Galbraith is finally done testing Canon's AF

[Khun_K]

I don't believe and don't think there is a perfect camera and I don't rely on a perfect camera to get work done.  The accuracy we are talking about today is much higher than those from 10 years ago when we use magnifier on film and thought most of them were perfectly focused. The degree of out of focus is increased when sensor resolution becomes higher, as the depth of field continue to shrink when ten of million pixels continue to add to a sensor. I do believe most of the camera makers, Canon included, made improvement on their system and mechanism, they don't make it worse. But then I do not think there is enough reason to allow maker to make less than perfect cameras, they should all improve endlessly.  
However, I also believe we should work with our camera like partners, learn to use them, learn how they behave and react/control their behavior, make the most out of it.  When a camera can deliver 90-95% of its function will be good enough for me, not that I can accept less, but I can accept reality. Just like when my digital back cannot get good enough quality at the highest ISO setting, I can go to use 1Ds MK2/MK3 or some may go to Nikon. But this is just me.

[Mark D Segal]

I don't believe and don't think there is a perfect camera and I don't rely on a perfect camera to get work done.  The accuracy we are talking about today is much higher than those from 10 years ago when we use magnifier on film and thought most of them were perfectly focused. The degree of out of focus is increased when sensor resolution becomes higher, as the depth of field continue to shrink when ten of million pixels continue to add to a sensor. I do believe most of the camera makers, Canon included, made improvement on their system and mechanism, they don't make it worse. But then I do not think there is enough reason to allow maker to make less than perfect cameras, they should all improve endlessly. 
However, I also believe we should work with our camera like partners, learn to use them, learn how they behave and react/control their behavior, make the most out of it.  When a camera can deliver 90-95% of its function will be good enough for me, not that I can accept less, but I can accept reality. Just like when my digital back cannot get good enough quality at the highest ISO setting, I can go to use 1Ds MK2/MK3 or some may go to Nikon. But this is just me.
[a href=\"index.php?act=findpost&pid=212976\"][{POST_SNAPBACK}][/a]

Khun K,

I agree with you about the nature of "perfection" largely because we don't know what that is - in the abstract. So I have a more operational definition of "perfection": things that work in the manner they are advertised or purported to work. That makes the whole discussion more tangible.

As for the quality of focus in cameras of yester-year, let us take care. We don't need to go back to the combination of a Leica M-series cameras with their Summicron lenses from decades ago, but a comment from Rob Galbraith which Tim Gray noted in this thread - to the effect that focus problems existing in the 1DMk3 did not exist in the 1DMk2. Doesn't this kind of make you think that sometimes technology regresses rather than progresses (perhaps because in the course of redesign things that were better before got messed-up now, and no-one caught it before releasing the product)?

I don't know of any technical reason why a higher sensor resolution should compromise focus in any way. If anything a higher resolution sensor should better reflect good focusing. Likewise, as far as I know, there is no technical relationship between depth of field and sensor resolution.

On your final point about accepting reality - sure - at some point in the sequence of events one has little choice, and no camera is equally good at everything even if there is nothing wrong with it. But when it comes to dealing with technical defects, the issue here is more up-stream about altering the reality so that camera manufacturers deliver what they advertise, or don't advertise it, or if they make a mistake, which is only human, acknowledge it and fix it as a matter of the highest priority. I think we the customers deserve this kind of consideration, and one is sometimes left wondering to what extent we really get it.  

Cheers,

Mark

[Eric Myrvaagnes]

It certainly seems that Canon has gotten a bit complacent recently. My hope, as one who is heavily invested in Canon glass, is that the new offerings from Nikon will frighten Canon into cleaning up their act a bit.

[Khun_K]

Khun K,

I agree with you about the nature of "perfection" largely because we don't know what that is - in the abstract. So I have a more operational definition of "perfection": things that work in the manner they are advertised or purported to work. That makes the whole discussion more tangible.

As for the quality of focus in cameras of yester-year, let us take care. We don't need to go back to the combination of a Leica M-series cameras with their Summicron lenses from decades ago, but a comment from Rob Galbraith which Tim Gray noted in this thread - to the effect that focus problems existing in the 1DMk3 did not exist in the 1DMk2. Doesn't this kind of make you think that sometimes technology regresses rather than progresses (perhaps because in the course of redesign things that were better before got messed-up now, and no-one caught it before releasing the product)?

I don't know of any technical reason why a higher sensor resolution should compromise focus in any way. If anything a higher resolution sensor should better reflect good focusing. Likewise, as far as I know, there is no technical relationship between depth of field and sensor resolution.

