Rob Galbraith's comments on 40D

[suhaskulkarni]

Now apart from the 1D MK III and iDs MK III he has turned his eyes on poor 40D too....

 

http://www.robgalbraith.com/bins/multi_page.asp?cid=7-8740-9068-9537

 

Q. Would switching to the EOS 40D for sports also be a decent workaround, perhaps until Canon has the autofocus

problems in their 1-series line completely sorted out?

 

No.

 

And looking at his comentry he has made some bold statements like

With an EF 300mm f/2.8L IS or EF 500mm f/4L IS lens attached, the EOS 40D produces an unacceptably low percentage

of properly focused or even usably focused frames of track events, soccer, rugby, football and basketball in our

testing. Even as a secondary body - for instance, acting as the short telephoto camera when the action comes too

close for a long lens - the camera's AF falls short. At best, we've seen it get about half the photos in a

sequence in focus, while for grab-and-shoot photography it has missed the focus on more frames than it has gotten

right.

 

 

Is this correct? Do people having 40D agree to this?

Now, I have 350D and while I have not done any analysis of the auto focus, I think I can get 50% of shots in

focus with 100mm USM macro lens for action shots. I usually feel that it's me and not the camera who is

responsible for the shots those were out of focus.

If what Rob says is correct then I can now blame my camera because if 40D is so bad then I can't expect 350D any

better in auto focusing... !

[PuppyDigs]

I can't speak for Robbie's experience, but I must say my 40D AF is pretty darn tootin' and rarely misses. Certainly is better

than my Elan 7NE, 10D, 20D and my manual focusing skill.

[bobatkins]

Maybe it's his testing procedure that's at fault rather than the cameras. Nothing is perfect and expecting perfection makes you doomed to be disappointed.

 

Has he ever tested a camera that he thought the AF actually worked on? What would be good enough for him? 50% in focus? 75%? 90%, 95%, 100%? At what point does AF tracking work as well as can be reasonably expected? Surely nobody expects perfect focus every time on a subject with semi-random movement. What about shots of something moving a constant or smoothly changing speed, like F1 race cars. Someone running is wobbling about all over the place making it difficult for tracking focus to predict exactly where they will be when the shutter opens.

 

My 40D seems fine, though I don't use the AF servo mode very often. In a few tests I thought the 5D did a slighly better job at AF tracking, but that could be because of the additional AF assist points in the 5D. I've never set the cameras on a tripod and shot subjects coming directly at me at a constant speed. Those aren't conditions that I actually shoot under.

[pjmeade]

Although 10fps would be nice, I find the 40D makes a very acceptable sports camera, certainly for cricket and polo. And as Puppy says, it's focus characteristics are better than those of the 20D.

 

P

[steve_eichorn]

"What about shots of something moving a constant or smoothly changing speed, like F1 race cars."

 

Bob, hey, I have a new 40D, maybe you could get me a press pass on the next F1 event. I promise I'll get as many shots for a focus test as anybody could want! {grin}

All kidding aside, for a test on auto focus, I couldn't think of a better subject than that.

0-60 around 2 seconds. By the time I could get the camera to pan, they would be gone!

BTW, I love this site, there is a ton of good, solid information here.

Thanks, everybody, Steve

[NK Guy]

> Has he ever tested a camera that he thought the AF actually worked on?

 

If I recall correctly, he's most satisfied with the AF performance of the 1D mark IIN.

[louie]

He certainly liked the 1D Mark II N:

 

"The EOS-1D Mark II N has the finest autofocus system we've ever used. It's not perfect: it can be slow to focus in dim light and its predictive tracking algorithm is easily tripped up by a subject that's continually speeding up and slowing down. All in all, however, we don't know of another camera that can match its level of autofocus performance. Well, except for the EOS-1D Mark II and EOS-1Ds Mark II, which feature the same system (with minor algorithm changes in the latter model to account for its slower maximum frame rate)."

 

As for testing procedure, it looks like it's mainly his experience in real world conditions of sports photography. The linked article is quite detailed about how he has come to his conclusions, and I find it very persuasive.

[richard_martin10]

I don't know who this guy is but he is way off the mark with his statement on the 40D. I was an XTi user for about 2

years and was pleased with that camera in all ways including AF. I switched to a 40D just recently for it's

ergonomics and the rebate. I expected similar AF results to my XTi but the 40D is way superior, and I use it mostly

for fast action auto racing. My keeper rate with the XTi was in the 50-60% range, the 40D is more like 80-90%. I

couldn't be happier with my new purchase. I've included an example photo from a race a couple of weeks ago, the

car is moving at about 140 mph.<div></div>

[eric merrill]

I'm happy with the AF results of a 40D + 300/4 IS. Even the 40D + 300/4 + 1.4x isn't bad.

