Patent: Electronic control for tilt-shift lenses

[Antono Refa]

An AF engine has to move a group of lens elements within the tube of the lens

A tilt engine has
- to tilt the whole tube with all lens elements.
- to rotate the whole mechanism to get the tilt orientation needed.

Frankly, I hardly can imagine this to happen (robustness).

And the user interface dealing with at least three independent focus points seems a challenge for me as well. (This is to be done by the body, not the lens)

I think there are two issues here.

The first is whether Canon will implement this patent, e.g. because it wouldn't be robust enough. If I had to bet $10, I would be on your side.

The second is, if Canon implements this patent, how much power it would consume.

 

[Del Paso]

If you have a look at the pictures in the article, you will see only a single lens is moved.

This doesn't look like a wide angle lens. The front element is in a housing which cut of the edges of wider lenses.
Probably this patent is aimed at unique portrait lenses or something different. These use cases would also more profit from electronic control. A landscape or real estate shooter is able to spent some seconds in adjusting the lens. They also are more interested in the shift function which this patent seems to lag.

You are right! Edit: you could (!) be right...
I was so excited I assumed this post was about the long-expected 14mm TS... so, I didn't even pay attention to the pictures.
Yet, if I think this over, this patent is just a theoretical representation of a concept. It still could become a 14mm TS reality.
Or, as you presume, a portrait lens.

 

[Nemorino]

Yes and I am just thinking.
Next thought: could this be realized in a shift lens or just a tilt only lens. I cannot imagine a wide angle with tilt but without shift.

 

[Alam]

This would be one hell of an expensive lens if it had that capability.
Personally I don't think it would be worth it.
I think one of the things that makes the TS-E great is that it is manual focus so you are far more precise in what you are doing as you are checking focus.
Motorised tilt is interesting but I wonder really. It's a function I don't use much. The current controls are not precise, hard to make minor adjustments.
I wonder would a motorised version be much better. It's like those motorised lens you can get for micro 4/3 or superzoom bridge cameras, they tend not to be that controlable. You'd want to be able to make micro adjustments.
Sometimes adding technology doesn't make it better.

Depends, imagine underwater photography with ts lens, there's no way you can do it with manual lens

 

[Hector1970]

Depends, imagine underwater photography with ts lens, there's no way you can do it with manual lens

Interesting, would never have thought of using one underwater .

 

[Hector1970]

It would, in my understanding, not just be a motorized tilt function, but a combination of tilt plus auto AF.
This would render obsolete the complex and boring calculation of focus distribution. Also ideal for handheld photography!

Yes both and additionally accurate exif data. It will be interesting. I don’t use tilt much. It’s tricky to finely adjust. The current knob is not very precise. The final product will be interesting and expensive. I’d highly recommend the 24 TS-E II for anyone considering it. Using it makes you a better photographer. You have to think when using it.

 

[InchMetric]

There is a "virtual tilt" alternative to tilt shift that can be implemented in firmware for existing bodies and lenses for relatively still subjects (like tilt shift is typically used for).

It might be described as "dual subject autofocus" or "multi-subject auto focus." It simply involves shooting rapid sequential images (at 20-30 FPS electronic shutter presumably) with each identified subject in focus. The image is the composited to put both subjects in focus. One typical application would be a portrait of two people at different focal differences. Getting both eyes on focus for an angled subject is another application. Face and goods for a model displaying a product such as jewelry.

This and related concepts are disclosed and claimed in US Patent Number 11,283,989, which was granted last month. Coincidentally, I'm the inventor and just received the printed copy of the patent in the mail yesterday.

Other disclosed concepts include applying this to focus stacking systems to provide focus for intermediate subjects between the main selected subjects.
For shooting angled planar subjects like building facades this does not provide the perfect plane of focus of a tilt shift. But when the tilt is used simply to get a couple important subjects into simultaneous focus, it's a suitable solution. And of course, it's a firmware implementation that doesn't require a new lens. In addition, it enables capturing more two subjects that are not in a common plane with each other.

As a multiple-patent inventor with experience sharing this sort of thing, to save time you may use the following response template codes if you wish:
a. "Cool, congratulations."
b. "Patents suck, and here's why based on my experience..."
c. "I didn't read the patent, but I'm sure it won't work."
d. "I don't like the idea and I would never use it."

