So hi everyone. I'm new here. Been grabbing useful info from here and entertaining myself with gear lust for a while, but I finally found a reason to post. So with that...
Warning: This is going to be a book. For those of you interested in the conclusion, just hop to where I wrote "conclusion". I'll try to be as brief as I can....
The Canon EF 24mm f/1.4L II USM. This is the newest lens I decided to add to my collection after finding myself in need of a good, fast, sharp, wide angle solution. So I picked one up for a relative steal on CL (rather than going for my other consideration, the venerable EF 16-35 f/2.8L II), and got to playing around. The first night, I tried it while with the seller, and decided that I'd take it home, despite that I wasn't 100% sure of the decision yet, or the image quality is was producing for me, but I figured that for the price at which I acquired it, I could probably make money if I resold it. That said, day 2 rolled around, and I decided to take it outside and start really seeing what it could do. I was impressed. Like, really impressed. It is an absolutely exquisite lens to shoot with, and produces some effects that I just haven't seen another lens capable of producing at this focal length, in part due to that beautiful f/1.4 aperture. It's remarkably sharp, even wide open (full disclosure, I'm on a 7D, so can't say that much about the full frame corner thing...yet, give me a few more days), and produces some wonderful blur outside of the DOF range.
So then day three rolled around and I started planning a shoot with it. While conceptualizing, I tried a few test shots of a setup indoors under some craptastic incandescent lighting, and noticed that my photo quality had kinda gone to poooo. I was suddenly a bit bothered, wondering how something so capable one day ago had produced this hideous shot that now sat upon my LCD. So I took a few more shots, trying to hold really still (yeah, I know it's 1.4 and all), and sadly, to no avail. About now, I became concerned. It appeared that the focus on the lens had grown very soft, and my images were looking downright terrible. Now I had read a thing or two about strange autofocus issues with this lens whilst perusing reviews of it prior to and after acquiring this one; hence the impetus for my concern. Had I acquired a bad copy? Is that why the guy sold it to me, and especially at such a decent deal? Well deuces...
Anyhew, after some reading, this got me to thinking that it was time for some good ol' scientific method. So I sat down at the table, pulled out some DVD cases, arranged them upright in progressively further distances, all perpendicular to my focal plane (so as to test for front and back focusing), pulled up the custom function for Micro Adjustments, and got to work. What I found was wild inconsistency. Sometimes it would be tack sharp, sometimes it was as sharp as a warm stick of butter. And all of this without moving the camera a single millimeter backward or forward, nor side to side. Needless to say, I was perplexed. How can I not move the camera a single iota, but still garner such maddeningly inconsistent results? Well, no scientific method is useful without controls and variables, so I set to discovering an applicable variable.
I tried adjusting the Micro Adjustment to +5, then took a series of three shots. Generally, one in the series would be sharp, with the other two displaying varying degrees of sharpness or softness. This would happen each time. I adjusted to -5. Pretty much the same results. I adjusted to +5 again. To 8. To 15. To 20. To 10. Though some were clearly worse than others, the problem seemed to persist despite the micro-adjustments. Clearly the problem was acting independent of this variable, so I had to try something different. With that, I moved to simply trying different apertures. One would think that to be intuitive, with 1.4 having such a narrow depth of field. Same results each time. Inconsistent, generally with one of three shots being sharper than its companions, though it was ever so slightly less of an issue at smaller apertures. What could be going on here? Was it really just a bad lens copy?
Enter light quality. Or perhaps quantity. Or both. I pulled out the old Fenix flashlight to shed some serious light on the matter and determine if it was simply an issue of there not being enough light for the auto focus system to function as well as it could/should. Admittedly, the kitchen lights at night can be a bit...insufficient, but with a 1.4 lens, I figured there would be no issue, especially since I was shooting on a table with a self timer to prevent any camera movement between shots in the sequence. Well, apparently I was dead wrong on that. Upon issuing that bright white light, I suddenly noticed that the variability between shots in the sequence had dramatically diminished. Interesting. So with that, I set the Micro Adjustments to zero, and began playing. I achieved relatively consistent results as long as that light was shining on the subjects. I moved through the aperture range. Even more consistent as I stopped down. Interesting. So I tried another light source, this one also a relatively pure white LED light, though more distant, and bounced off of the ceiling instead. Close to the same results, though not quite as accurate; minor deviations throughout, but negligible ones at worst.
