Trying to shoot small, fast flying birds with the sea and rocks in the background is a completely frustrating and pointless exercise with the Canon 1D Mk3. The camera is simply not capable of aquiring objects that are small in the frame, so the tracking isn't even an issue. All of my successful flight shots were made with the 1D2N.
I cannot see how this can be fixed with a firmware upgrade unless it is possible to re-map AF-point coverage and/or change the parity/priority between the individual sensors in the AF-sensor array.
For objects that are large in the frame, and for backgrounds without much detail/contrast, the Mk3 AF-system seems to work about as well as the Mk2, so in my opinion, the re-designed AF system in this camera has been completely over-hyped by the Canon marketing people and this has lead to expectations that are not matched by performance.
The high ISO performance and 14-bit colour, however, are real improvements, as I have noted in earlier posts, and since I was shooting most of the time at ISO 800 or above, these features were of significant benefit on this trip.
So if you are thinking of buying a Mk3, my advice is to hang on to your Mk2, rather than trade it in if you think you are going to be doing any flight shooting.
Also see the Note comment at the foot of this article. Chuck Westfall at Canon USA does seem to acknowledge a wider issue and more improvements to come by the looks of it.
Hi Phil, your experience with the MarkIII completely parallels mine. The camera is pretty much useless when it comes to pick up birds flying with a different background than the sky. In particular the sea seems to give the AF huge problems. I agree that it is not a question of tracking but initial AF. In Northern Norway I couldn't even pick up a flying white-tailed eagle with the sea as BG (the eagle probably was 5%-10% of the frame).
I haven't tried a MarkIIN, but my old trusted 20D would sometimes be able to pick up what the MarkIII can't, allthough it couldn't keep the focus.
In must other aspects the camera is really nice, but it is not a superb action camera!
Can you post, in this thread, an example that indicates the size of bird in the frame at the point you feel that it fails to perform.
Tom, I'm not sure how to respond to this. The only way I know how to photograph birds in flight is to:
1. select a bird that looks like it has a trajectory that will bring it close to my position when it is still way off and small (size of the central sensor) in the frame;
2. track it into the "window" that I have pre-visualized and where it is now large in the frame.
I don't shoot until the bird is in the pre-visualized position, either still coming onto me or parallel to my shooting position. I don't have any photographs taken when I'm trying to initially acquire the target when it is small in the frame, but in these circumstances and with the sea or other detail (contrast) in the background, the Mk3 just will not acquire the target. It acquires focus on the background and remains focused there.
Mike said:
Chuck Westfall at Canon USA does seem to acknowledge a wider issue and more improvements to come by the looks of it.
Mike, that isn't how I read what Westfall is saying. I take his meaning to be that Canon feel that the Mk3 is working to spec and that the failures are due to user error and that will be addressed in the up-dated White Paper that he refers to: in other words "RTFM".
So as I say above, shooting with the 1D2N, I was able to get sharp pictures of puffins and murres in flight with the sea and rocks in the background, which I was not able to do with the 1D3.
I will not be getting rid of the Mk3 because of the superiority of the files that it produces in terms of tonal qualities and low noise, and because I am not reliant on shooting birds in flight for my bread and butter. At an early point in the debate about the Mk3 AF-system I stated that I thought the new AF system was clearly different from that in the Mk2 cameras: it certainly is. I also said that we may have to accept it for what it is and re-learn how to use it most effectively for our particular applications. For birds in flight with challenging background conditions I do not think there will be a quick fix and we may need to accept this limitation until the "Mk3N" is rolled out. In the meantime, the Mk2 will still perform and produce this type of image as it has in the past.
I should explained further but your "size of the central sensor" remark may have answered it.
I was simply looking for a photograph, any photograph, showing a size of a bird (in focus or otherwise) relative to the size of the frame were you feel the focus will fail if the relative size of the bird gets any smaller.
