rscheffler
Offline
• • • • • •
Upload & Sell: Off
|
 p.8 #13 · Galbraith Update: MK III Not Improved | |
(Warning: Long post)
Garylv wrote:
As time goes by a number of people forget how many complaints there were on the various online forums. The new owners are often not even aware of it.
Excellent point Gary. For example about the Mark II:
http://www.sportsshooter.com/message_display.html?tid=14367
http://www.sportsshooter.com/message_display.html?tid=16709
http://www.sportsshooter.com/message_display.html?tid=17478
I'll go out on a limb and state that some of the problems seen in the Mark III are not entirely unique to it and have been present to various degrees in the previous 1D cameras.
If you go back to the SportsShooter forum archive, you'll find there were discussions by some who claimed issues with the 1D II shooting sports and getting nothing tack sharp (see especially the third link above). Was it a problem exactly the same as the III? Perhaps not, but some of the sample photos were similar to the main photo in this SS article about problems with the III: http://www.sportsshooter.com/news/1967 Interestingly Canon released the IIN fairly quickly and of course there were a lot of whispers by 'people in the know' claiming undocumented fixes to the Mark II problems. Kind of makes you go hmmm... As a result I expected a IIIN from Canon by now. But I'm glad they've in their own way acknowledged the III has problems and have worked towards fixing it.
My personal experience with the II was random soft images in action sequences. It was not super reliable and less so in low light situations. But there were specific times, such as shooting superbike races under variable cloudy conditions, where I noticed issues similar to those claimed now of the III. When the sun was shining there was an increase in overall soft images similar to those posted by Gust. When the clouds rolled in, tack sharp sequences...
The first major sports test I put my first III through in June 2007 was a triathlon. I thought it worked well tracking cyclists when shooting with the sun. When I switched to the run segment, I had to shoot backlit and noticed a significant drop in correctly focused images. The plane of focus jumped around and the III's AF system felt extremely hyper. This was a situation quite similar to that used by RG in his tests. It wasn't until the 1.2.3 firmware update that I found the camera to be much more stable, especially for 'normal' static subjects when shooting with faster than f/2.8 lenses.
For all those calling RG's initial findings into question, he was not the only one who saw these problems when shooting sports with action coming directly at the camera.
http://www.sportsshooter.com/message_display.html?tid=25247
Look for my posts. On June 11, 2007 I posted that the III was an improvement over the IIN but not perfect. Until that time I'd shot mostly indoor sports with it, such as ice hockey, and it really did work better in low light than my IINs. I was happy. It was the triathlon though that revealed to me the now debated issues. See my comments on June 17th and also those of Matthias Hangst directly above mine... That's two days before RG published his first review of the problem. And as noted by Paul B above this post, a discussion started on NatureScapes nine days before RG published. (though RG did perhaps plant the seeds of suspicion in his April 2007 Mark III preview)
keithreeder wrote:
Nope, you're not alone Colin.
I had a laugh a couple of days ago when I read a post on another website from someone who reckoned he could replicate RG's Mk III results on his Mk IIn by shooting similarly featureless, low contrast subjects as those RG is shooting in his Mk III tests...
Tedious or not, RG's sports samples and methodology are very representative of both the shooting conditions and the work produced by a lot of sports photographers. The tests needed to be repeatable and as consistent as possible while focusing on the specific AF issue, hence a sprinter running directly at the camera. While 'boring' at least they reproduced key elements that revealed the issue being debated. In real life sports situations it is not unusual to come across low contrast uniforms under less than ideal lighting (i.e. very harsh midday sun creating huge dark shadows, which compound the problem of the low contrast uniforms). And yes, the II or IIN could have similar problems in similar circumstances (based on my experience after 100,000+ images with each of those).
So why wasn't there as much of a stink about the II? What was the alternative back then? Either shoot with a 4MP 1D or use a 4MP Nikon with its own set of problems. Why so much more fuss about the III? Because the AF system was a clear departure from what everyone had been used to using since the EOS-3. The problems were extremely accentuated by the new, much more 'hyper' AF behavior. And the camera produced sharper files than the II, let alone more than 2x the resolution of the original 1D. All the better to reveal focus/sharpness issues. If you're shooting fast action with a 4MP camera, the lack of real resolution results in a greater zone of tolerance if the AF isn't perfect each frame. Having gone back through my original 1D sports archive from time to time, I've noticed a fair number of sequences where the focus was not perfect from frame to frame, but wasn't as obvious as would have been with a 10MP camera. While the II had nearly the same resolution as the III, it produced natively softer files, which could mask slight focus issues.
Early in this thread someone mentioned a comment by a SI photographer claiming he'd like to switch back to Canon but the III's AF remains too inconsistent. Both of my IIIs have had all of the fixes and one also had the mirror box replaced. After all this I will tentatively agree that I still find the III's AF inconsistent. Overall both cameras are better, especially with static subject AF accuracy. So far it looks like I can get by with no microfocus adjustment on either, which wasn't the case prior to the latest fixes. And prior to the mirror box replacement, that camera needed between +10 and +20 for many lenses. Having just shot over 1500 frames yesterday with subjects sometimes walking briskly (outside, bright overcast light), there remained sequences with random soft frames. And I'm still getting inconsistent static subject focus with the 50 f/1.2...
So in summary: Better overall performance than back in 2007, especially with the 1.2.3 firmware and latest AF fix, but still not perfect.
Ron
|
