[GregK8]

I have been playing with my new K100D for a couple of weeks now, and as I get familiar with it and investigate lens alternatives, I see some comments and I don't understand something. I see comments around the net where people say a particular lens focuses faster than another and it doesn't make sense to me, so either I don't understand something, or it's a subjective observation that is not objectively measurable.

I know the difference between the SDM AF system and the AF system that uses the body motor. But what I don't understand is how two different lenses that use the body motor for AF can perform differently in AF mode. If both lenses use the body AF sensors and the body's screw motor, how can one take more or less time to focus than the other. I mean, if the body takes the measurement and controls the lens focusing motor, how can one focus faster than the other. I can understand noise levels being different because different lenses will put different loads on the body motor.

Can you folks set me straight?

[monochrome]

Pentax cameras seem to be programmed to bring the image nearly into focus, then tighten the focus in a second step. Higher lens resolution could make this second step calculation by the focusing sensor faster.

The gearing in the lenses could different, and the "throw" (distance it travels to bring the image into focus) of the helicoid could be different from one lens to the next.

[GregK8]

I understand what you are saying and that makes a lot of sense, especially if the comparison is between 2 lenses of dramatically different focal length. I guess my confusion came when reading some of the comparisons between the Sigma 70-300mm vs the Tamron 70-300mm and the Pentax K100D kit lens vs the Tamron 28-80mm, which are also very close. I'd see comments that one was faster than the other, or that one hunted more than the other.

[monochrome]

Well I don't KNOW the answer, but I think the autofocus mechanism uses software to "read " the sharpness of the image on the sensor. Logically one lens should bring the image on the sensor plane into sharp focus as fast as another, so the distinction is confusing.

I bet the gearing designs in the lenses are proprietary, and there are cost savings involved with "slower" gear trains. Since lenses have chips that communicate with the camera, I'd bet that the programming algorithm on the chips is proprietary, as well. There could also be costs and proprietary design techniques associated with reducing friction in the helicoid, so the body motor can drive a higher cost, faster gear train into focus faster then in a lower cost lens.

Not every company has access to the capital necessary to constantly develop cutting-edge designs (sound like anybody we know?)

There must be lots of compromises made to find the right combination of cost, materials and design from one lens to another.

[RiceHigh]

See this experiment / example and you will understand more (for why there are differences):-

RiceHigh's Pentax Blog: Lack of Speed of the SDM

I have been playing with my new K100D for a couple of weeks now, and as I get familiar with it and investigate lens alternatives, I see some comments and I don't understand something. I see comments around the net where people say a particular lens focuses faster than another and it doesn't make sense to me, so either I don't understand something, or it's a subjective observation that is not objectively measurable.

I know the difference between the SDM AF system and the AF system that uses the body motor. But what I don't understand is how two different lenses that use the body motor for AF can perform differently in AF mode. If both lenses use the body AF sensors and the body's screw motor, how can one take more or less time to focus than the other. I mean, if the body takes the measurement and controls the lens focusing motor, how can one focus faster than the other. I can understand noise levels being different because different lenses will put different loads on the body motor.

Can you folks set me straight?

[Maxington]

See this experiment / example and you will understand more (for why there are differences):-

RiceHigh's Pentax Blog: Lack of Speed of the SDM

Why not link to the actual thread on DPReview, not your blog post which has nothing to do with it, since you didn't conduct the experiment, nor add any additional information, you are merely trying to direct traffic to your blog.

Real link:
SDM speed (or lack thereof): Pentax SLR Talk Forum: Digital Photography Review

The experiment is flawed as is pointed out in comments in your blog, and on the DPReview thread. It was merely a test of the lens travel distance, which is not a reliable indicator in itself of AF speed.

[flyer]

Some lenses, like macro lenses, have a longer focus throw, so the camera motor has more winding to do , increasing focus time. Then, some lenses don't have a large opening (does 5.6 mean something) so the focusing sensor take longer to figure out a focusing solution. Then, you have some lenses with deficiencies that have poor contrast, so AF has a hard time finding the proper focus. Then, there is THE perfect lens (f:1, short focus throw, super high contrast (does it exist)) where the reaction time of the camera would be almost instantaneous.

[Confused]

Hi flyer

Re your observation:

Then, you have some lenses with deficiencies that have poor contrast, so AF has a hard time finding the proper focus.

I am not disagreeing with you on this one, but if that proves to be the case, then why is it that other manufacturers seem to have overcome these deficiencies in even their lowest-spec DSLR's whilst the K10D/20D still appears to struggle on certain occasions ? Just out of interest, is poor contrast possibly more prevalent in zoom lenses rather than in primes and lastly, how can Pentax's engineers rectify this unacceptable situation in their forthcoming DSLR's ?

Best regards
Richard

[flyer]

I never said deficient lenses where Pentax. You have to include third party lenses in the lot, too. And all manufacturers get some deficient lenses. Manufacturers give themselves some tolerances when assembling/manufacturing lenses. If your lenses tolerances are at the minimum, you'll have an outstanding lens whereas if the tolerances are at the maximum, you'll have a lemon. Same goes for any manufacture goods. That said, you would have to be really unlucky to have ALL tolerances to the maximum. You can have one or two, but all of them would be exceptional. Unfortunately, the Pentax 16-50 f2.8 seems to have a rather high level of deficient lenses, more than likely due to poor workmanship, because when you get a good one, it is exceptional.

[GregK8]

I guess it is more complicated than I thought. I always thought things like contrast and had more to do with available light than the characteristics of the lense itself.

[GregK8]

See this experiment / example and you will understand more (for why there are differences):-

RiceHigh's Pentax Blog: Lack of Speed of the SDM

This has nothing to do with what I am asking about. I am not talking about the comparison of SDM to the body motor as this test attempts to make a conclusion on, but on why the Sigma 70-300mm lens and the Tamron 70-300mm might have different levels of AF performance using the body motor.

Perhaps there is no real difference in lenses this close, but a perceived difference. Obviously, there is also the possibility that mechanics play a role.