Manual focus on Z1U - sometimes coarse, sometimes fine - why?

[Martin Pauly]

The title says it, I think. There's probably a really simple answer to my question, I just haven't found it yet. When I adjust the focus manually using the focus ring on my Z1U, sometimes I get very coarse increments (like jumping from 10m to 100m or something extreme like that), and at other times I can make very fine adjustments. I just haven't figured out how to go from one mode to the other...

Dumb question? Waiting to be enlightened...

- Martin

[Bob Hart]

My imaginings are that the camera's distance measuring system is optical, not ultrasound which is another method.

It can only report what it once saw in testing and development in terms of the pixel resolution and has been taught to be true by the little wizards in the R&D workshops.

I imagine that the distance reporting is based on the most extreme differences in light level between adjacent pixels which suggests to the camera the sharpest areas. This would be complicated by colours and possibly the colour wavelengths.

There will have been monitoring built in to the lens servo system to report where the lens focus parts are at a given adjustment to help humans.

When the camcorder autofocus is turned loose to do its job, it has no need to report back to the operator. Flick back to manual without touching the focus ring and the camera then tells you where it put itself last.

As with human ability to see easier the shallower depths-of-field with long lenses or zooms set at the long end, the camera is likewise limited. A deep depth-of-field makes it harder to discern what is at the sharpest focus.

The camera distance reporting is set up so that it tells you no lies, only gives you an approximation of sharpest focus positions. With deep depths-of-field, the necessary focus movements to cause discernable differences at a pixel to pixel level will be quite coarse.

I don't think it does too badly. Short of taking a tape measure out, using a timebased measuring system like ultrasound or lazer or setting up a majorly inconvenient triangulation (3D) system and using a true manual lens control, there's probably nothing better for the price.

Let better brains than mine inform you. As I said, this is just my imaginings on the subject based on a very small amount of knowledge a danger unto myself.

[Martin Pauly]

Hi Bob,

Originally Posted by Bob Hart

As with human ability to see easier the shallower depths-of-field with long lenses or zooms set at the long end, the camera is likewise limited. A deep depth-of-field makes it harder to discern what is at the sharpest focus.

The camera distance reporting is set up so that it tells you no lies, only gives you an approximation of sharpest focus positions. With deep depths-of-field, the necessary focus movements to cause discernable differences at a pixel to pixel level will be quite coarse.

If I understand you correctly, you are saying that the course/fine switch occurs depending on the depth-of-field, so I should see a change when I open or close the iris - or did I misunderstand your post? I will give that a try!

Thanks,
Martin

[Bob Hart]

My imagining is that opening the iris will be unlikely to alter the precision of the distance display as I don't think the readout will be directly linked to what the sensor image processor determines.

Opening the iris however may enable the autofocus to work better and the manual focus to be more accurately set by the operator before iris closure back to the preferred setting as is practiced with film cameras.

I think the readout may be more likely simply reporting the position of lens components directly as best it can.

This may be generated by the camera from a system for counting pulses to a stepper motor or brushless DC motor from a zero-set reference point or maybe from a sensor in the lens system itself which measures the mechanical movment.

If the actual handling characteristics of the camera are any guide, with the lens on wide-angle, the travel of focussing movements will be very small. At the full zoom-in end, the travel of focussing movements will be much longer.

If some sort of step counter is being used, then more increments of readout can occur if a larger movement is happening.

As the head bookend of my comment states - my imaginings of how Sony might have gone about it, not proven fact.

It is a pity that each of the big players did not assign one of their employees to haunt these forums and get involved at the enthusiast developer level to respond to such a discussion as this.

[Chris Soucy]

Hmm, interesting thread.

Can any of you guys do this small test for me?

Camera on tripod, zoom in max (Z99 on a Canon A1) do the manual focus thing. Make a note of the jumps in notified focus distance (Start at minimum and just turn the focus wheel as slow as you can, if you have a "slow focus" setting in your menu's, turn it on)

Take the zoom out to Z89 (if the camera lets/tells you) and do it again.

Then Z79 etc.

If the camera under discussion has the same system as the Canon XH A1 (or very similar) you'll find the only way to get anywhere near smooth focus jumps using manual is to go to Z99, do the business, then zoom back out.

However, not owning one nor having access, can't confirm till one of you guys does.

This could well be where this discrepancy in manual is in fact happening.


CS

[Warren Kawamoto]

If you're wide, the lens focuses quickly. If you're telephoto, it focuses slowly.
It's that simple.

[Boyd Ostroff]

I have to agree that you're seeing a really simple phenomenon which I also see on my Z1. If you zoom all the way in the focus numbers will move in smaller increments which reflect the shallower depth of field. Zoom all the way out and the focus numbers jump in larger increments. This is because focus just isn't very critical at the wide end of the zoom, due to the huge depth of field.

[Martin Pauly]

Originally Posted by Warren Kawamoto

If you're wide, the lens focuses quickly. If you're telephoto, it focuses slowly.
It's that simple.

Thanks Warren - you are right, it really is that simple. I figured there must be a simple explanation, I just hadn't been able to figure out what it was...

- Martin