Faster (f/2.8) Lenses on 5DII or Slower (f/4) lenses on 5DIII?

[Loswr]

You apparently don't understand, or are ignoring, the reasons Why Focus-Recompose Sucks, to which agierke was alluding.

You apparently don't understand, or are ignoring, the reasons Why Focus-Recompose Sucks, to which agierke was alluding.

Neither of you apparantly understand that the shooting aperture has nothing to do with the focus accuracy.

I stand by it lads.

Utter tosh, and standing by it only makes you appear foolish.

To be clear, the issue isn't the accuracy of the focus system itself, but the accuracy of focus in the resulting image. Most people care about the latter, usually only CR geeks like me care about the former. Yes, we get that moving the camera after focus is achieved isn't, strictly speaking, a problem with focus system accuracy. We also understand that a deeper DoF with a narrower aperture is really just masking the problem. People who could easily understand that locking focus then walking two paces forward before taking the shot would lead to a misfocused image if the DoF is thin, sometimes have trouble understanding that shifting the camera in an angular motion can cause the same problem.

Are you saying that with a wide aperture and a relatively close subject (i.e., conditions resulting in thin DoF), focus-recompose doesn't lead to back-focused shots? Or are you merely quibbling over terminology?

 

[agierke]

Neither of you apparantly understand that the shooting aperture has nothing to do with the focus accuracy.

f4 might get everything reasonably sharp at 17mm on 135mm / leica / minature format DSLR's, but won't get you far anywhere else unless eveything in your scene is on the same focus plane.

The advice is useless, tosh. My biggest gripe is that somebody would actually read it and try it and miss out on important shots. Might as well have said get a lens with hyperfocal scales and shooting everything at f11. It's got about as much to do with how AF works. Jaysus.

Unless of course you meant that we should all buy lenses no faster than f4 so we can give our AF systems a chance for point and recompose. Or try just not using it

I stand by it lads.

i dont disagree with your technical assessment of either of your posts....but none of that has anything to do with what Neuro and I are suggesting to the OP.

to be clear, if AF accuracy is the desired result coupled with the focus and recompose technique then what i am advising to the OP is buying the 5D3 would be a better option for him using his existing lens lineup and using the improved responsiveness of the 5D3's outer points when composing RATHER than buying 2.8 zooms and continuing to do focus and recompose.

yes, buying 2.8 zooms offers better AF on the 5D2...but any accuracy gained is wasted on the focus and recompose technique as the very simple physics of the technique circumvent the accuracy of the original focus acquisition...unless of course your are shooting at F4 and above with the extra depth of field that would compensate for the discrepancies introduced with the recomposition. at this point....a wiser recommendation would to be just buy the 5D3 to take advantage of better, more accurate focus points outside of the center point and ditch focus and recompose.

this is sound advice...which does not ignore your completely valid points about how AF systems work coupled with different lenses capable of different maximum apertures. your posts do ignore however the very real problems the focus and recompose technique introduces to focus accuracy....especially at apertures greater than F4. the OP should be made aware of this to help inform him of his decision.

i appreciate your enthusiastic points...but you are arguing something that 1. i dont disagree with and 2. doesnt have anything to do with the actual point that i am discussing.

 

[chas1113]

In addition to the points above, my technique could very well change based upon the technological improvements and ergonomics of the 5DIII AF. Having been a focus/recompose shooter since my Nikon F days in the '60s, the 5DII outer points never engendered the confidence to change my shooting style. Maybe the tech in the 5DIII can teach this old dog new tricks.

For the record, I fully understood agierke's point about focus/recompose technique and its ramifications regarding faster, shallower depth-of-field lenses. And agree.

To mrsfotografie: regarding pre-focusing (f/8 and be there) techniques, distance scales have all but disappeared on newer lenses making my tired eyes all the more dependent on sensor tech. I do manual focus all the time with 30-year-old Contax lenses — on things that don't move. My interest is improved AF on things that DO move.

I appreciate all the viewpoints and spirited debate. Is that popcorn I smell?

 

[paul13walnut5]

Agierke: you had made you focus recompise point before the op had even mentioned that style of shooting.

Neuro: i'm saying there are better ways to solve the particular problem, one of them is a camera with more af points, more accurate af points, for whatever reason, say low light, f4 may not be desirable.

I'm really saddened that i look foolish to you Dr. Brain, breaks my heart that does. I guess that puts me in with just about everybody else who doesn't agree with you. What can I do to cheer myself up? i know i'll have a perusal of your smashing pics of all your flash stands lined up, or all your camera bags. That always lifts me and reminds me what photography is really all about.

 

[chas1113]

C'mon Paul, lighten up a little.

