Patent - Canon Tilt-Shift Adapter for EF Lenses

[keithcooper]

...
And I am a heretic. I want TS to use for interview effects (portraiture I guess) and fake minature. Burn him, burn the witch.

Ah, that might be why I've never addressed it in any of the articles I've written about tilt and shift ;-)

But seriously, what is it you feel would be useful?

I'd suggest that understanding how tilt moves the focal plane around, through a combination of tilt and focus setting is all you actually need to know? (that and practice ;-)

 

[Tinky]

Things like how to shift focus according to tilt for max effect, typical scenarios with typical settings, an emphaisis on a little does a lot, folk talk about a wedge of focus, a profile diagram of what this means, and importantly, what is the counter effect...

Even a 101 would have helped me lots and lots back in the day... tilt into the plane to increase, tilt away to decrease etc, what will happen to near and distant objects on that plane..

In those days the only guide I had was 2 pages of a kodak encyclopedia, which was great for shift and shooting round objects, not so useful for playing around with dof..

Digital changed it all for me, especially live view. My 24 was quite subtle and the risk was on film that i over tilted...

 

[keithcooper]

Ah, there would be my problem.

"...shift focus according to tilt for max effect, typical scenarios with typical settings"

This doesn't mean that much for me, since I don't know what the 'effect' is (or why/what you would want) and hence have no idea of what settings would be 'typical'.

It still comes down to understanding where you want the plane of focus to go. That just needs one distance measuring (or estimating) and looking at a tilt table - something you quickly remember if you do it very often.

Unfortunately I've discovered that even a simple list of numbers scares off a significant number of photographers, that and visualising a plane of focus in free space, for some people, is not the trivial matter I assumed.

I'm busy with a lot of work on the Northlight site at the moment, but I'll add another tilt article to the 'to do' list ;-)

 

[Tinky]

It could be as simple as...

"For greater depth of field when the desired focal plane is -for example- to your left, tilt to the left, very slightly at first, you might find A-DEP mode is helpful or at least select the largest AF pattern and tilt half checking the shutter until the greatest number of points light up, use live view to make focus adjustments, if the near distance or far distance isn't too far out, if so, tilt less or stop down more if you can, modern cameras can go up to surprising isos to give you that extra depth of field without having to decrease the shutter speed, be careful to check the right side for artefacts such as unusual bokeh shapes of loss of focus over detail, over titling will reduce depth of field, probably more than not having a tilt to work with at all.

If you actually desire the minumum depth of field, then use an opposite tilt, if your camera is pointing down then tilt up, the more you tilt the greater the effect but remember to go through your frame carefully, as unintended near or distant objects may fall into the plane or wedge of focus, negating the intended effect, the usual rules of wider aperture, closer focus distance all apply, shift can also be deployed where available to position the sweet spot"

And key phrase:

"On 135 and APS-C based digital formats there is an element of trial and error."

No numbers, and half a clue.

 

[keithcooper]

Sorry - just not a way I'd look to use such lenses at all ... I suspect our definitions of simple probably vary ;-)

For me it starts from knowing where you want the plane of focus, and setting the lens to do what you want. The numbers get things in the right area, and you can tweak stuff to look good. I even do this using such lenses hand held, so there needn't be any long deliberation over it (my own minor heresy)

The numbers can save a lot of the random (unpredictable) settings and adjustment I all too often seen when people are trying out such lenses. Almost if you keep twiddling the lens adjustments, you'll eventually hit on something that (may) work.

It's this seeming randomness that seems to put a lot of people off using such kit. Then again it does ensure a good market for used lenses, as people give up on them.

Guess that's why I'm foremost an architectural/industrial photographer

 

[dolina]

Looking forward to the full frame mirrorless body to match this adapter.

 

[keithcooper]

Looking forward to the full frame mirrorless body to match this adapter.

Indeed, but unless a lot of EF lenses magically get enlarged image circles, its utility may be limited for a full frame sensor...

