrote:
Not that i ever intended to, but likewise! It's certainly not the galactic leap some described it to be. I'm sure Nikon users will be happy upgrading from d70s though, which is what matters i guess.



For the D100 users it is giant leap.



Dec 23, 2005 at 12:51 PM
Sam Bennett
Offline
• • • •
Upload & Sell: Off
p.3 #19 · canon 20D and nikon D200 samples


10DFT wrote:
Also, it seems the the highly-touted 17-55 lens doesn't have any advantage over the 17-40L Canon lens.


Except that it's 17-55, not 17-40 and it's an f/2.8 and not an f/4. I'd buy a 17-55 f/2.8L. I wouldn't touch a 17-40L with a 15 foot pole.



Dec 23, 2005 at 01:26 PM
jdaily
Offline
• • •
Upload & Sell: Off
p.3 #20 · canon 20D and nikon D200 samples


Jeff, it's a bit hard to follow your arguments sometimes.

Like many, I believed the bunk that Canon's wide offerings were nowhere near as good as Canon's.

and...


...but my prediction is that Nikon is going to show low chroma noise, and higher luminance noise than the Nikons.




Dec 23, 2005 at 01:34 PM
1       2              4       end




FM Forums | Canon Forum | Join Upload & Sell

1       2              4       end
    
 

Welcome back
Log in to your account