Still room for improvement on the 1DMkIII

[John_Nevill]

RG went public a couple of hours ago with yet another installment in the ongoing 1DMkIII focus saga and it looks like he's still not overly satisfied with the fix or later production units.

From the 8 tests performed the 1DMkIIN was superior on 3, even on 3 and lost out on 2.

It interesting to note that he recommends a -1 AF micro adjsutment to help improve AF on extended sequences.

He concludes that he has no knowledge on whether Canon are continuing to work on the problem and is uncertain whether the same issues affect the 1DsMkIII.

Canon must love this guy!

 

[Asher Kelman]

RG went public a couple of hours ago with yet another installment in the ongoing 1DMkIII focus saga and it looks like he's still not overly satisfied with the fix or later production units.

From the 8 tests performed the 1DMkIIN was superior on 3, even on 3 and lost out on 2.

It interesting to note that he recommends a -1 AF micro adjsutment to help improve AF on extended sequences.

He concludes that he has no knowledge on whether Canon are continuing to work on the problem and is uncertain whether the same issues affect the 1DsMkIII.

Canon must love this guy!

John,

This is actually pretty good news! We need MFRS feet to be held to the fire and have them know that they can't just create a new txation system for photographers by enticing us to shell out for essentially minor or raz-ma-tazz upgrades. Except for pixel count, Nikon wil hold its own against Canon and also add to the pressure for Canon to take things seriously. I personally feel that Canon has amazing technology sitting on the shelf but until now has felt little pressure and therefore no need to really bother to deliver us the best. Now that Nikon has levelled the playing field somewhat, hopefully Canon will wake up!

Asher

 

[Ken Tanaka]

I have no interest in sports photography and pay little note to those who are interested in this genre. It's just not on my (ever-dwindling) list of interests.

But I am sure that Canon wishes that this problem would just go the hell away. Television images of pro sporting events featuring a blizzard of white lens barrels in the background have established Canon's market leadership more than any other marketing factor. The overwhelming majority of these 1d3 units will land in the hands of amateurs who will never experience AiAF problems that Galbraith rails against. (Many won't even learn how to activate AiAF on the camera.) Nevertheless, the scent of inferiority can gradually gray that white lens sea. And that's something Canon really doesn't want to see.

 

[Asher Kelman]

This has to be encouraging for Pentax, Leica and others who now must see what was obvious a year ago is no longer so clear!

Asher

 

[Ken Tanaka]

Perhaps any apparent weakness in the Canon juggernaut will be encouraging to Pentax and Olympus, Asher. (Leica less so, since the market for their now non-existent R and their M8 is centered mostly in brand nostalgia rather than performance.) But these brands became sidelined largely because they catered to amateur budgets just as ready credit redefined that slice. They're still cheaping it out rather than taking pointers from other brands and products that price up.

No, if there will be any brand to pick up crumbs that fall from Canon's gaping maw it will be almost exclusively Nikon. Pentax and Olympus are not truly competitors to the big dogs.

 

[Steve Saunders]

I think the Nikon D3 has raised the bar considerably. I can't recall ever knowing anyone ditching Canon to go Nikon until now, previously it was all one way traffic. Every day I'm meeting Canon shooters who are now buying into Nikon, the 1DIII fiasco isn't to blame but certainly isn't helping.

Having said all that, I think RG's report puts the MkIII marginally ahead of the MkII in the fast AF stakes if "1DMkIIN was superior on 3, even on 3 and lost out on 2" is right.

 

[Joseph A. Kurkjian]

It interesting to note that he recommends a -1 AF micro adjsutment to help improve AF on extended sequences.

John, interesting is the understatement of the year. I was absolutely blown away by the "-1" statement; it leads me to believe that there is a lot of "work in progress" that will be revealed in the near future (just a WAG on my part). At first it seemed to me the "-1" discovery was similar to a million monkeys typing away to eventually, somehow, type a complete sentence. If the "-1" thing was not a discovery but a S/W switch not encoded into the CFn matrix then I believe Canon is still diligently working the AF problem and "-1" actually changed the AF response (as versus adjusting static AF error).