 

[Tinky]

Sorry - just not a way I'd look to use such lenses at all ... I suspect our definitions of simple probably vary ;-)

For me it starts from knowing where you want the plane of focus, and setting the lens to do what you want. The numbers get things in the right area, and you can tweak stuff to look good. I even do this using such lenses hand held, so there needn't be any long deliberation over it (my own minor heresy)

The numbers can save a lot of the random (unpredictable) settings and adjustment I all too often seen when people are trying out such lenses. Almost if you keep twiddling the lens adjustments, you'll eventually hit on something that (may) work.

It's this seeming randomness that seems to put a lot of people off using such kit. Then again it does ensure a good market for used lenses, as people give up on them.

Guess that's why I'm foremost an architectural/industrial photographer

Sorry Keith, I'm a bit confused, you say you are wary of writing a guide using specific numbers, then criticise me for providing such a form of words?

All the best mate.

 

[dolina]

Looking forward to the full frame mirrorless body to match this adapter.

Indeed, but unless a lot of EF lenses magically get enlarged image circles, its utility may be limited for a full frame sensor...

Good point.

 

[keithcooper]

Sorry - just not a way I'd look to use such lenses at all ... I suspect our definitions of simple probably vary ;-)

For me it starts from knowing where you want the plane of focus, and setting the lens to do what you want. The numbers get things in the right area, and you can tweak stuff to look good. I even do this using such lenses hand held, so there needn't be any long deliberation over it (my own minor heresy)

The numbers can save a lot of the random (unpredictable) settings and adjustment I all too often seen when people are trying out such lenses. Almost if you keep twiddling the lens adjustments, you'll eventually hit on something that (may) work.

It's this seeming randomness that seems to put a lot of people off using such kit. Then again it does ensure a good market for used lenses, as people give up on them.

Guess that's why I'm foremost an architectural/industrial photographer

Sorry Keith, I'm a bit confused, you say you are wary of writing a guide using specific numbers, then criticise me for providing such a form of words?

All the best mate.

Sorry, it seems we're at cross purposes - I've already written guides with numbers, that's what I use.

If I was wary of writing something, it would be -without- the numbers...
That's why I wondered what you'd want?

 

[Tinky]

Unfortunately I've discovered that even a simple list of numbers scares off a significant number of photographers,

I picked that bit up wrong, apologies, I took it to mean that you wouldn't use numbers period.

I don't know how useful numbers would be for the tilt function as the planes could be at any angle from the camera, each scenario is potentially unique. I've always found shift fairly straight forward if you excuse the pun, one good bit of advice I was given was to slightly undershift and let a tiny ickle little bit of perspective in, it stops the top of the structure appearing too un-natural.

 

[alliumnsk]

Re: Sensor movement

If you consider how few people understand (or want to know) how to make effective use of T/S lenses

That is because T/S lenses are fully manual heavy expensive cludges. When everything done via automatic it'll save users from complex manual operations and even can function in "green" modes.

 

[keithcooper]

Re: Sensor movement

If you consider how few people understand (or want to know) how to make effective use of T/S lenses

That is because T/S lenses are fully manual heavy expensive cludges. When everything done via automatic it'll save users from complex manual operations and even can function in "green" modes.

Good luck with any user interface on that one
Never underestimate the ineptitude and lack of understanding of your audience when designing 'easy' interfaces for complex kit.

Light field and computational imaging will eventually be far more flexible for any 'Auto' modes, and not need the complexities (and limitations) of moving tilting elements.

Even then I suspect I'll still be enjoying my "fully manual heavy expensive kludges"

 

[alliumnsk]

Re: Sensor movement

Good luck with any user interface on that one
Never underestimate the ineptitude and lack of understanding of your audience when designing 'easy' interfaces for complex kit.

Well there is already interface called A-DEP. I don't see why it can't be automatic in green modes.

Light field and computational imaging will eventually be far more flexible for any 'Auto' modes, and not need the complexities (and limitations) of moving tilting elements.

Yes. Someday. Now there's very heavy pixel tradeoff (if you need 4x depth you need 16x number of pixels) and exit pupil of the lens has to be coupled with the sensor which makes it er... limited to fixed lens cameras.
Sensor tilting isn't more complex than double reflex mirror.

Even then I suspect I'll still be enjoying my "fully manual heavy expensive kludges"

Why not. Some people still enjoy shooting film and manual lenses when there's digital and AF.