I did a quick AI Servo test of my 40D taking sequence of about 20 shots of a duck swimming across a pond. This duck was very small in the frame but every shot was in focus.
I was using the 70-200 f/4L IS. I haven't done anything straight at me yet but the AI Servo seems to work great.
I'm having a hard time buying some of the reviews that I have seen about this camera's AI Servo perfromance. Drew's test looks like he never had the subject in focus at the start of his test and as the subject got closer it was more in focus.
It is a known fact that the sensor points of the 20D/30D/40D series are larger and when the subject is small in the frame the chances of getting the surrounding environment in focus are greater.
It is a shame that perspective buyers are so influenced by opinions of people that do flawed test and create rumors of defective cameras when most likely they don't exist.
I don't expect the AF performance of the 1 series in the 40D, and you shouldn't either, but I can tell you it is better than any other camera in xxD series and a welcomed improvement.
What is not encouraging is that the guy has essentially a plain white shirt while running at you, and you can only guess where the AF point is in the images. Not a whole lot to lock on to, so this should be considered a fairly extreme test, AS THE TESTER POINTED OUT.
Anyway, back to your normally scheduled program....
There's another thread started by Alan321 about this very subject. You might have noticed it. Any how he also provided a series of pictures depicting focus problems. And while his first group were taken of a stationary subject he does later provide picures taken of a couple/baby in a park walking. Again you do see the focus going in and out.
And as for Drew's pics...sure they start off out of focus but they still go in and out of focus as well. As you scroll down the series of photos.
So as for starting rumors...I don't think so. Just casual observations.
It's great that some of you aren't having any AI Servo issues. Perhaps when you have time (J. Victory) you could take a series of shots of a subject coming towards you in AI Servo and post them. Are they all in focus? We'll see.
I mean, if there's a logical explation for Alan321's AI Servo issues. Please explain. Right now he's sent his 40D back to Canon USA to be looked at. Prove my observations wrong. Also, I take most things with a grain of salt.So I haven't totally bought into these claims.
Alan you don't have a position because you don't have a 40D and haven't shot with it.
Your just repeating what others have said without clarifying how they reach their determination. I don't find anything in Alan321's test to be valid or controlled enough to make any educated decision about the AI Servo performance of the 40D.
I'm afraid I have to rely on my past experience and ability to make that determination and I certainly feel no obligation to prove anything to you.
As far as your observations they have been limited to reading the impressions of owners of the 40D and not you actually experience, ie. the title of this thread ( Hands-On EOS 40D Body)
Jim
Sep 19, 2007 at 02:18 PM
Allen Maestas Offline • • • • • • Upload & Sell: Off
Yakim Peled wrote:
Well, I jumped on the wagon and ordered it. Should be getting it next week.
Canon, be ashamed on yourself. You could have got 4500$ from me if you only you'd fix the AF issues in the 1D Mk III. Now you lost more than 3000$. That is no way to run a mega-business.
I would bet that his test is sort of a worst case scenario - the subject is in white against a higher contrast background, and the subject doesn't move much relative to the sensors.
So, think of it this way, the guy was shooting 6.5 FPS. Even at 40% success rate, you are getting on average over 2.6 keepers per second, at worst. Show me a camera with a higher keeper per second rate in the same price range.
I think the people buying this camera will be very happy with 2.6 keepers per second. Heck, I've had entire days without keepers
python2000 wrote:
Regarding the AI-Servo test of the guy running...
So, think of it this way, the guy was shooting 6.5 FPS. Even at 40% success rate, you are getting on average over 2.6 keepers per second, at worst. Show me a camera with a higher keeper per second rate in the same price range.
I think the people buying this camera will be very happy with 2.6 keepers per second. Heck, I've had entire days without keepers
I certainly would have been more pissed off at the 1 series cameras that cost 5X as much and don't have 95% keepers.
python2000 wrote:
Regarding the AI-Servo test of the guy running...
So, think of it this way, the guy was shooting 6.5 FPS. Even at 40% success rate, you are getting on average over 2.6 keepers per second, at worst. Show me a camera with a higher keeper per second rate in the same price range.
I think the people buying this camera will be very happy with 2.6 keepers per second. Heck, I've had entire days without keepers
I certainly would have been more pissed off at the 1 series cameras that cost 5X as much and don't have 95% keepers.
python2000 wrote:
Regarding the AI-Servo test of the guy running...