 

Are all shots in a sequence using AI servo in focus? No. Nor do I expect them to be.

 

Its autofocus is good enough that I don't flinch about using it professionally.

 

 

Eric

[milton-chris]

No complaints with my 40D on AI-Servo mode. I find the only time that focus becomes even slightly a challenge is when I am shooting through the fence of a municipal basebal diamond, and then, what can one expect. That is certainly not the fault of the camera - it is doing what it was designed to do.

 

RG's comments against the 40D for bird photography might also be disputed by someone who makes thier living shooting birds - oh, maybe, /arthur Morris for example, who has been using the 40D for several months and raving about how it is an excellent camera for BIF and birds in general.

[jay a. frew]

I am very happy with my 40D. I have shot my dog, running full-tilt, both towards and across the lens' field of

view, with very good results. However, I have no deadlines and I haven't done that type of shot 14,000 times.

 

When I read Mr Galbraith's comments, I considered the frame of reference under which he conducted his tests. He

is a busy pro who shoots many more frames (orders of magnitude) in a short period of time than I ever will. His

success-rate with a photo system, and the availability of special tools that function as designed, affect his

lively hood and reputation - directly.

 

I also considered the fact that (I might be wrong here) Canon did not engineer the 40D as a "pro sports shooter"

platform. If I am shooting my dog and I miss a few frames here and there, well, I just get to shoot some more -

and I don't have to force my dog to run ;~))

 

When I bought the 40D, I did not delude myself into believing that there were no limits to its capability...I

just knew it was more than good enough for what I wanted to do...and...thanks to Mr Galbraith, I now know a

little bit more about the limits of this great camera without having to push thousands of frames through it.

 

Cheers! Jay

[canfred]

I must say never having used auto focus SLR's before my 40D has been an eye opener.

I rarely see it miss enogh to refocus. I don't shoot sport and looking at his tests make me think

just how much more could be achieved I would be happy with the results would you do better with manual

focus?

[david_amberson1]

I have the 40D and a 1D3 and have never been happier. My 40D is excellent in AF tracking. In fact I only upgraded to the 1D3 for the increditable low light focus speed and high ISO performance. I think everyone should should quit reading this BS and use their cameras and only question it if it continually fails. I believe that this leads people to try and find fault based on what they are reading where they never found fault before.

 

Folks, take it for what its worth. This is big business(reviewing photography equipment). No doubt it had faults before, thank him for helping get it addressed, but its done. Its time to move on.

[ryepixels]

Hmmm. 40D shooting F1 race cars. Check out my gallery from this year's <a href="http://www.ryepixels.com/other-photos/20080315-GP/1_Sat_AM_Qualifying/index.html">Melbourne F1 here</a>.

 

That was an early 40D (Aug '07) with 24-105 f4L hand held. Yup, not all my shots were perfect, but enough goodies for my personal enjoyment.

[pgranone]

We should all ignore Rob's BS. He appears to have a flawed technique or flawed testing method. My 40D does not have the focusing problems he suggests. Robs method of testing is where the failure is.

[bobatkins]

I think he's simply using the most difficult possible test method, but that's what he shoots. Runing/jumping people

don't maintain a constant velocity. They wobble about a bit. It would be really nice if the AF was smart enough top

know with 100% certainty which way and how fast they might wobble when shooting at 10 frames/sec, but it's not.

 

Predictive AF works by calculating how fast the subject is moving, then calculating where it will be when the shutter

opens and moving the AF to that point. The real test of that ability would be to shoot an approaching car driven at a

constant speed. If it constantly gets focus wrong, there's clearly a design issue. If it hits focus fairly consistantly,

then there isn't. If the 1D MkIII AF tracking is perfect for autoracing but less reliable for athletes, then the problem

may be as much with the subject as the camera. Maybe the algorithms just can't cope with athlete wobble 100% of

the time.. That's not a design flaw any more than the inability to shoot at 30fps or at ISO 400000 is a design flaw.

Both would be nice, but they're not currently possible.

 

It seems to have gotten to a point where he can't let go and his fans don't want him to let go. I'd assume he's not

making Canon very happy at this point either. The vast (and silent) majority seem to be OK with the AF on the 1D

MkIII (and 40D). It's not perfect, they don't expect it to be perfect and it could, I guess be better. However again for

the vast majority of users, it's more than good enough.

 

The question now is whether there's an actual design flaw, of whether the system is working as it should, but simply

isn't as close to perfect at RG would like. Where do you draw the line? What percentage of perfectly focused shots

is "acceptable". If they were all out of focus, then there's a problem, but if 50% of the shots are in focus under

conditions where the probablity is you'd be lucky to get any in focus with manual operation, then is that reasonable

performance.? The other question is whether anyone else does better. The only tests I've seen between the 1D MkIII

abd Nikon D3 showed that the 1D MkIII got a somewhat higher percentage of shots in focus than the Nikon did.