FWIW I had an opportunity to test a manufactured blue-dot (as versus repaired) 1DmkIII that belongs to a friend. I shot several sequences of a constant velocity car (his car doing 30 MPH) and although the camera was pretty darn good there were still some OOF shots (too many IMHO). Shooting constant velocity cars (at either 25 or 30 MPH) with my 30D, 40D, or 1DmkII yields a 100 percent success rate. I should point out my friends camera had near zero AF error on static objects whether the camera was in One-shot AF or AI Servo AF; I mention this because from my perspective this was an excellent overall camera that had some problems tracking a constant velocity car. A constant velocity object is the most trivial subject for a predictive AF system and the cameras should be 100 percent successful (assuming the shooter aims the camera correctly).

Did you eventually buy a 1DmkIII or are you still waiting in the bushes like me?

Regards,

Joe Kurkjian

 

[John_Nevill]

Joe, I'm not sure what to get next.

I have used two 1DMk3s. One for long term and they both had issues. I'm not convinced that Canon have totally solved the problems yet.

Seeing as I use the 1DN mainly for bird in flight stuff and the 5D for higher res stuff, I'm happy to stay with them, although I may consider picking up an end of stock 1Ds2 early next year.

 

[Philip Chong]

I have found no problem with the EOS-1D Mark III's AF system in tracking moving subjects. And this image is one of the thousands I had shot with a first shipment (May 25, 2007) unit. Speed of the bike, rode by American Nicky Hayden, was above 160 mph. Shooting data: 1/400 & 1/500 at f/8, Av mode, ISO 200 w/EF 500mm f/4L IS USM

 

[Bart_van_der_Wolf]

I have found no problem with the EOS-1D Mark III's AF system in tracking moving subjects.
...
Shooting data: 1/400 & 1/500 at f/8, Av mode, ISO 200 w/EF 500mm f/4L IS USM[/B]

They look fine, well done. Maybe the f/8 also helped a bit to mask the inevitable small variations. Is there a specific reason why you chose f/8? I looks to me that a slightly more blurred background could have helped the images. Did you shoot on monopod, a gimbal mount, or otherwise?

Bart

 

[Philip Chong]

1/500 is the maximum I would use to capture bike images. Any faster and the panning effect would have been lost, and although slower shutter speeds would produced better motion/panning effect, it would require even smaller minimum f-stops to be used = 1/250 at f/11, 1/125 at f/16 and so on ...

1 or 2 stops down from maximum aperture would provide adequate depth-of-field to cover from front to rear of bike in most cases. If shot at maximum of f/4 or f/5.6 (if using 300mm w/2X Extender), the zone of focus would be just enough to keep the rider in sharp and everything else to be soft. Here's a link to an f/5.6 image shot via 300mm lens w/2X Extender - http://www.velocity-media.net/philip/MGP069.jpg

The above sequence and the linked larger image were all shot handheld, with the IS feature on lens set to 2 for panning mode. Had tried shooting via a monopod previosuly but didn't quite like it.

 

[Joseph A. Kurkjian]

1/500 is the maximum I would use to capture bike images. Any faster and the panning effect would have been lost, and although slower shutter speeds would produced better motion/panning effect, it would require even smaller minimum f-stops to be used = 1/250 at f/11, 1/125 at f/16 and so on ...

1 or 2 stops down from maximum aperture would provide adequate depth-of-field to cover from front to rear of bike in most cases. If shot at maximum of f/4 or f/5.6 (if using 300mm w/2X Extender), the zone of focus would be just enough to keep the rider in sharp and everything else to be soft. Here's a link to an f/5.6 image shot via 300mm lens w/2X Extender - http://www.velocity-media.net/philip/MGP069.jpg

The above sequence and the linked larger image were all shot handheld, with the IS feature on lens set to 2 for panning mode. Had tried shooting via a monopod previosuly but didn't quite like it.

Your pictures are very nice pictures but I think the real point Bart was trying to make is "if" you post pictures to prove/demonstrate the AF on your camera is working properly in AI Servo AF mode then the lens should be at maximum aperture (and IMO 100 percent crops should be included with the big picture shots). If on the other hand you want to show some nice action shots, then mission accomplished. BTW, I love those types of bike shots, you did great!