So at this point I had a hypothesis and decided to make my testing as accurate as I possibly could. I switched on mirror lockup, pulled out the wireless remote, and thus made sure that there was no influence at all from my touching the shutter button. I repeated the experiment with all three types of light, and each time, the strongest white light seemed to bring the camera to a relatively perfect focus each time. I tried getting a focus point with that light, then switching to manual focus and shooting a three series. ZERO variation. I tried getting auto focus with that light, turning the light off, turning AF off, THEN taking a three shot series under the incandescent. ZERO variation. Dead tack sharp every time. My hypothesis was beginning to pan out pretty much exactly as I expected. Finally, I started moving the camera back and forth in distance from the subject while repeating each different phase of the experiment. Each time, the variations were slightly greater when closer to the subject, and also, each time, the variations were slightly greater at the wider apertures, exactly as one would expect, but why the variation at all? What was causing this problem?
I hooked up my EF 100mm f2.8L IS USM Macro lens to see if the issue was present in it. There was no variation at any point, in any light, whatsoever. That lens, I know for a fact is as solid as a rock. So how can I relatively succinctly explain the conclusion that I've come to on the issue with this lens? As follows:
Conclusion:
So I've heard that the 5D Mark III doesn't seem to display this anomaly. My conclusion is this: The EF 24 f/1.4 II obviously has a relatively razor thin aperture at f/1.4, as we all know. And the autofocus system on many slightly older dSLRs, (the 7D is beginning to age just a tad now) use slightly older and thus less accurate and proficient algorithms to acquire their focus each time that you half depress the shutter button to initiate said sequence. Depending on infinitesimal differences in focusing location, and especially their application to this algorithm as the processor processes them, there are going to be minor variations in the appropriated plane of focus as dictated by the autofocus system's acquisition of a given focus point. That said, when you're working on such a paper thin focal plane, as is associated with the f/1.4 aperture, you are more liable to experience a loss of focus due to tiny movements of your hands, or what have you. We all know and accept this. What I'm proposing, is that the slightly more antiquated, and thus relatively inferior (when compared with the 5D III) autofocus system found on the 7D (and lesser bodies), in cooperation with their lower light performance, is causing them to search a bit harder to find their focus point in LOWER LIGHT than a more advanced, more recent autofocus system, and therefore it is the addition of more, or higher quality light that allows the AF algorithm to repeatedly reproduce accurate results. When the system has to hunt a bit more because of lower levels of light, or lower quality of light, it is inherently less likely to lock focus on EXACTLY the same millimeter as it did in the frame that was shot immediately prior to this now most recent acquisition. When the light amount or quality is improved, it has a much easier time nailing the same point again and again, thus giving you consistent focus. This is why I believe that I experienced inferior results whilst purchasing the lens under those shoddy indoor lights at night, as well as in my kitchen under those same craptastic incandescents, as opposed to during the day when I took it outside in broad daylight and experienced such phenomenal results. It is a combination of the lens' inherently high level of sensitivity, and extremely narrow potential DOF, in combination with the AF system's tendency to be less accurate in low light situations, that produces the discussed variations.
Now what of the EF 85mm f/1.2L USM? I'm afraid I can't discuss that point, since I don't own that piece of glass. I don't know if it experiences similar problems, and if not, then why that would be the case.
SO, what does this mean? Well, to me, it means that this lens may not be quite as stellar for low-light applications on a camera body that is not as good at low-light performance as I had initially hoped. This theory would hold consistent, given that the majority of complaints that I have seen with this lens have been on lesser bodies than the 5D II or the 5D III or the 1D or 1Ds series. The lens is capable of amazing things, there is no doubt about that, but I think that if not married to the right body, or if married to such a body under less than optimal conditions, one must simply expect less consistent, and most probably, inferior results. Unfortunate, its true, but I'm not sure that it's enough to make me sell this lens that can perform so fantastically under the right circumstances.
Now I know that this post was extremely long winded, and I apologize for that. I actually tried to keep it as curt as I could, leaving out some minor details, but I'd love to hear discussion from anyone who owns this lens to see if this theory can be confirmed.
Do you have the 24 II?
Do you also have a 5D or higher body, or a 7D or lower body?
Do you shoot mostly in well lit situations, or do you shoot more in low light?
When shooting in low light, and if you own a 7D or lower body, do you discover that your lens' performance is less consistent, or gives consistently poor results?
If so, do you find that these AF issues resolve with the addition of more light, better light, or perhaps stopping down your aperture?
I'd really like to get some confirmation on this, because I know there are a lot of people out there concerned about this, and a lot of them are on forums asking these very questions. So if anyone's ready and willing, chime right in!
Thank you to everyone for all of the sound advice you've provided me so far on this forum, and thanks to those of you who stuck through this beast of a post; and especially, thanks in advance to those of you who participate in the discussion!
All my very best to all,
-The Brains of Ape