I would bet that his test is sort of a worst case scenario - the subject is in white against a higher contrast background, and the subject doesn't move much relative to the sensors.
So, think of it this way, the guy was shooting 6.5 FPS. Even at 40% success rate, you are getting on average over 2.6 keepers per second, at worst. Show me a camera with a higher keeper per second rate in the same price range.
I think the people buying this camera will be very happy with 2.6 keepers per second. Heck, I've had entire days without keepers ...Show more →
Hiya python2000,
Good point. I guess getting misses are simply part of owning a DSLR no matter the brand/model. Though with certain cameras more and some less due to the size of the sensors. Or so I've been told.
And as for Jim Victory's comments...I think any one can have a "position". Especially if they are being acused of spreading rumors. Which I think is uncool. And the "position" I was referring to was that I didn't think I was spreading rumors. Lastly, when I asked him to "prove my observations wrong" I meant that in the most friendly way. As I'd love to find out that Alan321's 40D is perectly fine. That it was all "user error". And as for Drew's testing it was an extreme or "worst case" test.
Btw JV my name is not Alan. Any how try and take it easy. We're all friends here. Well most of us.
Edited by Andrew Chase on Sep 19, 2007 at 04:05 PM GMT
Andrew Chase wrote:
I don't think I'm spreading rumors.
There's another thread started by Alan321 about this very subject. You might have noticed it. Any how he also provided a series of pictures depicting focus problems. And while his first group were taken of a stationary subject he does later provide picures taken of a couple/baby in a park walking. Again you do see the focus going in and out.
And as for Drew's pics...sure they start off out of focus but they still go in and out of focus as well. As you scroll down the series of photos.
So as for starting rumors...I don't think so. Just casual observations.
It's great that some of you aren't having any AI Servo issues. Perhaps when you have time (J. Victory) you could take a series of shots of a subject coming towards you in AI Servo and post them. Are they all in focus? We'll see.
I mean, if there's a logical explation for Alan321's AI Servo issues. Please explain. Right now he's sent his 40D back to Canon USA to be looked at. Prove my observations wrong. Also, I take most things with a grain of salt.So I haven't totally bought into these claims.
Just thought I should defend my position. ...Show more →
From what I've seen on the forums and my own use of the camera, I would say that the majority of the cameras focus just fine (ie as expected). I was shooting a MTB event on the weekend and I was shooting in the forest in low light, mid day in fulll sun and everything in between and had no issues at all.
I think people are forgeting that you still need to be able to take pictures. Just because a camera has AI servo focus doesn't mean that you can just push the button and expect the camera to magically do everything for you. Alan mentioned something about using the servo on static objects to avoid misfocus from body sway What did people do when they used manual focus, did the constantly adjust focus based on body sway. I would guess that they either upped the f-stop to bring more of the image in focus, or used a tripod.
Anyways, my report on the camera is that everything is working wonderfully. The menu system is great, the fps are great, the shutter lag is so much improved from the 300d.
I'm sure that all the people with cameras that are working well are out shooting with them and not wasting time doing 'testing' and posting results on the forums.
jctriguy wrote:
From what I've seen on the forums and my own use of the camera, I would say that the majority of the cameras focus just fine (ie as expected). I was shooting a MTB event on the weekend and I was shooting in the forest in low light, mid day in fulll sun and everything in between and had no issues at all.
I think people are forgeting that you still need to be able to take pictures. Just because a camera has AI servo focus doesn't mean that you can just push the button and expect the camera to magically do everything for you. Alan mentioned something about using the servo on static objects to avoid misfocus from body sway What did people do when they used manual focus, did the constantly adjust focus based on body sway. I would guess that they either upped the f-stop to bring more of the image in focus, or used a tripod.
Anyways, my report on the camera is that everything is working wonderfully. The menu system is great, the fps are great, the shutter lag is so much improved from the 300d.
I'm sure that all the people with cameras that are working well are out shooting with them and not wasting time doing 'testing' and posting results on the forums....Show more →
Hello jctriguy,
Thanks for your post. Good to know you're happy with your 40D.
It's looking more and more to me that Alan's woe's could be "user error" and being that Drew's tests were extreme. There's probably nothing to worry about.