Regards,

Joe Kurkjian

 

[Philip Chong]

Your pictures are very nice pictures but I think the real point Bart was trying to make is "if" you post pictures to prove/demonstrate the AF on your camera is working properly in AI Servo AF mode then the lens should be at maximum aperture (and IMO 100 percent crops should be included with the big picture shots). If on the other hand you want to show some nice action shots, then mission accomplished. BTW, I love those types of bike shots, you did great!

Thanks for your comments. However, in actual assignment situations, the number of in-focus images such as the ones I have shown here are far more important than relying on 100% magnification to show whether the images are really in focus or otherwise.

Looking at 100% crop from the PC monitor is not the right way to view actual photographs - what may look soft at this magnification won't translate into a soft image in a magazine double-spread or even an A2-size print being churned out from a Canon ImagePrograf 5100 or Epson Stylus 4800.

Also, I found RG's tests on 1DIII to be somewhat flawed despite the presence of Canon USA staff. In some of his tests, out-of-focus images were taken from a non-active AF point rather than on the area where the AF point was active. I certainly won't use the camera's central AF point tracking the chest and later based my overall opinion from a soft looking image of the face, nose or lip area. IMO, the fact that the 1DIIN may be sharper at those areas could be due to slight front- or back-focus rather than precise focusing.

While I may not shoot non-motorsports images, I do keep myself updated on other sporting events and so far, I haven't come across comments from photographers at Sports Illustrated magazine having problems with the 1DIII's AF - most of the claims of poor AF came from other online forums similar to OPF.

 

[Joseph A. Kurkjian]

Thanks for your comments. However, in actual assignment situations, the number of in-focus images such as the ones I have shown here are far more important than relying on 100% magnification to show whether the images are really in focus or otherwise.

Of course I agree "if" your "stated subject" was actual assignment performance, but it wasn't. Think about this for one more minute, if you had said your camera had great noise performance at ISO-3200 it would not have served you well to drop down two stops in ISO and show examples at ISO-800, right? That's all I was trying to say regarding shooting wide open instead of stopping down by two stops to demonstrate AI Servo AF tracking accuracy, nothing more was implied regarding your personal skills (which obviously are very good) or your camera AF performance (which I can't comment on based on your examples).

Also, I found RG's tests on 1DIII to be somewhat flawed despite the presence of Canon USA staff. In some of his tests, out-of-focus images were taken from a non-active AF point rather than on the area where the AF point was active. I certainly won't use the camera's central AF point tracking the chest and later based my overall opinion from a soft looking image of the face, nose or lip area. IMO, the fact that the 1DIIN may be sharper at those areas could be due to slight front- or back-focus rather than precise focusing.

Yup, I couldn't agree with you more. IMO he should have used constant velocity cars (a simple target that the AF system should track with 100 percent razor sharp accuracy) instead of runners (a very complicated target with many complex motions that cloud the evaluation of OOF issues).

While I may not shoot non-motorsports images, I do keep myself updated on other sporting events and so far, I haven't come across comments from photographers at Sports Illustrated magazine having problems with the 1DIII's AF - most of the claims of poor AF came from other online forums similar to OPF.

I personally find it interesting (and disturbing) that unlike other 1-series cameras (e.g. 1dmkIIN) SI has not published a guide to their photographers regarding how to set up the 1DmkIII for sports. It's been nine months and still no "how to" guide; maybe the guys at SI can't figure out "how to" in a consistent manner, no longer have the time and/or interest to publish a "how to" guide, or all in a sudden have lost their photography skills and are incapable of advising others?

BTW, as a side note and just out of curiosity, what is your opinion regarding sportsshooters.com and the contributors there?

Regards,

Joe Kurkjian

 

[John Harper]

EOS 1DMKIII - versus Birds in Flight

Hi There

I thought i would add my take on the 1DMKIII debate and the accuracy of its AI servo after the fix. I was able to have a play with a sub mirror fixed and Firmware 1.3 version MK3 at my usual haunt of the Hawk Conservancy last weekend.

