Canon's Rumored 75mp Camera To Have a Non Bayer Multilevel Sensor?

[Canon Rumors]

Tweet

Non Bayer Multilevel Sensor

[NL] has been told that the rumored 75mp image sensor Canon is testing is a “totally usable photosite count” and is a non bayer multilevel sensor. This patent for such technology appeared back in May of 2013.

Source: [NL]

cr

[Drizzt321]

So something sorta like the Foveon X3 sensor? Or are they doing some other kind of tricks?

[clicstudio]

Finally someone gets the idea of multi sensors...one for highlights and one for shadows or even a third one for mid tones...
Instant Dynamic Range heaven!
I can't wait for a camera that sees what my eyes see...
Canon has to beat everyone else.
Progress in technology has been made with the dual pixel AF on the 70D.
multiple sensors is the way to go.
200 years later cameras haven't changed much, only in the past 15 years.
I can't wait for the future! 

[neuroanatomist]

“totally usable photosite count”

Which means the "75 MP" sensor is really a 25 MP sensor in terms of spatial resolution.

[rs]

Finally someone gets the idea of multi sensors...one for highlights and one for shadows or even a third one for mid tones...
Instant Dynamic Range heaven!
I can't wait for a camera that sees what my eyes see...
Canon has to beat everyone else.
Progress in technology has been made with the dual pixel AF on the 70D.
multiple sensors is the way to go.
200 years later cameras haven't changed much, only in the past 15 years.
I can't wait for the future! 

More likely to be foveon-like than high DR, so a genuine 25MP output. This looks like the patent which is mentioned from May 2013:

http://www.canonrumors.com/2013/05/patent-canon-foveon-sensor/

[Drizzt321]

Finally someone gets the idea of multi sensors...one for highlights and one for shadows or even a third one for mid tones...
Instant Dynamic Range heaven!
I can't wait for a camera that sees what my eyes see...
Canon has to beat everyone else.
Progress in technology has been made with the dual pixel AF on the 70D.
multiple sensors is the way to go.
200 years later cameras haven't changed much, only in the past 15 years.
I can't wait for the future! 

Doesn't necessarily mean that, it might be stacked photosites to detect R, G, and B separately, but specially the same so that you get the true color at that pixel, instead of having to interpolate by de-bayering. Sorta like what they do with 3CCD cameras, except stack the photosites so you have a single chip.

[Snaps]

Finally someone gets the idea of multi sensors...one for highlights and one for shadows or even a third one for mid tones...
Instant Dynamic Range heaven!
I can't wait for a camera that sees what my eyes see...
Canon has to beat everyone else.
Progress in technology has been made with the dual pixel AF on the 70D.
multiple sensors is the way to go.
200 years later cameras haven't changed much, only in the past 15 years.
I can't wait for the future! 

So if I'm undestanding this correctly, it's a sensor with 75 million diodes which are split up into three parts for the highlights, midtones, and shadows?

[neuroanatomist]

Finally someone gets the idea of multi sensors...one for highlights and one for shadows or even a third one for mid tones...

So if I'm undestanding this correctly, it's a sensor with 75 million diodes which are split up into three parts for the highlights, midtones, and shadows?

You're correctly understanding clicstudio's desire, but not the patent on which this rumor is based.  That patent is for a Foveon-like sensor, where the three layers (different depths in the silicon wafer) are detecting blue/green/red.

[bchernicoff]

You're correctly understanding clicstudio's desire, but not the patent on which this rumor is based.  That patent is for a Foveon-like sensor, where the three layers (different depths in the silicon wafer) are detecting blue/green/red.

Isn't Sigma's 46 megaphotosite sensor supposed to have the equivalent spatial resolution as a 30mp bayer sensor? I wonder how a 75mps sensor from Canon would compare under ideal circumstances with one of their bayer sensors.

[Snaps]

Finally someone gets the idea of multi sensors...one for highlights and one for shadows or even a third one for mid tones...

So if I'm undestanding this correctly, it's a sensor with 75 million diodes which are split up into three parts for the highlights, midtones, and shadows?

You're correctly understanding clicstudio's desire, but not the patent on which this rumor is based.  That patent is for a Foveon-like sensor, where the three layers (different depths in the silicon wafer) are detecting blue/green/red.

Gotcha! I just did a quick read on the Foveon X3 sensor. Quite interesting tech I must say, although the Sigma Foveon cameras seem to fall short of what's possible with those sensors.

[Mt Spokane Photography]

The reference patent is for a Foveon type sensor.  It proposes to overcome the light loss in the deepest layer of photo sites by making the photo site three dimensional and resonant at the wavelength of the color (apparently red).  By making the photosite resonant, the sensitivity to that color of light is greatly increased, which makes tiny photosites more practical.
 
