Lens for dog agility / action

[HDNitehawk]

I think the price would exceed the 70-200mm f/2.8L IS II that I'm trying to avoid buying.

I knew if you worked the thought process long enough you would stop avoiding the right answer. I know from experience the 70-200mm II will do the job.

Denise did have the right idea though, primes will rule over the zooms for the fast moving little doggies. The primes just to fast and accurate, where the zooms are less so and the AF point varies over the range. You are not going to match the sharpness either. A thing I always weigh when choosing one over the other, will the truly great shots out of the prime make up for the good shots I miss because I do not have a zoom.

[jrw]

Provided you only want to shoot at outdoor events, the 70-200 f4 should be adequate. There may be times, heavy cloud cover, that you would prefer to have the f2.8 to allow fast enough shutter speeds to freeze the legs and paws if you want them to not be blurred.

7D Servo AF should be up to the job. Hardest part is keeping the active focus point where you want it on a bouncing target while also watching the framing and adjusting zoom to keep up with changes in subject distance. Can be hard enough with runners until you get used to it, let alone a smaller faster target. Seems to me that there is at least one reviewer of camera bodies that has used running dogs to test the Servo abilities. It is a challenge for both camera and photographer.

As far as the f2.8 goes, I'd suggest that you start with it in the first place if you can possibly swing it. It will save you money in the long run as there will come a day when you will, not might, be buying it for the wider aperture.

[neuroanatomist]

I think the price would exceed the 70-200mm f/2.8L IS II that I'm trying to avoid buying.

But...the 70-200mm f/2.8L IS II really is the best lens for what you describe... You discount the f/2.8 non-IS for sharpness, not weight - the IS II is not significantly different is size/weight from the non-IS.

I'm puzzled by Rich's statements about slower focusing, too. The aperture doesn't directly affect focus speed, AFAIK. While an f/2.8 lens provides more precise AF, that precision does come at a slight cost in speed - that's why the majority of the data used to drive the AI Servo calculations actually come from the f/5.6 sensor lines - they have a faster response time (similarly, with the high-precision AF points, the AF system uses the co-located f/5.6 line(s) for a 'coarse focus' then refines that with the f/2.8 line(s).

The main problem (perhaps that's too strong a term) with a variable aperture zoom is when you shoot in M mode, since zooming changes exposure. You're shooting with a 7D, though - if you use M mode but set ISO to Auto, the camera will adjust the ISO to provide metered exposure in M mode (and the forthcoming 7D firmware update will allow you to set a cap on Auto ISO lower than 3200, if you prefer).

No idea about focus hunting with the 70-300mm, but that issue notwithstanding, I'd pick the 70-300mm f/4-5.6L IS over the 70-200mm f/4L IS.

[Richard Lane]

Richard, this is the first I've heard of the maximum aperture affecting the focus speed. I know there is the f/2.8 focus sensor for higher precision, but that an f/4 lens would focus half the speed of an f/2.8, or and f/5.6 being quarter speed, I've certainly never heard of that.

Sorry, about the confusion. I combined my thoughts of AF f/2.8 sensor speed and aperture focusing speed. To clarify, I have edited my original post and I have changed the words "focusing speed" to "lens speed." The f/2.8 is definitely much faster, but my two f-stop, 4 times faster focusing example is overstaing the difference. I meant for my example to reflect "lens speed" as in the faster shutter-speed that may be obtained. It is true however, that the more light the maximum aperture lens lets in, the faster the AF sensor will acquire focus.

Rich

[HDNitehawk]

I'm puzzled by Rich's statements about slower focusing, too. The aperture doesn't directly affect focus speed, AFAIK. While an f/2.8 lens provides more precise AF, that precision does come at a slight cost in speed - that's why the majority of the data used to drive the AI Servo calculations actually come from the f/5.6 sensor lines - they have a faster response time (similarly, with the high-precision AF points, the AF system uses the co-located f/5.6 line(s) for a 'coarse focus' then refines that with the f/2.8 line(s).

Interesting. Why do the supertele's focus slower when you put a 1.4x or 2x extender on? Decrease in aperture, increase in length or addition of hardware?

Has any one clocked the focus speed of a variable aperture zoom to see if the speed increases or decreases from one end to the other?

[Jayson]

You could always look at the 70-200mm f/2.8 L (non is version). You are going to be shooting at a high shutter speed anyway so IS isn't going to do much good if your just using for dog agility. In addition you get the high precision points of having a 2.8 lens. I believe that is priced right in that range also. I personally use my 70-200mm f/4 L IS, 85mm f1.8, and my 100mm macro USM If you limit the focus on the macro it is just as quick as the others. There is also the 100mm f/2 which is the twin of the 85mm. Lots of people love that lens.

[neuroanatomist]

It is true however, that the more light the maximum aperture lens lets in, the faster the AF sensor will acquire focus.

It's more about contrast than amount of light (although the two are related). But a high contrast transition at an f/5.6 AF point will lock on faster than a low contrast transition at an f/2.8 point. The AF sensor lines have a minimum threshold at which they operate, and as long as the intensity of light is over that threshold and there is sufficient contrast, there is no speed difference with more light. As I stated, a dual-sensitive point will actually use the f/5.6 part to 'pre-focus' because the f/5.6 line(s) are faster than the f/2.8 line(s).

As Canon has stated, "Using f/2.8 light rays is advantageous for two reasons. Because the aperture is larger and optically brighter, it is possible to detect the focus in darker conditions. And because the base line used in the triangulation principle is longer, more accurate focus measurements are possible. The disadvantage is that it may take some time to search and find the focus when the subject is well out of focus because the detection range at f/2.8 is narrow."

The 'darker conditions' apply to light levels down near the low end of the AF sensor sensitivity (-0.5 EV to -2 EV), where more light can push it over the threshold.

Interesting. Why do the supertele's focus slower when you put a 1.4x or 2x extender on? Decrease in aperture, increase in length or addition of hardware?

It's implemented due to the camera firmware (non-reporting TCs do not slow down AF). Canon has stated the speed reduction is important for greater accuracy.

[Richard Lane]

The aperture doesn't directly affect focus speed, AFAIK. While an f/2.8 lens provides more precise AF, that precision does come at a slight cost in speed - that's why the majority of the data used to drive the AI Servo calculations actually come from the f/5.6 sensor lines - they have a faster response time (similarly, with the high-precision AF points, the AF system uses the co-located f/5.6 line(s) for a 'coarse focus' then refines that with the f/2.8 line(s).

I was referring to the concept that twice as much light from the larger aperture was being directed towards the AF Sensor, which will increase AF acquisition speed, and why an f/2.8 lens set at 5.6 will focus faster than a f/5.6 lens set at 5.6. Of course this will be accentuated in lower light conditions and/or at faster shutter-speeds in which the aperture will be open for a shorter duration.

Rich

[DavidEccleston]

So it seems the consensus is we should keep saving our pennies, until can afford the 70-200mm f/2.8L IS USM II.

The 70-200mm f/4L IS USM *should* work, as *should* the 70-300mm f/4-5.6L IS USM. It's a coin toss as to whether people prefer the constant f/4 vs. the longer reach.

The next question becomes, if the f/2.8 AF sensor slows down AF, is it actually going to reduce servo performance? The camera will have to work harder to lock onto a bouncing non-constant speed object. Will the increases sensitivity combined with the slower performance mean fewer focus locks, and therefore, fewer shots?

I'm glad there isn't a simple answer to all of this.

[jrw]

The really short answer is that both the f4 and the f2.8 II versions of the 70-200 focus really quickly.