On your final point about accepting reality - sure - at some point in the sequence of events one has little choice, and no camera is equally good at everything even if there is nothing wrong with it. But when it comes to dealing with technical defects, the issue here is more up-stream about altering the reality so that camera manufacturers deliver what they advertise, or don't advertise it, or if they make a mistake, which is only human, acknowledge it and fix it as a matter of the highest priority. I think we the customers deserve this kind of consideration, and one is sometimes left wondering to what extent we really get it. 

Cheers,

Mark
[a href=\"index.php?act=findpost&pid=212985\"][{POST_SNAPBACK}][/a]

I agreed with you totally, just that I look at the camera a smaller part of photography, but, sure, any improvement from camera makers is welcome.  Perhaps I don't use the focus tracking at the maximum firing rate often enough to tell the defects or I just got lucky to have a good sample.

[Chris_Brown]

It certainly seems that Canon has gotten a bit complacent recently. My hope, as one who is heavily invested in Canon glass, is that the new offerings from Nikon will frighten Canon into cleaning up their act a bit.[a href=\"index.php?act=findpost&pid=212993\"][{POST_SNAPBACK}][/a]

If the replacement of the Firewire connection with USB is any indication, then Canon has lost site of making a perfect tool for working professionals. Only those whose income rests on the consistent quality of image will find faults like sloppy focus and crooked viewfinders. What market share is that? If that market share left and went to Nikon would it have an adverse effect on Canon? After all, the best selling digital camera of all time is the EOS Rebel. In addition, Canon has R&D in many areas. Could they be looking at other, more profitable industries?

[Slough]

I don't pretend to know anything about Canon gear - apart from second hand information - but there is a thread on this same issue in the Canon EOS forum on Photo Net. The overwhelming opinion is that there is no problem, apart from RG not being able to use the camera.

Whatever the truth of this matter, it is a gift to Nikon, as it has surely damaged Canon's reputation.

[Plekto]

So is this a problem with the optics or is it that the thing has too many focusing zones and choices for the software/AI to deal with?   Or is it just shoddy quality control and some units are junk while others are fine(kind of like VW - heh)?

[Mark D Segal]

I don't pretend to know anything about Canon gear - apart from second hand information - but there is a thread on this same issue in the Canon EOS forum on Photo Net. The overwhelming opinion is that there is no problem, apart from RG not being able to use the camera.

Whatever the truth of this matter, it is a gift to Nikon, as it has surely damaged Canon's reputation.
[{POST_SNAPBACK}][/a]

Slough,

I went to Photo.net and read each and every one of those posts. My interpretation of all that material is that a few of them took serious issue with RG's analysis. Others agreed with him and others were more agnostic one way or another.

Interesting also is Atkin's reference to [a href=\"http://www.prophotohome.com/forum/pro-photo-reviews-articles/76674-canon-canon-1d-camera-canon-mkiii-camera-vs-nikon-nikon-d3-camera-sports-autofocus-showdown.html]Drew's published analysis[/url] which is really worth looking at. He compared the 1DMkIII (about which he's been writing for quite some time regarding the focus issue in particular) and the Nikon D3. He found both of them to have less than stellar performance, but the Canon was less flawed than the Nikon.

Reading that makes me wonder whether we're not simply dealing with a situation of over-reaching and hype - not on the part of web reviewers, but by the manufacturers themselves. Maybe they are trying to deliver more than they really can, raising false expectations. It gets more interesting as it goes!

The next thing is that over at Photo.net the moderator abruptly closed the thread after 25 posts, none of which were rude, blasphemous, libelous or anything or the sort. One wonders why that happened?

Cheers,

Mark

[Ray]

He compared the 1DMkIII (about which he's been writing for quite some time regarding the focus issue in particular) and the Nikon D3. He found both of them to have less than stellar performance, but the Canon was less flawed than the Nikon.

Reading that makes me wonder whether we're not simply dealing with a situation of over-reaching and hype - not on the part of web reviewers, but by the manufacturers themselves. Maybe they are trying to deliver more than they really can, raising false expectations. It gets more interesting as it goes!
[a href=\"index.php?act=findpost&pid=213019\"][{POST_SNAPBACK}][/a]

Mark,
I wondered about that. One usually complains about a camera's performance because it's not up to the standard that is possible and evident in other models or brands. There's an implication in all this brouhaha that perhaps the latest Nikons do a better autofocussing job. If this is not true, then I think you are probably right in your assessment that Canon has simply raised false expectations in its advertising hype.