I was asking as I know what size the subject has to be in a 1DS2 viewfinder before I can confidently expect the central point focus to achieve focus on that bird; I wanted to compare each.
Posted by Mike Smith on August 7, 2007 at 12:34 PM
Phil, perhaps it was my interpretation but I I read the article as being 3 issues - (i) incomplete firmware fix, (ii) "situation concerning this camera's autofocus performance is still evolving as of this writing" (ie (i) hasn't worked) and (iii) arising from feedback of issues encountered a better technical guide on how to programme the complex AF system
I am sure that a fix will evolve eventually, there is too much Pro Sports business potentially tied up in this body and Canon will not want pro sports shooters returning to Nikon for a more reliable AF system
Discussion
6 repliesI cannot see how this can be fixed with a firmware upgrade unless it is possible to re-map AF-point coverage and/or change the parity/priority between the individual sensors in the AF-sensor array.
For objects that are large in the frame, and for backgrounds without much detail/contrast, the Mk3 AF-system seems to work about as well as the Mk2, so in my opinion, the re-designed AF system in this camera has been completely over-hyped by the Canon marketing people and this has lead to expectations that are not matched by performance.
The high ISO performance and 14-bit colour, however, are real improvements, as I have noted in earlier posts, and since I was shooting most of the time at ISO 800 or above, these features were of significant benefit on this trip.
So if you are thinking of buying a Mk3, my advice is to hang on to your Mk2, rather than trade it in if you think you are going to be doing any flight shooting.
It would be much appreciated.
T
I haven't tried a MarkIIN, but my old trusted 20D would sometimes be able to pick up what the MarkIII can't, allthough it couldn't keep the focus.
In must other aspects the camera is really nice, but it is not a superb action camera!
cheers, Helge
Tom, I'm not sure how to respond to this. The only way I know how to photograph birds in flight is to:
1. select a bird that looks like it has a trajectory that will bring it close to my position when it is still way off and small (size of the central sensor) in the frame;
2. track it into the "window" that I have pre-visualized and where it is now large in the frame.
I don't shoot until the bird is in the pre-visualized position, either still coming onto me or parallel to my shooting position. I don't have any photographs taken when I'm trying to initially acquire the target when it is small in the frame, but in these circumstances and with the sea or other detail (contrast) in the background, the Mk3 just will not acquire the target. It acquires focus on the background and remains focused there.
Mike said:
Mike, that isn't how I read what Westfall is saying. I take his meaning to be that Canon feel that the Mk3 is working to spec and that the failures are due to user error and that will be addressed in the up-dated White Paper that he refers to: in other words "RTFM".
So as I say above, shooting with the 1D2N, I was able to get sharp pictures of puffins and murres in flight with the sea and rocks in the background, which I was not able to do with the 1D3.
I will not be getting rid of the Mk3 because of the superiority of the files that it produces in terms of tonal qualities and low noise, and because I am not reliant on shooting birds in flight for my bread and butter. At an early point in the debate about the Mk3 AF-system I stated that I thought the new AF system was clearly different from that in the Mk2 cameras: it certainly is. I also said that we may have to accept it for what it is and re-learn how to use it most effectively for our particular applications. For birds in flight with challenging background conditions I do not think there will be a quick fix and we may need to accept this limitation until the "Mk3N" is rolled out. In the meantime, the Mk2 will still perform and produce this type of image as it has in the past.
I should explained further but your "size of the central sensor" remark may have answered it.
I was simply looking for a photograph, any photograph, showing a size of a bird (in focus or otherwise) relative to the size of the frame were you feel the focus will fail if the relative size of the bird gets any smaller.
I was asking as I know what size the subject has to be in a 1DS2 viewfinder before I can confidently expect the central point focus to achieve focus on that bird; I wanted to compare each.
I am sure that a fix will evolve eventually, there is too much Pro Sports business potentially tied up in this body and Canon will not want pro sports shooters returning to Nikon for a more reliable AF system