Is a 150mp Canon EOS R camera on the way? [CR1]

[KeithBreazeal]

No need to stitch frames for panoramas.

 

[Berowne]

Are full frame lenses really resolving much of anything at that pixel density?

Yes: More Ultra High-Resolution MTF Experiments

 

[wtlloyd]

Same with me. You will have to carry a complete workstation with you to get the images post-processed. Plus, DLA (diffraction limited aperture) would have to be f/3.5 or something like that I guess, so you could forget to shoot e.g. landscapes the classic way (f/16 or 22) and get substantially more real image information out of such a camera than out of a, say, a 5D3. Plus, you would have to shoot even moderately moving subjects with 3x higher (am I right?) shutter speeds than with a 50 MP camera to freeze them sharply on the pixel level. I think if Earth moves closer to the sun, Canon should bring out such a monster

This is what I had in mind, although I limited my comment to the hand-held aspect of 35mm format. This sensor could be good in the studio, but anything that moves (and that includes landscape!) would require very high shutter speed.

 

[felipeolveram]

RELEASE THE R5 ALREADY

 

[dtaylor]

Not a practical resolution in a hand held 35mm format body. The technique demanded to get the most of that pixel size, as anyone knows who has shot with a 5DS/DSr...

My technique for shooting handheld with a 5Ds is simply "don't adjust your Tv expectations from crop" since pixel density is similar to the 7D. If I'm shooting without IS then I still think of 1/(1.6 x focal length) as my minimum shutter. Of course that's with the 24-70 as everything else has IS. And any future high end R bodies will no doubt have IBIS.

Even I think 150mp would be a bit too much for a 35mm sensor size. But you wouldn't be at a disadvantage with it and there are lenses/situations where you could exploit the full resolution. I would be pretty happy with 75-100mp as I'm pretty happy with 50mp now.

As others have pointed out, this rumor is probably due to someone misunderstanding a number for a 75mp DPAF sensor.

 

[Mikehit]

so you could forget to shoot e.g. landscapes the classic way (f/16 or 22)

Why?

 

[dtaylor]

Plus, DLA (diffraction limited aperture) would have to be f/3.5 or something like that I guess, so you could forget to shoot e.g. landscapes the classic way (f/16 or 22) and get substantially more real image information out of such a camera than out of a, say, a 5D3.

You get more real image information out of a 50mp sensor than a 22mp one even at f/16. Resolution and diffraction "limits" do not work they way people generally believe they do. This is an artifact of the terminology we use, i.e. "limit" and "limited."

The final resolution of any imaging system is going to be less than the weakest component. But improving any component improves the final result, bringing it closer to the theoretical best of the weakest component.

Put another way, a 50mp sensor at f/16 is not as sharp or detailed as the same sensor at f/5.6. But it's still sharper than a 22mp sensor at f/16. Both cameras are "limited" by the f/16 aperture, but the higher resolution sensor still pushes the final result closer to the theoretical best it can be for f/16.

 

[KeithBreazeal]

Just think of all the chores you can get done while processing the photos. When processing a project of stacked images, you can mow the grass, go grocery shopping, or go on another photo adventure.

 

[melgross]

L lenses handle it just fine.

Canon Unveils a Monster 250-Megapixel Sensor

Canon today announced that it has created a monster of a CMOS sensor with the world's highest pixel count (for its size). The DSLR-format sensor manages

petapixel.com

I’d like to see some evidence of that. A lot of properties of this sort of photography is done in single color where it’s easier. I seem to remember Canon saying that they were naming lenses that would work with a 200mp sensor, and that would make sense.

the question is where is that sensor now? We do know that current lenses do break down as the image gets more off center. So in that one third middle portion, it might do fine, but elsewhere it will be a problem. Additionally there are problems with every camera, and every lens in regards to planar parallelism. This gets worse as resolution goes up. Don’t pretend these aren’t issues.

 

[scyrene]

Using existing cameras/pixel-densities data from TDP I would guess around f/3.5 so a f/4 big white lens would be just fine. But this would exclude the use of teleconverters (at least the 2X ones) and f/7.1 zoom lenses.