The weather conditions were far from ideal but i post below a couple of pictures taken at the end of the day in what might best be described as available dark.

I set the ISO at 2000 for the Great Grey Owl Shot and 3200 for the Saker Falcon. Now i appreciate this is not a scientific test, just my opinion against my experience using a MKIIN for just over a year.

In low light I got a higher percentage of shots in focus on the eye with the MKIII than i have been able to with my IIN. It seems to snap into focus a lot faster than the IIN.

I will have to wait for better weather for the sunlight test but certainly in low light i feel it is better than the IIN and the lower noise at high ISO make it a winner as far as i am concerned.

John

Shot 1

EOS 1DMKIII
EF 500mm F4L
ISO 2000
1/640 @ F4

Shot 2

EOS 1DMKIII
EF 500mm F4L
ISO 3200
1/800 @ F4

 

[Philip Chong]

Of course I agree "if" your "stated subject" was actual assignment performance, but it wasn't.

Thanks for your comments, Joe but those images were part of my actual assignment, and not a test to simply gauge the camera's AF performance. I just happened to be using this sequence of images to prove my point regarding the AF issue of 1DIII.

 

[Joseph A. Kurkjian]

To John Harper:

John, wow, I have to agree with you. Looks like you have a definite 1DmkIII winner in your bag, congratulations. I hope you come back here and report on your "high illumination" results. I'm quite impressed with what you have accomplished so far as this is the best post AF fix performance I've seen to date.

Regards,

Joe Kurkjian

 

[John Harper]

To John Harper:

John, wow, I have to agree with you. Looks like you have a definite 1DmkIII winner in your bag, congratulations. I hope you come back here and report on your "high illumination" results. I'm quite impressed with what you have accomplished so far as this is the best post AF fix performance I've seen to date.

Regards,

Joe Kurkjian

Hi Joe

Well as soon as i get the chance when the weather improves over here i will let you know. My main interest in the MKIII is the better high ISO performance, to a certain extent i would trade a slightly worse AF performance for cleaner images.

Birds in flight are about as hard a target as it gets for the AF as their movements are often erratic and they can be quite a low contrast subject, so its pushing the limits on any camera. But i have to say i was pleasantly surprised at the performance that the camera turned in

John

 

[Philip Chong]

BTW, as a side note and just out of curiosity, what is your opinion regarding sportsshooters.com and the contributors there?

I sort of missed out on this part of your previous reply to me.

To be honest, I am not a regular visitor to that site and my previous visits there were limited to curiosity as I have two fellow photographers from my country who have been accepted as members there. One of them, Marcus Yam, told me a few things of what real American pro photogs do (both sportshooters members and non-members) that he came across while covering the same assignments. These were what he had told me:

1. Almost all of them relied on either DPP or Capture NX to edit their RAW images. Very few use 3rd party make RAW converters.

2. AF issue in 1D Mark III? Never heard of it by the majority of them.

3. Rob Galbraith? Err ... who the heck is this guy???

4. Many existing 1D-series shooters prefer the two-button style of Mark II than the one-hand version on Mark III

5. Despite the fact that the higher speed 1D Mark II N and 1D Mark III are aimed at the sports and news photographers, an increasing numbers of them are turning to the full-frame 5D and 1Ds Mark II as their cameras of choice (this was before both the EOS-1Ds III and Nikon D3 started shipping).

6. The larger 3.0-inch LCD nor its 230,000 pixels resolution in the Mark III didn't make any difference to most of them since "chimping" is not something they do most of the time.

How many of the above are the truth? Well, I have no idea. These are basically what have been told to me and I would say I believe him since he was also my former protege. He is currently in the USA and had been interned for Buffalo News as well as the Associated Press.

 

[Philip Chong]

Former 2006 MotoGP World Champion Nicky Hayden posted the fastest ever lap at the Sepang circuit for an 800cc bike
with a time of 2min 0.326 secs astride a Honda RC212V fitted with a 2007 engine and attached to a 2008 chassis

EXIF Data: EOS-1D Mark III, EF 500mm f/4L IS USM w/1.4x EF Extender II, 1/500 sec. at f/5.6, ISO 200. Handheld shooting.