[michael_dimarzio]

Sam S-Great photos, that looks like it was a fun day.

[Spearhead]

Nice summary Bob, too bad Galbraith, and the people here who insist on linking every time he posts something new, don't recognize the same issues.

[neil_everitt]

I am beginning to suspect that RG has got himself down a blind alley with this one. I put off buying a Mklll until about six

months ago because of the adverse criticism of this camera. Since buying one I can tell you unequivocally that the hit rate

of my Mklll in predictive AF beats my former Mkll hands down.

 

So would I be put off by RG's comments on the 40D? No!

[david_amberson1]

I think Bob hit the nail on the head. Galbraith seems pretty happy with the Mark IIn. Well, take what Bob said.

Maybe the Mark IIn deals with "His" kind of shooting better. But most will say that it doesnt do so well where the

new 1D3 shines. Ex: Low light focus speed, aquisition, accuracy, High ISO performance, FPS, IQ etc.

 

Maybe Canon said, "Most people who shoot "Galbraiths" type sports are happy with a 60 -75% hit rate, but would

much rather have better, low light, initial focus aquisition/accuracy much more".

 

And thats where the 1D3 shines. Sounds like you cant have the best of both worlds. Maybe they figure that the

1D3's new features are more important than the style Galbraith shoots. But not to him, because this is what he

shoots the most. Much more important to a man that derives 75% of his income of this particular style. And as you

can see, he still uses it, because he too knows there isnt a better choice. And

certainly going back in tech. isnt worth the down grade in IQ. For me, it works great! Both my 1D3 and the 40D.

 

Bob, good job. I couldn't have sumed it up better.

[rhaytana__tim_adams_]

My two cents, for whatever they're worth from a 5D owner:

 

When I read a review, I ask myself a few questions:

 

(1) Is the reviewer a good photographer?

 

In browsing Mr. Galbraith's site, the answer seems clear: he's a superb photographer. A top professional, one of the best in the business.

 

(2) How is he/she regarded by other knowledgeable photographers?

 

If you visit the forums at SportShooter, which caters to professional sports photographers, the answer becomes clear -- he's held in exceptionally high regard. Consider the posts that refer to him by SportShooter's Robert Hanashiro:

 

http://www.sportsshooter.com/message_display.html?tid=30240

 

(3) Do the tests/trials of the equipment seem fair and thorough?

 

To this reader, they seem exceptionally fair and thorough.

 

In short: I'm inclined to put a lot of faith in Galbraith's opinion. However, I agree that most users -- who don't depend on their cameras for income -- may not be bothered by the issues that he highlights in his reviews.

[PuppyDigs]

"In short: I'm inclined to put a lot of faith in Galbraith's opinion. However, I agree that most users -- who don't depend on

their cameras for income -- may not be bothered by the issues that he highlights in his reviews."

 

Or those that shoot weddings, portraits, airshows, skaters, motocross, landscape, travel, macro, birds in flight, etc. There

is much more to photography than AF on track events...

[kevin cassar]

People should stop taking every negative comment as if it was intended to put down their brand/equipment. Now I am by no means an expert or a professional photographer but I shoot in very difficult low light conditions like concerts, gigs and sports. I understand exactly what he means! I shoot a Nikon D200 and I know that because of the "limitations" of the camera I sometimes have to anticipate the shot or use the centre focus sensor and recompose to get the shot. If I had a D3 or D300, I probably would not need to do this because of Superior AF sensors. Having said that it does not mean that I don't get the shots I want but that if I spent more money on a better body, I'd be able to do it more easily. The Canon 40D vs the 1D mk3 is probably in the same line of reasoning. Just my opinion. [link removed]

[hawkman]

Suhas,

I disagree with Rob G. please see a similar thread below, AF in BIF, if not more, is as challanging as sports.

 

http://forums.dpreview.com/forums/read.asp?forum=1019&message=28839295

 

 

Regards,

Arash

[victor_kunkel2]

Could just be me, but I don't have the impression that the author took much time to learn the camera. He might be a pro, but my opinion is that to provide a review, or a partial review, one must learn more than just a few basics. For example, I recently did long bursts in AI Servo with my 40D and 17 - 85 IS, and had much better results that his photographs seem to show. I did two such bursts; first a bride and groom walking quickly down the isle just following their service; and, second, children playing frenetically at the same event. The couple elected to keep approx. 75% of the photographs in my above examples preferring to choose what they were going to put into their LCD album. But they were very sharp shots. I am happy with the 40D. Surely there are better cameras out there, but this is well worth the money and, crop or no crop, it yields excellent reslts for me.