I would say a 80-82mp camera would be more practical. It would have the 90D's DLA (f/5.2)

Diffraction isn't a brick wall. If the DLA is f/4, it doesn't mean you can't use lenses slower than that. Besides, pixel densities are still much higher for smaller format sensors, and people seem to manage fine with them. It's just diminishing returns.

Sure, it’s possible, but what kind of per pixel IQ could we expect from that, and can lenses be made sharp enough for that rez? We’re hitting the point of max resolution now, as the laws of physics are telling us that we’re running right near the edge.

Well nobody is viewing single pixels, but whole (or cropped) images, so that question is rather academic. The history of digital photography surely teaches us that increasing resolution can go along with improved IQ overall, though as I say above, you get diminishing returns, so doubling the res isn't going to improve other areas anything like that much. But there is no 'point of max resolution'.

File size is the main disadvantage.

 

[Diko]

@CR, give a glass that would support that resolution. And THEN I will call it a substantial rumor

 

[masterpix]

I am not sure who needs 150MP sensor, and since pixel size will limit the sensor flexibility and therefore the dynamic range. I suppose that arthcitectes and fashion photographers will "jump" on the new sensor.

WIll the R5 will be acompanied by simialr 5Dv4?

 

[Sporgon]

I wonder what happened to the 70-80mp camera that was previously mentioned? Just a rumor I guess.

It met a 2 mp camera and bred

 

[tron]

Diffraction isn't a brick wall. If the DLA is f/4, it doesn't mean you can't use lenses slower than that. Besides, pixel densities are still much higher for smaller format sensors, and people seem to manage fine with them. It's just diminishing returns.



Well nobody is viewing single pixels, but whole (or cropped) images, so that question is rather academic. The history of digital photography surely teaches us that increasing resolution can go along with improved IQ overall, though as I say above, you get diminishing returns, so doubling the res isn't going to improve other areas anything like that much. But there is no 'point of max resolution'.

File size is the main disadvantage.

No objection that's why I mentioned 2X teleconverters and not 1.4X ones to the f/4 lenses. In fact I have used 2XIII teleconverter with a f/4 lens (500II) at f/9 with my 5DsR with success. Due to the use of 2X contrast took a hit but that's a whole different issue (in some other shots it didn't). In fact some of these 1000mm shots had given me wonderful results.

 

[joestopper]

Continue reading...

I dont think 100+ mp is unrealistic. Why?
- smart phones have 100+ already
It does not mean we have 100mp images. The pixels can be used for electronic IBIS i.e. the sensor is larger than FF and while the image through shake is projected toout-of-sensor area, the FF frame is read.
That is a better and potentially more accurate (since faster) way of IBIS. Additional benefit: No moving parts which is good for durability. This technique will come for certain (but maybe not yet in the RS)

 

[joestopper]

Probably sensor-shift technology.

No.
Here they talk about sensor size, not image size.

 

[caffetin]

Anyway, they are cooking something

 

[Canon1966]

It met a 2 mp camera and bred

That was funny!

 

[Sharlin]

@CR, give a glass that would support that resolution. And THEN I will call it a substantial rumor

The vast majority of cameras on the planet have a pixel density higher than a 150 MP small-format sensor. And even if you dismiss phone cameras, 20MP 1" sensor cameras make great images at a similar pixel density, as has already been mentioned in this thread.

 

[Dragon]

With that kind of sensor, Canon could win all the DXO Mark lens tests . Seriously, it is not a bad idea when you look at the issues with Bayer grid reconstruction. Such a sensor would dramatically reduce the occurrence of false color and other nasty artifacts arising from deBayering. As an example, when was the last time you saw Bayer artifacts on a 1/2.3 P&S sensor? Of course, many of those are over 500 MP FF density. 150 MP is essentially the same density as a 20 MP 1" sensor. With 4-way pixel shift just to cancel any remaining artifacts from the Bayer grid, this could produce some truly awesome images. No real need for the 8-way shift to try to get additional resolution. It does look like Canon is queuing up to change the game in a bunch of areas at once. I don't think we have yet seen all they are working on in sensor performance, either.