Focus speed: Lens vs Camera

I have several different lenses, and I know from using them that some are faster to focus than others. But I have also read about the new 40d and how much faster focusing it is supposed to be than my good ole' 300D, and the 1D series even faster than the xxD series.

My question is, what percentage of overall focusing speed can be attributed to the lens and what percentage can be attributed to the camera body?

I would have thought the electronics involved in focusing would react as near instantaneously as makes no difference and that any delay in achieving actual focus would be attributed entirely to the lens.

The camera body will however be doing all the computations involved, and also providing the power to drive the lens, so I think it perfectly feasible that the body would play a major part.

I'm thinking closer to 50/50 60/40.

Lens 30% --- Camera 80%
My 50-1.4 focuses fastest and most accurately on the 1D.
My 135L and 200L focus fast on everybody I own.
Lens 30% --- Camera 80%

Psychic1 wrote in post #4329351
Lens 30% --- Camera 80%
My 50-1.4 focuses fastest and most accurately on the 1D.
My 135L and 200L focus fast on everybody I own.
Lens 30% --- Camera 80%

That math doesn't add up

Psychic1 wrote in post #4329351
Lens 30%
My 135L and 200L focus fast on everybody I own.

What about the people you don't own?

Hard to give a firm answer - both contribute. The body does the computations, while the lens has to actually provide the movement. The 85/1.2, for example, is very slow to AF regardless of what body it's on. On the other hand, a lot of Canon's USM zooms focus very quickly on nearly any body, at least in good light.

An f/2 lens will allow the camera to detect and compute focus more quickly than a slower lens, say f/5.6 due to the increased amount of light available to the focus sensors.

A 1-series body with a dedicated AF processor and will compute focus very quickly, which is necessary for high-speed tracking in AI servo.

1/2 body as it provides all the smarts, power, and the focus sensor is in the body as well. The other 1/2 is lens. It's all dependent on the speed of the motor, the aperture which will allow the focus sensor to do it's thing, and just the design of the lens. I have the 85 f/1.2 and the 100 f/2.8 and both are just plain slow no matter what body they are on. The 17-55 f/2.8 however is real quick.

A more interesting test would be how dependent is the body in lens focus performance. Take one lens, put it on a old Rebel and measure focus performance, then move it on up ending with the MkIII measuring focus performance along the way.

It seems to me that my lenses focus faster on my MkIII than my 40D but Ive got nothing scientific to support that...

That would be an interesting test if there were an actual way to measure it. Basically take a lens that is known to be a really slow focuser and test it on a xxxd, xxd, and xd. Then do the same thing with an average speed focsing lens, and then again with a really fast focusing one.

It's something that I've wondered about for a while. Although I've vowed to keep shooting my 300d until it dies, I sometimes (ok, often) temp myself with the idea that a new body that would allow my lenses to focus faster would yield me more sharp pics when trying to capture kiddo pics. I have a 1, 6, and 8 year old and everything in our house is in constant motion. I've managed to keep my urges in check for 3 generations of camera bodies. I guess we'll see how much longer I last.

...Is that xmas music I hear?

Mr B Snappy wrote in post #4329103
The camera body will however be doing all the computations involved, and also providing the power to drive the lens, so I think it perfectly feasible that the body would play a major part.

The way I see it is this : When the shutter button is half pressed the lens starts to adjust for focus more or less immediately. This is because the camera has either immediately computed the focus error and sent a focus command to the lens or that it cannot sensibly detect a focus point and has sent a commend for the lens to oscillate through it's complete focus range in the hope of achieving a focus point. The important bit here is that the camera is always waiting for the lens and not the other way around. Electronics operate in nanoseconds, lens mechanisms don't. This means that any delay in achieving focus must be down almost entirely to the lens. The only other issue here which I can see would put a question mark over the body is if the focusing mechanism in the lens is being operated by a motor in the camera body. In this case the body is still playing a part in the focusing action and the speed at which the motor can drive the lens will obviously be a contributing factor to the speed at which lens achieves focus.

gofer wrote in post #4328645
I would have thought the electronics involved in focusing would react as near instantaneously as makes no difference and that any delay in achieving actual focus would be attributed entirely to the lens.

Hmmm. Nope. Same lenses on my 1D II focus a helluva lot faster than they do on my 10D, regardless of the light levels...even my manual focus lenses.

If what you're saying is true, there wouldn't be a reason for a 1D series other than the build.

Can't say exact percentages. But the lens is clearly more at play than the body, but how much more I don't know. If the body was the main factor, then my slowest focusing lens should be as fast as my fastest focusing lens - if you follow my drift.

But my slowest focusing lenses focus faster on the 40D than on the 10D. The fastest focusing lens - well, it focuses too quickly for me to spot the difference