[MichalW]

Just read about new Olympus E-M5 Mark II. And I find this feature very interesting: "it can also record 40 MP still shots by shifting the sensor and creating a composite image" And since Pentax cameras can shift their sensor too... Why not incorporate this feature into Pentax bodies? Possibly by only firmware upgrade? Imagine Pentax K-5 IIs or K-3 able to produce 40 mp file when you need it!!

Of course, I'm aware of the fact that Ricoh will never incorporate that in the already discontinued K-5 or discounted K-3 but still.. Sounds like a good idea to me

What do you think? Possible?

[Adam]

Just read about new Olympus E-M5 Mark II. And I find this feature very interesting: "it can also record 40 MP still shots by shifting the sensor and creating a composite image" And since Pentax cameras can shift their sensor too... Why not incorporate this feature into Pentax bodies? Possibly by only firmware upgrade? Imagine Pentax K-5 IIs or K-3 able to produce 40 mp file when you need it!!

Of course, I'm aware of the fact that Ricoh will never incorporate that in the already discontinued K-5 or discounted K-3 but still.. Sounds like a good idea to me

What do you think? Possible?

Definitely not via firmware, but I it would certainly be a possibility, especially considering that Pentax already has the AA filter simulator feature.

[MichalW]

AA simulator yeas, but also gps tracker(O-GPS1) controlled sensor movement on K-5 for instance. I know I'm simplifying things here but this sounds like a good option to have.

[rrstuff]

Definitely not via firmware, but I it would certainly be a possibility, especially considering that Pentax already has the AA filter simulator feature.

I don't know if AA simulator is controlled for speed, or is it just a vibration at a certain frequency.

And why it couldn't be a firmware update, out of curiosity?

It should be possible to employ sensor shift to do this kind of improved resolution image. Sensor shift has to compensate for camera motion within a sub-pixel accuracy. I see no reason why it couldn't be done in software.

[Clavius]

Considering the fact that Pentax has been looking for more ways to exploit their SR mechanism. Evidence being the AA filter simulator. I doubt if they haven't looked into this option for using the SR for "superresolution" yet. Maybe the Pentax SR implementation can't be controlled precisely enough yet for this purpose?

[rrstuff]

Considering the fact that Pentax has been looking for more ways to exploit their SR mechanism. Evidence being the AA filter simulator. I doubt if they haven't looked into this option for using the SR for "superresolution" yet. Maybe the Pentax SR implementation can't be controlled precisely enough yet for this purpose?

That was my first though, but in order to get a sharp image during camera motions, the camera has to record and then shift the sensor within at most one pixel, or less, to get a sharp image. The only thing preventing them from using it that I can think of would be processing power. Not that I can think of all the possibilities, of course. There might be some more limitations.

[MichalW]

Considering the fact that Pentax has been looking for more ways to exploit their SR mechanism. Evidence being the AA filter simulator. I doubt if they haven't looked into this option for using the SR for "superresolution" yet. Maybe the Pentax SR implementation can't be controlled precisely enough yet for this purpose?

Well, I'm no expert but stars sensor shifted tracking via gps unit is very precise in my opinion, maybe there are other obstacles we are not aware of..

[rrstuff]

Well, there is always a motorized stage with a precise spatial control, like this one: Dimension Icon - Overview | Bruker Corporation

This thing is more than capable of moving a camera at precisely one pixel in each direction . One can then assemble everything in post.

[Clavius]

That was my first though, but in order to get a sharp image during camera motions, the camera has to record and then shift the sensor within at most one pixel, or less, to get a sharp image. The only thing preventing them from using it that I can think of would be processing power. Not that I can think of all the possibilities, of course. There might be some more limitations.

Well, I'm no expert but stars sensor shifted tracking via gps unit is very precise in my opinion, maybe there are other obstacles we are not aware of..

True... Maybe there is no technical limitation, but more a legal limitation perhaps? Or... Marketing wise, it's better to spread such surprises. The K-3 got the AA filter simulator. The next one gets some more big news of its own.

[Hamiltom]

I believe this is possible using "composition adjustment".

Take 1 image centered, then 9 images while tilting the sensor to maximum in all directions and merging the images.

Here the video explaining it:

You Tube

You Tube

[btnapa]

Just read about new Olympus E-M5 Mark II. And I find this feature very interesting: "it can also record 40 MP still shots by shifting the sensor and creating a composite image" And since Pentax cameras can shift their sensor too... Why not incorporate this feature into Pentax bodies? Possibly by only firmware upgrade? Imagine Pentax K-5 IIs or K-3 able to produce 40 mp file when you need it!!

Of course, I'm aware of the fact that Ricoh will never incorporate that in the already discontinued K-5 or discounted K-3 but still.. Sounds like a good idea to me

What do you think? Possible?

The technology is not new and exclusive to Olympus. Back in the late 90's early 2000, Imacon offered a similar technology to boost resolution for a medium format back. If my memory serves me well, they took a 16MP image and shifted it 6 times to create a 96MP image. The target market was mostly studio still product shooters.

The thought is intriguing as I am interested in such a solution for landscape work. This however might put a dent in the 645Z business which I am sure Ricoh is not interested in.

I believe, currently there is a Hassy based solution too offering a 200MP resolution. The technology is there. My guess is that technically Pentax can do it but whether it make economic sense, or it fits their business strategy, I am not sure.

[VisualDarkness]

Considering that Hasselblad already do it and now Olympus (looks like patents are loose enough then) I scratch my head over why Pentax haven't done it yet.
Hasselblad unveils pixel-shifting 200MP H5D-200c MS: Digital Photography Review

[BrianR]

At least one earlier thread about this: https://www.pentaxforums.com/forums/6-pentax-dslr-discussion/281225-olympus-w...ml#post3057690

I'd be happy if they added a 'one-touch composition adjustment bracketing' where it walks the sensor around some user programmable route taking a shot each time, to be combined later in post by the user. Not useful for ever shooter, but some would find it very handy.

There's tons of info out there, look up 'superresolution' and also the photoacute software.

[Edgar_in_Indy]

Ricoh has already offered a similar feature in a consumer camera in the early days of digital cameras. The model slips my mind at the moment.

EDIT:

The Ricoh camera I was thinking of was the 3.3MP RDC-7 from 2000. This is from their marketing brochure:

"There are three Pro Mode options-all based on Ricoh's innovative Image Enhancement Technology. Pro-L mode shifts the CCD by one pixel to take two shots that a Ricoh-developed algorithm composes into one image. This boosts resolution and definition by 20% without increasing image size.

The default Pro Mode uses a Ricoh-developed interpolation algorithm that eliminates "jaggies" when the output resolution increases to 7 megapixels. The resulting image size is 3,072 x 2,304 pixels. Pro-H mode raises output resolution to a maximum 7 megapixels by simultaneously incorporating both the default Pro and Pro-L modes."

It seems like the pixel-shifting may have been successful, to a degree, in getting more resolution. However, it had a lot of limitations. The camera had to be on a tripod, the subject had to be stationary, and it took 35-seconds to 2-minutes to process each shot. Seems like the same kind of basic idea though.

[interested_observer]

I've been asking this question for a couple of years now. Technically the capability is there, just produce the individual EXIF tagged frames, and post process everything together, thus a firmware update could easily take care of it. It has to be Pentax's world famous marketing/management gurus putting the kibosh on it. It's not very likely that I will be running out any time soon to buy a 645Z with a 24-85 lens - a relative bargain at $13,000.

While were on the subject, they could also have a second mode to stack each of the RGB colors at each pixel location - similar to a Floron sensor.

This would be a very nice addition for landscapers.