You points are just as valid and appreciated. Having shot for for decades myself, I understand stop-down metering and depth of field issues that your posts have addressed. Nobody's rightest here which infers that nobody's wrongest either.

Cheers.

 

[agierke]

excellent position to take Chas, i really wasn't trying to let this thread digress into questioning the validity of other members photography. i really only wanted to offer my opinion from my direct experience.

i hope it proves helpful and whatever you decide, you will certainly have an exciting purchase upcoming!

 

[celltech]

If you are in continuous, high speed servo AF shooting, does the lens/body have time to open the aperture/grab focus/close the aperture to proper setting between every shot?

 

[chas1113]

I'm not sure that was an issue in my case. I normally shoot in One-shot mode and back button focus. Seems from all I have heard and read that this is the most effective setup for non-predictive action... Assuming that you have basically pre-focused in advance, which I try to do.

 

[Loswr]

Agierke: you had made you focus recompise point before the op had even mentioned that style of shooting.

Neuro: i'm saying there are better ways to solve the particular problem, one of them is a camera with more af points, more accurate af points, for whatever reason, say low light, f4 may not be desirable.

I'm really saddened that i look foolish to you Dr. Brain, breaks my heart that does. I guess that puts me in with just about everybody else who doesn't agree with you. What can I do to cheer myself up? i know i'll have a perusal of your smashing pics of all your flash stands lined up, or all your camera bags. That always lifts me and reminds me what photography is really all about.

I'm sorry, Paul, but I don't see where I (or agierke, from what I can see) suggested shooting at f/4 or narrower as a solution to the problem of backfocus resulting from focus-recompose. Feel free to point out where I suggested that, ideally with a direct quotation. The point was that switching from an f/4 lens to an f/2.8 lens might reveal this problem, and that turned out to be a relevant observation to the OP.

You apparently jumped to the erroneous conclusion that those statements constituted advice of some sort (beyond the simple description of the problem), proceeded to call that description 'tosh', accuse me (and agierke) of not understanding the very concept being accurately described, and then descended into sarcastic insults. If you want to be saddened by something, you need look no further than your own immature behavior.

 

[frumrk]

Not that this matters... but I have been thinking of upgrading my 70-200 F4 to a 2.8. I also thought about upgrading from a MK II to a MK III instead. So I took 3 of my lenses...

50 F1.8
24-70 F2.8
70-200 F4

I was able to find a spot in a dark area that was difficult to focus and then attempted to focus on that area with all three lenses.

I used both the 24-70 and 70-200 at 70mm and I moved up slightly for the 50 1.8.

Now I am not talking about focus accuracy... simply focus lock. In my results... which I didn't expect... here is the order in which the lens performed the focus lock the quickest.

70-200
24-70
50

I thought it would be exactly the opposite.... can anyone explain?

 

[chas1113]

Interesting result. Assuming the testing procedure is not in question, I'd like to hear any conjecture as well. My own interpretation is that if your 24-70 is older, even the f/2.8 aperture advantage of that lens might not best the newer 70-200 f/4 IS focusing abilities. The 50mm I have no clue about. One would think the f/1.8 would focus appreciably better than either zoom. But this does reflect on my original query: Does the centerpoint 2.8 sensitivity of most Canon dSLRs (and the accompanying fast glass) really pay off when comparing different generations of sensor technology.

 

[Loswr]

In my results... which I didn't expect... here is the order in which the lens performed the focus lock the quickest.

70-200
24-70
50

I thought it would be exactly the opposite.... can anyone explain?

Actually, I don't find that particularly surprising. Here's the thing - an f/2.8 line is more accurate, but that accuracy isn't free. The price you pay for accuracy is speed. In fact, the predictive algorithms for AI Servo AF use mostly data from the f/5.6 lines vs. the f/2.8 lines, because the former are faster. Lens AF speed also plays a role, and the 50/1.8 is the slowest of the bunch.

 

[Drizzt321]

Not that this matters... but I have been thinking of upgrading my 70-200 F4 to a 2.8. I also thought about upgrading from a MK II to a MK III instead. So I took 3 of my lenses...

50 F1.8
24-70 F2.8
70-200 F4

I was able to find a spot in a dark area that was difficult to focus and then attempted to focus on that area with all three lenses.

I used both the 24-70 and 70-200 at 70mm and I moved up slightly for the 50 1.8.

Now I am not talking about focus accuracy... simply focus lock. In my results... which I didn't expect... here is the order in which the lens performed the focus lock the quickest.

70-200
24-70
50

I thought it would be exactly the opposite.... can anyone explain?

So the 70-200 was fastest, 24-70 next, and 50 f/1.8 slowest? That's a 24-70 v1?

I'm not surprise the 50 f/1.8 was the slowest. While it has a faster f-number, it has a slow and old AF motor, and it's actual transmissive qualities may or may not be as good as the rest. Which that can certainly matter, especially in a location with a low amount of light.

