1d MkII, 1D, 10D a słabe światło [ang. link]
z cyklu Chuck Westfall wyjaśnia
dSLR low light
miłej lektury
pozdrawiam,
g3
Oto i text
<Chuck Westfall>
As promised, I have looked into the questions that were raised in the previous thread about EOS-1D Mark II AF performance in low light. After a series of tests with EOS-1D Mark II, original EOS-1D and EOS 10D with a variety of popular EF lenses (16-35/2.8L, 50/1.4, 70-200/2.8L IS), I am reasonably certain about the conclusions I’ve reached. To make a long story short, the focusing systems on all 3 of these cameras are functional in low light, with some limitations.
TESTING METHOD: I used a fairly large room with lighting that could be turned on or off. When the room lights were off, there was enough spill light from surrounding areas that the exposure setting for ISO 1600 at f/2.8 was approximately ¼ second. This is equivalent to EV 1 at ISO 100, about 1 stop brighter than the cameras’ rated AF sensitivity threshold. The test target was a large box with a printed side panel. The camera-to-subject distance was approximately 25 feet. All cameras were set to ISO 1600. Focusing methods were intentionally varied between One-shot and AI Servo, and also between automatic and manual focusing point selection. Each camera was tested with 3 EF lenses (16-35/2.8L, 50/1.4, 70-200/2.8L IS), first with a 550EX Speedlite and then once more using each lens with no focusing aids.
So many relevant issues emerged that I decided to break them down, one by one, in order to make a clearer explanation. Please read each of the following points for more information.
1) Lens communication: This should go without saying, but I’m putting it right up front to cover all the bases. Make sure that your lens is communicating properly with your camera before checking any kind of functionality. After mounting the lens on the camera and turning the camera on, check the camera’s top LCD data panel to ensure that the aperture readout is functioning. If the aperture display reads 00, the lens is not communicating electronically with the camera. Rectify this condition, by cleaning the lens contacts or having the lens serviced, before proceeding further.
2) Low Light Threshold: All EOS cameras are rated in terms of low-light sensitivity for autofocus as well as exposure metering. The specifications for each camera can be found in the owner’s manual. For EOS-1D, EOS-1D Mark II, and EOS 10D, the low-light threshold for AF sensitivity is rated at EV 0. Accurate exposure at this light level requires camera settings equivalent to 2 seconds at f/1.4 at ISO 100. If the ISO is reset to 1600 and the aperture is set to f/2.8, the equivalent shutter speed would be 0.5 seconds. Do real-world shooting conditions ever fall below this light level? Absolutely. Wedding receptions at night are a good example. But flash is normally used in order to freeze the action and get a good exposure. Read item 7 for more information.
3) Subject Contrast: All EOS cameras utilize a passive AF system that requires a readable level of subject contrast in order to function. In other words, the camera won’t focus on a blank wall or a clear blue sky (unless there’s some sort of texture) even when there’s plenty of light. Is this a bug? Not really. It’s a limitation to be sure, but it’s also a trade-off for excellent focusing accuracy at any subject distance provided a reasonable level of light and contrast, compared to active AF systems which can only function over a limited distance range.
4) Focusing Speed: Like it or not, autofocusing speed in the EOS system is largely dependent on the following factors:
i) Light Levels: The darker it gets, the longer it takes for the AF system to lock on to the subject, regardless of the lens in use.
ii) Maximum Aperture: In low light conditions, the faster the lens, the faster the AF.
iii) Focal Length: Longer focal length lenses like the EF 70-200L have a much greater range of defocus than standard or wide-angle lenses. In other words, when an image goes out of focus on a fast telephoto lens, it *really* goes out of focus. This can cause delays or sometimes even failures in low-light AF.
5) One Shot vs. AI Servo: Beyond the basic distinction between "locking" focus for stationary subjects (One Shot AF) and "tracking" focus for moving subjects (AI Servo AF), One Shot AF in general performs slightly better in low light than AI Servo AF. There’s a very simple reason: the pixels in the AF sensor are allowed more time to accumulate light in One-Shot than in AI Servo. Therefore, if maximum AF sensitivity in low light is the objective, select One-Shot AF.
6) Automatic vs. manual focusing point selection: When the camera is set for automatic focusing point selection, you are letting it pick a focusing point for you. With a multi-point AF system, this method inevitably takes a fraction of a second longer than selecting a single focusing point manually. Moreover, the central focusing point on any EOS camera has the highest performance of all focusing points. Therefore, if maximum AF speed is the objective and you have enough time to identify your subject before you shoot, consider using the center focusing point. BUT REMEMBER: If you are using a 1D or 10D with an E-TTL Speedlite, recomposing after you lock focus may throw off the flash exposure. In this case, we recommend the use of Custom Function 4-1 or 4-3 as outlined in item 8 below. EOS-1D Mark II users do not have this limitation because of E-TTL II.