As I've said. I've seen a lot of excellent pics taken using the 40D. And looks like a solid camera.
Happy shooting!!
Edited by Andrew Chase on Sep 19, 2007 at 04:21 PM GMT
Edited by Andrew Chase on Sep 20, 2007 at 02:01 PM GMT
I installed the EF-D screen and it's great. The lines are very fine and not distracting at all. I only take note of them when I need them. One less thing to worry about now when I do a landscape or a architecture composition.
As for IQ. Fantastic, and coming from a 20D the perks, bells and whistles are nice.
therock wrote:
I installed the EF-D screen and it's great. The lines are very fine and not distracting at all. I only take note of them when I need them. One less thing to worry about now when I do a landscape or a architecture composition.
As for IQ. Fantastic, and coming from a 20D the perks, bells and whistles are nice.
Would have loved to have seen Canon put out a focusing screen with a "rule of thirds" grid. Being that you can't enable the same "rule of thirds" grid you see when in Live View mdoe.
Edited by Andrew Chase on Sep 20, 2007 at 02:01 PM GMT
Sep 19, 2007 at 04:25 PM
Allen Maestas Offline • • • • • • Upload & Sell: Off
therock wrote:
I installed the EF-D screen and it's great. The lines are very fine and not distracting at all. I only take note of them when I need them. One less thing to worry about now when I do a landscape or a architecture composition.
As for IQ. Fantastic, and coming from a 20D the perks, bells and whistles are nice.
Where did you get that screen? How hard was it to replace?
Where did you get that screen? How hard was it to replace?
Do a search on ebay. The guy is in Hong Kong. I have used him several times. He seems to come up with nice deals and hard to find items. The screen is an official Canon product.
The install is a snap. The tool has a hook on top to flip the hinged mounting frame down. And a clamp to handle removing and installing the screen. A steady hand and a no hurry approach is a key. The install may have taken 5 minutes tops, this includes the reading.
Reading the paper work, be sure to note not touching the screen with fingers.
I just bought a 40D to go with my main camera the 5D. I bought the 40D for use with telephoto lenses (70-200 F4L IS, 300mm F4L IS, & 400mm F5.6L).
I have used the 10D, 20D, & 30D on multiple occasions, and owned an XTI.
This is what I noted about the 40D on the first outing with the 400mm F5.6L lens:
-- improved focusing is nice. I had no problem locking onto birds both big & small, a road runner, and a coyote. The focusing seems more sure of itself when it grabs focus.
-- improved viewfinder is a notable improvement. I wear glasses so this is a welcome change. It is easier to see the entire frame, and the ISO in the viewfinder is a welcome change.
-- high ISO performance. I did not get a chance to test this much. The ISO 400 images were a fine grain that was not distracting to the image. I tested with high ISO noise reduction set on in the camera. The 5D is cleaner at 400 in my opinion.
-- 14 bit color. The images seem to have a more natural color then I remember getting from the 20D. The RAW files seem to have that better color similar to what I get with the 5D.
-- I like that the 40D uses the same battery as the 5D. The BP-511A have excellent battery life. This is especially so with the battery grip with gives you 2 batteries.
-- The BG-E2 battery grip is a necessity with long lenses for balance. I also think the portrait controls make it easier to use as well.
-- LCD screen.-- I had no issue with using the LCD screen to determine focus
Live view -- yawn, whatever, I do not think DSLR's needed live view. If people wanted to frame with the LCD they could buy a point and shoot. But maybe I will find a use someday.
More RAW converter options are available. I tested with Lightroom v1.2, Photoshop CS3 with ACR v4.2, and DPP 3.1 that shipped wth the camera.
Personally I prefer using Lightroom v1.2
The 40D is a keeper camera to go with my 5D keeper camera.
An excellent 2 camera combination.
Edited by abqnmusa on Sep 19, 2007 at 10:58 PM GMT
Edited by abqnmusa on Sep 20, 2007 at 10:07 PM GMT
therock wrote:
I installed the EF-D screen and it's great. The lines are very fine and not distracting at all. I only take note of them when I need them. One less thing to worry about now when I do a landscape or a architecture composition.
As for IQ. Fantastic, and coming from a 20D the perks, bells and whistles are nice.