 

[frumrk]

Not that this matters... but I have been thinking of upgrading my 70-200 F4 to a 2.8. I also thought about upgrading from a MK II to a MK III instead. So I took 3 of my lenses...

50 F1.8
24-70 F2.8
70-200 F4

I was able to find a spot in a dark area that was difficult to focus and then attempted to focus on that area with all three lenses.

I used both the 24-70 and 70-200 at 70mm and I moved up slightly for the 50 1.8.

Now I am not talking about focus accuracy... simply focus lock. In my results... which I didn't expect... here is the order in which the lens performed the focus lock the quickest.

70-200
24-70
50

I thought it would be exactly the opposite.... can anyone explain?

So the 70-200 was fastest, 24-70 next, and 50 f/1.8 slowest? That's a 24-70 v1?

I'm not surprise the 50 f/1.8 was the slowest. While it has a faster f-number, it has a slow and old AF motor, and it's actual transmissive qualities may or may not be as good as the rest. Which that can certainly matter, especially in a location with a low amount of light.

Not only am I saying that it was slower... but actually less successful at locking. The test procedure wasn't really scientific... however I did test 3 times with each lens (swapping lenses each time) and the results were always the same. The 50 rarely was able to achieve focus... and the 70-200 F4 was able to achieve focus most of the time.

 

[Drizzt321]

Not that this matters... but I have been thinking of upgrading my 70-200 F4 to a 2.8. I also thought about upgrading from a MK II to a MK III instead. So I took 3 of my lenses...

50 F1.8
24-70 F2.8
70-200 F4

I was able to find a spot in a dark area that was difficult to focus and then attempted to focus on that area with all three lenses.

I used both the 24-70 and 70-200 at 70mm and I moved up slightly for the 50 1.8.

Now I am not talking about focus accuracy... simply focus lock. In my results... which I didn't expect... here is the order in which the lens performed the focus lock the quickest.

70-200
24-70
50

I thought it would be exactly the opposite.... can anyone explain?

So the 70-200 was fastest, 24-70 next, and 50 f/1.8 slowest? That's a 24-70 v1?

I'm not surprise the 50 f/1.8 was the slowest. While it has a faster f-number, it has a slow and old AF motor, and it's actual transmissive qualities may or may not be as good as the rest. Which that can certainly matter, especially in a location with a low amount of light.

Not only am I saying that it was slower... but actually less successful at locking. The test procedure wasn't really scientific... however I did test 3 times with each lens (swapping lenses each time) and the results were always the same. The 50 rarely was able to achieve focus... and the 70-200 F4 was able to achieve focus most of the time.

Something to actually know, the newer L lenses (forget if those 2 are included) seem to have a better feedback and higher accuracy than older lenses, of which the 50 1.8 definitely is among.

 

[Rienzphotoz]

Hi chas113, I believe you've already made up your mind to go with the body upgrade ... 5D MK III is a fantastic camera I'm sure you will see a far much snappier performance in the way your f/4 lenses work on 5D MK III over the 5D MK II.
5D MK III is my first full frame DSLR as I upgraded it from 7D ... but I took a different approach i.e. upgraded my f/4 lenses to f/2.8 lenses before upgrading to 5D MK III ... no "scientific" approach, other than wanting to invest in f/2.8 zooms (16-35, 24-70 & 70-200) before I made the body upgrade, it took several years but I eventually did it.
Good luck with your upgrade ... may it serve you well.

 

[chas1113]

Hi chas113, I believe you've already made up your mind to go with the body upgrade ... 5D MK III is a fantastic camera I'm sure you will see a far much snappier performance in the way your f/4 lenses work on 5D MK III over the 5D MK II.
5D MK III is my first full frame DSLR as I upgraded it from 7D ... but I took a different approach i.e. upgraded my f/4 lenses to f/2.8 lenses before upgrading to 5D MK III ... no "scientific" approach, other than wanting to invest in f/2.8 zooms (16-35, 24-70 & 70-200) before I made the body upgrade, it took several years but I eventually did it.
Good luck with your upgrade ... may it serve you well.

Good to know, thanks. It's nice to have confirmation from multiple sources. My approach to kit building was "slow zooms/fast primes that are lightweight and sharp enough"...but with the 24-70 II coming down in size and weight....it's very tempting. For me, my 17-40 and 24-105 serve me well enough, but there are times when the AF of the 5DII just is so frustrating. Hence my original post. I don't shoot a lot of thin DOF and everyone here has pretty much confirmed upgrading body would be more worthwhile. If the 16-50 f/4 IS comes to pass — (and is great!) — I could see it replacing both zooms. That and my 70-300L and I'd be set.