7) Focusing aids: The EOS-1D and EOS-1D Mark II do not have a built-in AF Assist beam. The EOS 10D does, but it exists in the form of a stroboscopic burst from the built-in flash. This method can be effective over a limited distance range, but it is annoying to most human subjects, and thankfully unavailable when using an EX Speedlite. For best low-light AF performance with these cameras, we strongly recommend the use of the near-IR AF Assist beams built into EX Speedlites and the Speedlite Transmitter ST-E2. These devices project a red, patterned beam onto the subject and are effective over varying distance ranges depending on the accessory you choose. Speedlite 550EX has the most powerful AF Assist beam. It is typically effective at distances up to approximately 10 meters/33 feet or more. However, the camera MUST be set to One-Shot AF in order for the AF Assist beam to function. (And despite reports to the contrary in this thread, the 550EX’s AF Assist Beam is fully functional with the EOS-1D Mark II, as long as the camera is set for One-Shot AF.) Speedlite Transmitter ST-E2 provides an AF Assist beam without firing a flash, so it’s useful for non-flash photography in low light. If that is your objective, be advised to stay away from Program mode on the camera, since the lowest shutter speed will be limited to 1/60 second, resulting in massive underexposure in low light without flash. Pick any other exposure mode, but stay away from Program unless you are using flash.
8 ) Custom Functions/Personal Functions: Personal Function 14 on the EOS-1D and EOS-1D Mark II can be helpful to reduce delays in low-light AF without the use of an EX Speedlite or ST-E2. This P.Fn prevents the AF system from searching when it can’t find focus, which may occur due to lack of contrast, low light, etc. Custom Function 10 on 1D, Mark II and 10D controls the behavior of the red superimposed display over active focusing points in the camera’s viewfinder. Some users may find that increasing the brightness of the display is helpful in low light. Custom Function 4-1 or 4-3 can also be very useful for low-light AF, especially with an EF lens that supports Full-Time Manual (FTM) focusing. Under this condition, you can manually focus the lens at will if the AF system isn’t able to lock on. You can also stop the lens from searching by simply lifting your thumb off the AE lock button.
9) Shutter release technique (half-press vs. mash): If you want the AF system to function at all in low light, you need to give it some time to get a reading by pressing the shutter button halfway and holding it there before you shoot. If you simply mash the shutter button down without waiting for the AF to function, the shutter release will either lock up (in One-Shot AF with C.Fn 4 set to its default) or release without focusing (at all times in AI Servo or with the combination of One-Shot AF and C.Fn 4-1 or 4-3 if your thumb is not on the AE lock button). This is true with or without the use of an AF Assist beam, so get used to it.
10) Color temperature: It’s a fact that low-light scenes with tungsten illumination are very low in color temperature, but my tests show that the color temperature of the illumination has no effect on the accuracy of the autofocus. As long as the subject is parallel to the camera and has a sufficient degree of contrast and illumination, the camera will autofocus accurately at any color temperature from 2000K to 10000K.
<Conclusions
Even under the most ideal shooting conditions such as the use of a fast wide-angle lens with a contrasty subject and the presence of a 550EX AF Assist beam with One-Shot AF and center-point focusing, it may take an EOS camera almost a full second to lock focus on the subject in extremely low light. This is especially true if the lens is starting from its closest focus setting and attempting to autofocus a distant subject. This is simple physics: it takes X amount of time vs. Y amount of light vs. Z amount of contrast for the AF sensor to accumulate enough information to function. That’s the way it is, and firmware updates will not change it. Therefore, if fast response is more important than accurate AF, consider the use of manual focus instead. If accurate AF is more important than fast response, have the patience to wait for the AF system to work. If you can’t get a lock in One-Shot AF even with an AF Assist beam, consider using manual focus to get "in the ballpark" and then let the AF take over to fine-tune if necessary.
EOS-1D Mark II Low-Light AF Performance vs. EOS-1D and EOS 10D
Despite conflicting reports from other users in the previous thread, I found that the EOS-1D Mark II’s AF system performed at the same level as the original EOS-1D, and both of these models were noticeably faster for low-light AF than the EOS 10D. In the particular case of the EF 70-200mm f/2.8L IS USM, the 10D was the only camera that failed to lock focus without a focusing aid in low light with my test, whereas the original 1D and the Mark II had no problem.
Last but not least, the tests showed that AF speed increased rapidly as overall light levels improved. This may not do you any good under some conditions, but if you have any control over the light level for your photography, try to bump it up a little if you want faster AF. Every little bit helps!
--------------------
Chuck Westfall
Director/Technical Information Dept.
Camera Division/Canon U.S.A., Inc.
</Chuck Westfall>
