SAL70400G

[stevecim]

I got a new filter today. They have exchanged it without any problems. They also ran some test, and it was remarkable to see the difference. The problem is not that obvious shooting subjects 5-7m, but going futher 10-20m the results are shocking!

At least I know now how to test focus, back focus and I have met the head technician of Sony South Africa. I guess it was all worth it.

Thanks for the help.

Can you please post some 100% crops , I would love to see the differance?

Cheers Steve

[Jonathan K]

Just to clarify matters - unless a lens is f/2.8 or faster, and activates the central f/2.8 sensor, all the AF sensors in all current Sony models read the lens as if it had an aperture of f/7.1 (the only exception is the 500mm f/8 which has a special code to enable the central sensor only, and over-ride this).

So if you have an f/4, f/5.6, f/6.3 it makes no difference at all to the AF in terms of light level for a given condition. They all are seen by the AF sensors as if you had an f/7.1 lens fitted. Other things such as focal length and exit pupil size have subtle effects on how the AF works, but whatever the case is with the 70-400mm, any change in focusing ability as you zoom to 400mm is not down to the aperture change. It is probably caused by the focal length change.

Better light will still help of course, but helps all lenses equally (and those which are f/2.8 or faster, even more). Better light often also means better contrast and that can aid focusing.

David

Thanks for the info, David. Maybe I haven't been too clear in my post, but I never wanted to suggest that there was a relation between aperture and the phenomenon I experience, but rather the focal length. There is a very quick and abrupt drop in focus speed, once you reach the 300mm.

Since there are other, more experienced users who notice the same thing, (thanks to all for your reply), I suppose I can rule out that there is something wrong with this particular sample.

So what I would like to know: Why does that happen? What influence does the focal lenght have on the way the AF-sensor reads the information? (since the information remains roughly the same... between 290mm and 310mm for example)...

Today I believe I have the opportunity to take some action shots... there is a boat race down the river in town.I will see if I can get a few "test shots".

Thanks, Jonathan

[David Kilpatrick]

At 300mm-400mm the depth of field is so limited that fast AF may overshoot the system's ability to identify the sharp image and halt the focus action.

This is what happens with the Sigma 70-200mm HSM if you use a Teleplus converter - the converter does not slow the focus down, and the camera simply can not focus. It goes wild, even the human eye looking through the finder can't spot the split-second moment when a subject is sharp, as the lens flies backwards and forwards hunting for a lock.

The Sigma converter correctly slows the focus down, the same way the SAL 1.4X and 2X converters do with the Sony lens, so that the camera can operate.

I imagine the 70-400mm is equipped with a chip that commands the AF motor to operate more slowly past a certain focal length, in order to ensure that the AF system of the camera is able to get a focus lock.

David

[Jonathan K]

Hello David,
this makes sense.
Is there a reason for slowing the focus down more one way (1.5 to infinity) than the other way (infintiy to 1.5)?
Thanks a lot for your qualified information, as always.

Jonathan

[Winston]

Hello David,
this makes sense.
Is there a reason for slowing the focus down more one way (1.5 to infinity) than the other way (infintiy to 1.5)?
Thanks a lot for your qualified information, as always.

Jonathan

I don't believe it's slower. When the focus sytem doesn't know what to do, it it focuses closer until it hits the limit and then reverses. Turning on the limiter will reduce the effect considerably. As I mentioned before, I routinely prefocus at infinity.

If I was the designer, I would make the default direction selectable.

[David Kilpatrick]

I notice differences in focus speed depending on direction of travel with all lenses, I guess it depends on the way the helicoids are designed and also - as with zooming - on the orientation of the lens. There may be a difference in close/far focus direction speeds when aiming any lens up or down, because the motor is either opposed or assisted by gravity. If you observed an extreme difference, apparently not affected by gravity, it would be worth checking against another similar lens to see that the behaviour was normal.

David

[Winston]

I notice differences in focus speed depending on direction of travel with all lenses, I guess it depends on the way the helicoids are designed and also - as with zooming - on the orientation of the lens. There may be a difference in close/far focus direction speeds when aiming any lens up or down, because the motor is either opposed or assisted by gravity. If you observed an extreme difference, apparently not affected by gravity, it would be worth checking against another similar lens to see that the behaviour was normal.

David

I just ran a brief test on my 70-400G. I did it on my deck so that I could focus on something straight down or straight up at about the same distance. With a prefocus at infinity, the time to focus lock was the same pointing up or down. I repeated the test except that I prefocused at the near limit. Again, no difference in time to lock.

I suppose there might be a some small difference but I couldn't detect it...certainly nothing to complain about.