I'm looking forward to picking up one of these screens. I'm definitely challenged when it comes to shooting the horizon correctly!
Andrew Chase wrote:
Thanks for posting the URL. Not too encouraging. Glad I haven't yet picked up my 40D. Allegedly some 40Ds are fine in AI Servo. SOme not.
While I won't comment on problems with the set up of the test, I will comment on shortcomings in the analysis.
First, the reaction of readers - "Wow! Thank goodness I didn't get a 40D, it sucks!". The 1D2 which costs a couple times more, isn't accused of having a fatal focus flaw and is considered a top tier sports camera is still 44% OOF on the "same" test. Good thing you didn't buy any of the 1D series either. I think it would be unreasonable to expect the 40D to score as well as any 1 series, and it didn't. Who's shocked? Not me.
Second, there's no analysis of the focus error rate compared to subject distance or, alternatively, how large the focus target is in the frame. In how many of those shots is our bouncing jogger test subject presenting a plain white T-shirt to the chosen AF sensor? Put him in a referee shirt and try again. How many of the OOF shots were at far distance (small subject) and how many at close distance (good target)? I know on the L zooms that I've used, as the subject gets closer the focus elements in the lens need to move a lot farther for each incremental change in distance. There is more to nailing focus than just the body and this test only used one lens. At 6.5fps that's a lot of lens movement too - no just camera body. The fact that NO camera (even a 1-series) did very well implies to me that either the test is flawed or the expectations are too high.
Third, there's only a comparison of the 40D against to 1-series, no 30D. Again, the reaction is "wow, glad I didn't order a 40D" with no data on how this body compares to the previous offering in the same class - the 30D. If the 10,20 and 30D score the same or worse I think it's unfair to call the 40D a failure.
Fourth, based on physics, what results are actually reasonable? The camera was shooting at 6.5fps, that's one exposure every 0.153 seconds. The Canon whitepaper on the 40D also lists that the viewfinder blackout time is 0.1 seconds. If we make an assumption, that's 0.05s to raise the mirror and 0.05s to return it. So with a "budget" of 0.15s per exposure, two thirds of that is spend moving the mirror - during which no focus evaluation can be performed and no exposure can be taken. If we ignore the actual exposure time that leaves 0.05s for the AF system to evaluate the scene and instruct the lens to focus. The subject then continues to move during the 0.05s it takes to raise the mirror. Does the projection of movement work properly? How good was the focus target (for 0.05s)? How fast can the lens focus (0.05s budget)? How accurately did it move?
Fifth, what aperture settings was used? If DOF is minimized and the subject is moving, I'm not shocked that a lot of shots were OOF.
Now, if this were a Canon vs. Nikon discussion, then the entire Nikon line isn't represented. If this is a 40D vs. other Canon models discussion I think there's room for improvement in testing, the conclusions drawn, and the expectations of the audience.
Sep 20, 2007 at 07:10 AM
Allen Maestas Offline • • • • • • Upload & Sell: Off
therock wrote:
Do a search on ebay. The guy is in Hong Kong. I have used him several times. He seems to come up with nice deals and hard to find items. The screen is an official Canon product.
The install is a snap. The tool has a hook on top to flip the hinged mounting frame down. And a clamp to handle removing and installing the screen. A steady hand and a no hurry approach is a key. The install may have taken 5 minutes tops, this includes the reading.
Reading the paper work, be sure to note not touching the screen with fingers.
Here! Here! I second the motion! Well put and I hope no-one takes it the wrong way. We do agree to disagree here, right?
I read of, and view images from the 40D here and am thankfull for all the time and efforts of all the members here. Right, wrong or indefferent. Hands on is what this thread is about.
I am very green, a guy with a nice camera I'm having a ball. I have no resource to post a nice slide show for viewing a series of AI Servo shots but am working on it.
I will say that I shot my 9 year old running at me from 100 feet with the 40D and EF-S 17-55 f/2.8 IS, in AV. 2.8, ISO 400 AI Servo. Only the closest one to me, about 12 feet away was OOF. Lucky?
Also, I take most things with a grain of salt.So I haven't totally bought into these claims.
What did people do when they used manual focus, did the constantly adjust focus based on body sway. I would guess that they either upped the f-stop to bring more of the image in focus, or used a tripod.