My other camera-related patents (granted - more pending) are:
Controlling added digital zoom (cropping) using the optical zoom ring on a conventional lens - US Patent 10,868,965.
Lens mount adapter with latch to prevent detachment from lens (semi-dedicated) - US Patent 11,212,451.

 

[Del Paso]

Yes both and additionally accurate exif data. It will be interesting. I don’t use tilt much. It’s tricky to finely adjust. The current knob is not very precise. The final product will be interesting and expensive. I’d highly recommend the 24 TS-E II for anyone considering it. Using it makes you a better photographer. You have to think when using it.

24 TSE : my third most often used lens, after Summilux 1,4/35 and EF 100-400 L II when travelling (cities, nature etc...)

 

[Del Paso]

If these lenses are fully electronic this way, I'm guessing $5k USD.

Brian

Just look at Nikon's 19mm TS lens to get an idea...$$$$$$$
In French I'd say it will cost "la peau des fesses" (the skin of one's a*s)

 

[neuroanatomist]

Just look at Nikon's 19mm TS lens to get an idea...$$$$$$$
In French I'd say it will cost "la peau des fesses" (the skin of one's a*s)

Ouch, that makes Canon's TS-E 17 look like a bargain!

 

[Del Paso]

Ouch, that makes Canon's TS-E 17 look like a bargain!

In Europe, Nikon's TS 19mm costs as much as Canon's TSE 17 + TSE 24...
But is said to be optically fantastic (Lens Rentals).
And, for what it is, basically a MF ultra wide-angle, the TSE 17mm is indeed a bargain!

 

[pzyber]

Did you ever use any TS lens?
There is a difference between
- tilting (the topic of the patent)
- and shifting (your not so wise proposal)

No, a 10-500 zoom will not do the job.

It was a sarcastic comment aimed at how Canon have designed the latest lenses with very high distortion and corrects it with software. I'll make that more clear in the future since not everyone seem to understand sarcasm

And yes I own and use TS-lenses regularly.

 

[cayenne]

I wonder, would these also still be capable of manual control of focus, tilt, etc...?

If not, then it would kill the lens' usefulness on any other cameras.

I know I find a lot of pleasure using full manual lenses adapted for use on a variety of mirrorless cameras.

If it is manual, you can use it on any camera which is nice....and to me, makes it more valuable to use across my camera stable.

Just a thought.

cayenne

 

[stevelee]

The current controls are not precise, hard to make minor adjustments.
I wonder would a motorised version be much better. It's like those motorised lens you can get for micro 4/3 or superzoom bridge cameras, they tend not to be that controlable. You'd want to be able to make micro adjustments.
Sometimes adding technology doesn't make it better.

My fear would be that the precision might have a feel a bit like manual focusing on the G5X and G7X cameras.

I can't justify the expense for the current TS-E lenses, so I shouldn't worry about the motorized ones.

 

[Del Paso]

My fear would be that the precision might have a feel a bit like manual focusing on the G5X and G7X cameras.

I can't justify the expense for the current TS-E lenses, so I shouldn't worry about the motorized ones.

It's a case of conscience.
Mine tells me I have to buy the TSE 14mm, even if I must sell my car.

 

[becceric]

It's a case of conscience.
Mine tells me I have to buy the TSE 14mm, even if I must sell my car.

I understand. I just bought the TS-E 17mm. If it wasn’t for the crazy values on used cars, my Honda would be worth less than that lens.

 

[gdanmitchell]

I hate to imagine the cost of such a special-purpose beast of a lens!

However, there are lots of interesting possibilities here, too. For example, if the operation of the tilt/shift mechanism (and, of course, focus) are electronically controlled, the camera will also be getting information about these settings and that information could be used to inform image correction. I can imagine things like graduated CA correction that recognized how CA would vary across an image from a TS'ed lens, for example.

But the cost! ;-)

 

[stevelee]

I hate to imagine the cost of such a special-purpose beast of a lens!

However, there are lots of interesting possibilities here, too. For example, if the operation of the tilt/shift mechanism (and, of course, focus) are electronically controlled, the camera will also be getting information about these settings and that information could be used to inform image correction. I can imagine things like graduated CA correction that recognized how CA would vary across an image from a TS'ed lens, for example.

But the cost! ;-)

One advantage might be that the tilt and shift info would be included in metadata and perhaps focus data.