This is one of the big drawbacks with existing Foevon Sensors, they have poor light sensitivity due to light loss or diffusion of light traveling to the lower layer.  It gets very bad as photo sites get smaller.
 
Its very likely that the sensor is being tested, but that's normal, you have to test to verify that a new technology works and that its practical.  The big question is still whether it overcomes the inherent weaknesses of a Foveon type sensor.  They have issues with color accuracy, high ISO noise, and reduced DR.  We want more of all those, so it would be a big breakthrough if that actually happened.

[bchernicoff]

Its very likely that the sensor is being tested, but that's normal, you have to test to verify that a new technology works and that its practical.  The big question is still whether it overcomes the inherent weaknesses of a Foveon type sensor.  They have issues with color accuracy, high ISO noise, and reduced DR.  We want more of all those, so it would be a big breakthrough if that actually happened.

It sure would be neat if Canon released info on their sensor desigm/testing/manufacturing lifecycle on the Canon Museum site. It would be interesting to read about their "process". It might also give us some insight into how long after a "being tested" rumor surfaces we can expect a new product release.

UPDATE: it seems like it sorta exists, but is terrible(cheesy/marketing BS): http://www.canon.com/camera-museum/design/making/index.html

[neuroanatomist]

You're correctly understanding clicstudio's desire, but not the patent on which this rumor is based.  That patent is for a Foveon-like sensor, where the three layers (different depths in the silicon wafer) are detecting blue/green/red.

Isn't Sigma's 46 megaphotosite sensor supposed to have the equivalent spatial resolution as a 30mp bayer sensor? I wonder how a 75mps sensor from Canon would compare under ideal circumstances with one of their bayer sensors.

 A Foveon-type "75 MP" filter is giving you a spatial resolution based on 25 MP (because they're separately counting three stacked photosites that are in the same physical location on the image sensor.  But it's not a simple relationship, because it depends on the strength of the AA filter of a sensor with a Bayer CFA.  So a 'regular' 25 MP sensor (with a CFA and an AA filter) isn't giving you 25 MP of real spatial resolution, but something less.

[Mt Spokane Photography]

Its very likely that the sensor is being tested, but that's normal, you have to test to verify that a new technology works and that its practical.  The big question is still whether it overcomes the inherent weaknesses of a Foveon type sensor.  They have issues with color accuracy, high ISO noise, and reduced DR.  We want more of all those, so it would be a big breakthrough if that actually happened.

It sure would be neat if Canon released info on their sensor desigm/testing/manufacturing lifecycle on the Canon Museum site. It would be interesting to read about their "process". It might also give us some insight into how long after a "being tested" rumor surfaces we can expect a new product release.

UPDATE: it seems like it sorta exists, but is terrible(cheesy/marketing BS): http://www.canon.com/camera-museum/design/making/index.html

Thanks for the link, I'd forgotten where I'd seen it.
 
The big issue is that you have to kiss a lot of frogs to find that prince. 
 
I doubt if one in a hundred test designs makes it to production, but they need to be done to identify weaknesses that can't be discovered in a lab environment, and that allows for a chance to fix them, or go to plan B.
 
Canon now has their robotic manufacturing targets in place, so I'd expect the capability to be robotically assembled is a new factor in designs.  I'm suspecting that every sub assembly will have to be redesigned, since so much is now hand assembled, and its difficult.  Even more modular design and throw away assemblies that can't be repaired.  I was thinking that their new packaging for sensors on the SL1 may be part of that, they needed a smaller package, so why not kill two birds with one stone?
 
 
http://www.robots-dreams.com/2012/05/canon-building-new-robotic-factories-in-japan.html

[bchernicoff]

A Foveon-type "75 MP" filter is giving you a spatial resolution based on 25 MP (because they're separately counting three stacked photosites that are in the same physical location on the image sensor.  But it's not a simple relationship, because it depends on the strength of the AA filter of a sensor with a Bayer CFA.  So a 'regular' 25 MP sensor (with a CFA and an AA filter) isn't giving you 25 MP of real spatial resolution, but something less.

Maybe I was using the term spatial resolution wrong. I understand that an AA filter and the debayering algorithm reduce actual resolved detail compared to an ideal 25mp sensor with no color or AA filters at all. The problem is that we have gotten used to relating resolved detail to mp count with today's sensors. What I am wondering is what current-technology equivalent mp value would a 75 megaphotosite sensor maps to in terms of resolved detail? The Sigma cameras certainly resolve more detail than a camera with a traditional 15mp sensor does (like the 50D). I doubt it's comparable to 30mp as they claim, but it's surely better then 15mp.