I have to leave for a while. When I get back, I'll try the same on my old Minolta 100-400 APO. I suspect I will see a considerable difference with that one.

I'm back.

Repeating the above test on the 100-400 APO yields the same results...no difference up vs. down.

One notable difference is that the APO achieves focus lock in half the time that the G needs for large changes in subject distance.

Both lenses have a much higher slew rate at short focal lengths than longer focal lengths.

In the real worlds of birding and near-focus close-up photography, the G runs circles around the APO. The G's reject rate for focus and camera-motion errors is a small fraction of what I was experiencing with the APO.

[Jonathan K]

Hi Winston and David,

yesterday I tried out the lens on a boat race... I had much less difficulties to get a lock, even on moving objects.

So my take on this:
I noticed the same difference in focus speed with my old 100-300 APO when it was supposed to work in bad light. (you could even hear the pitch of the AF-motor noise changing)
The focus speed difference is only noticeable when I work in bad light. Around 300-400 mm the problem appears even when the light is okayish. In good conditions with clear contrasts and easy-to-focus objects there are no such problems.

So I still believe it has to do with the ability of the AF-system to lock the focus and not with a mechanical problem. The factor of ambient lighting is crucial here.
I just don't get why there is such a sudden drop in focus speed (again, in good lighting conditions it won't happen) exactly around 300 mm... So this is more intuitive: What I have considered to be enough light to work for such a lens, seems to work up to 300mm. Beyond that, there is something changing.... It could very well be that David's assumption that the DOF becomes that shallow and makes it more difficult for the focus to lock is correct...

David, I actually held another lens in the shop, which worked the same way. (if not worse). So I suppose that there is noting wrong with the lens and that I will have to learn to work with it.

The results I got yesterday are pretty stunning to my eye. I am leaving for a few days, after that I can post some pictures of the boat race I shop yesterday.

Winston: I don't really understand your explanantion here (Professionally, I am supposed to be a-techinal.. ): As far as I understood, the phase-detect AF-system is supposed to be reading the direction in which it should be adjusting? So why would your trick (prefocusing to infinity, a trick which I tried and works well) be necessary?

The problem with me (and many others) fearing there is someting wrong with the equipment is that you have to know what to expect... If such a lens (which looks like quite a technical achievement to me) forbids faster handling because of its construction and this kind of behaviour is inherent to its construction, than everything will be ok. I just have to know it.
The lens is great anyway, no question here.

Cheers

Jonathan

[Winston]

Winston: I don't really understand your explanantion here (Professionally, I am supposed to be a-techinal.. ): As far as I understood, the phase-detect AF-system is supposed to be reading the direction in which it should be adjusting? So why would your trick (prefocusing to infinity, a trick which I tried and works well) be necessary?

When out of focus at long focal lengths there is no information until it gets pretty close. Look in the view finder as you manual rack the focus in and out and remember that the focus sensor is only seeing a very small portion of what you see. There may be some information there but it isn't reliable. If the system tried to use it, hunting would render the lens useless.

The problem with me (and many others) fearing there is someting wrong with the equipment is that you have to know what to expect... If such a lens (which looks like quite a technical achievement to me) forbids faster handling because of its construction and this kind of behaviour is inherent to its construction, than everything will be ok. I just have to know it.
The lens is great anyway, no question here.

Thank David profusely for this forum

[David Kilpatrick]

I have test the lens. The lens was sharp. Then I have put the filter back on, and the photos ware not sharp at all.

The filter is a pro-digital Kenko UV filter. All the trouble, while meantime it is something as obvious as a filter.

Is it possible that the filter has a defect? I foought that I bought a good filter.

Hennie

I just bought a Kenko 'Digital' (not labelled Pro Digital) 77mm. It only cost £10, which is ridiculously cheap, including postage from a UK eBay vendor and is brand new and sealed, was delivered in 24 hours. Observations: this 'Digital' Kenko is either uncoated, or single coated. The glass is of a thickness where if you hold the filter a foot in front of your eye, and then move it round and tilt it, you see whatever is behind it move. That's not good sign, and may indicate a lack of true plane parallelism. The filter though sealed and new had a distinct film on the glass, invisible at a casual glance, like an atmospheric deposit (similar to the layer left on a car windshield after a few weeks). This would not wipe off dry, and I had to apply lens cleaner to remove it.

Without this cleaning process, I think the Kenko filter would have diffused fine contrast. Why was it SO cheap - and clearly genuine? I would have to suspect this is stock from a rescue, possibly an insurance recovery. The nearly invisible film on the glass is the kind of thing which happens if there has been a fire and 'undamaged' pallets are rescued. For the moment, I've stuck it on the 70-200mm but even on visual examination it looks like a low grade filter glass.

David