More Big Megapixel Talk [CR1]

[Canon-F1]

Anyone been over at nikonrumors and hear what they complain over? It seems that the guys here love nikon more than they do...

no wonder .. both companys have switched their school of thought drastically.

we user need some time to adapt. no matter if nikon or canon user. 
 
the poor nikon guys sitting there with a D800 on their knees and don´t know what to do with all the megapixels.  it´s a heavy jump from 12 to 36 MP.

while we canon guys suddenly see that sony, out of nowhere, produces the best sensors.
and we wonder why our sensors have only 22MP and no better DR then a tiny, dwarf-size-photosite equipped D800 sensor.

[Drizzt321]

Anyone been over at nikonrumors and hear what they complain over? It seems that the guys here love nikon more than they do...

no wonder .. both companys have switched their school of thought drastically.

we user need some time to adapt. no matter if nikon or canon user. 
 
the poor nikon guys sitting there with a D800 on their knees and don´t know what to do with all the megapixels.  it´s a heavy jump from 12 to 36 MP.

while we canon guys suddenly see that sony, out of nowhere, produces the best sensors.
and we wonder why our sensors have only 22 and no better DR then a tiny, dwarf-size-photosite equipped D800 sensor.

Actually, I find that 22 MPx is about perfect for me, although I would like more DR at low ISO, but my 5d3 still beats out the D800 at high ISO which I use quite a bit   Different strengths for different people is all.

[verysimplejason]

... a very new sensor design/overhaul. The emphasis is in the dynamic range of the sensor ...

New sensor technology?  If so I wonder what it is?  Any patents to hint at it?

I'm interested because technology tends to trickle down to consumer cameras that I can afford.

We'll see... We'll just have to wait for about 2 to 3 years before we can get it down to 5D2 price level then that's the time I'll get an FF. 

[pwp]

i also think it will be very expensive.
in the range of 8000-10000$
maybe called EOS-S  (for studio).   

It wouldn't make any sense for Canon to price a DSLR that high.  You might as well buy a used Hassy at that price.  I bet it will be under $5000 - if this rumor is true, of course.

That's true, you might just get a pre-owned Hassy for similar money. But if the option is there to run with just one set of lenses, it makes Canon's megapixel monster hugely attractive if it's priced in the $8-10K range. Do a quick search for pricing on lenses for that Hasselblad. Ouch!

-PW

[LetTheRightLensIn]

All of this info points to it being expensive as hell. I'm guessing $6K minimum.

Yeah the new naming scheme makes me afraid it will be the EOS 1$   
expensive and brick sized and slow.

But talk of super DR is pretty exciting.

Imagine a 5D4 in a few years with tons of DR, 38MP, super video, 6fps, same top AF.  All thoughts of going to the dark side might end pretty quickly. 

[LetTheRightLensIn]

How long until we hear about people who switched to Nikon wanting to switch back to Canon..... 

yes you have to sit the dry years out.

Anyone been over at nikonrumors and hear what they complain over? It seems that the guys here love nikon more than they do...

 

[heheapa]

High Megapixels for Landscape and Studio? That's 1DS Grade. I bet this should be sold for 10k+

[Zlatko]

Actually, I find that 22 MPx is about perfect for me, although I would like more DR at low ISO, but my 5d3 still beats out the D800 at high ISO which I use quite a bit   Different strengths for different people is all.

"Different strengths for different people is all." — Well said!  22mp and great high ISO is perfect for me too.

[RuneL]

If the sensor is a 16-bit sensor with some kind of active cooling (no, not necessarily a fan), and Canon doesn't completely botch the ISO 100 and 200 electronic noise, then it could stomp all over the D800. With an extra two bits of information they could push 15 stops of DR, maybe even a little more (but no more than 16.0.)

My guess is that its still probably their same old sensor tech, but with some kind of efficient cooling to keep the sensor below room temperature (thereby reducing electronic noise), and extra bit depth. Canon needs the active cooling because they are either incapable of innovating and patenting technology similar but different enough to Sony Exmor, or there simply ISN'T another way to reduce noise electronically like Exmor, and Canon either has to pay Sony royalties, or do something entirely different.

Am I wrong in thinking that the amount electronic noise stems from the placement of the image processing unit? Too close to the sensor, too much heat for say video,  far enough away to eliminate heat more electronic noise as it passe through the camera?
Maybe they should redesign the chip and do a pure imaging camera meant for the absolute best stills possible.

[c.d.embrey]

I do mainly studio work and 46.1 megapixel is overkill. For magazines ads and catalog work a 16Mp 1Ds2 is all I need. I've used a 5D3, and it main advantage is it's lighter weight when used out of the studio.

DR is no big deal 'cause I've got enough Profoto watt seconds to cause sunburn        

But I'm sure 46.1Mp will sell well with the Doctor and Lawyer Photo Enthusiasts                                                      

[jrista]

If the sensor is a 16-bit sensor with some kind of active cooling (no, not necessarily a fan), and Canon doesn't completely botch the ISO 100 and 200 electronic noise, then it could stomp all over the D800. With an extra two bits of information they could push 15 stops of DR, maybe even a little more (but no more than 16.0.)

My guess is that its still probably their same old sensor tech, but with some kind of efficient cooling to keep the sensor below room temperature (thereby reducing electronic noise), and extra bit depth. Canon needs the active cooling because they are either incapable of innovating and patenting technology similar but different enough to Sony Exmor, or there simply ISN'T another way to reduce noise electronically like Exmor, and Canon either has to pay Sony royalties, or do something entirely different.

Am I wrong in thinking that the amount electronic noise stems from the placement of the image processing unit? Too close to the sensor, too much heat for say video,  far enough away to eliminate heat more electronic noise as it passe through the camera?
Maybe they should redesign the chip and do a pure imaging camera meant for the absolute best stills possible.

You are thinking of "read noise". Thats one kind of noise that can be introduced by electronics, and the most frequently talked about noise. Its usually added by the ADC, which in most cases tend to operate at a pretty high frequency (which has the tendency to introduce noise. Analog-to-Digital conversion also tends to introduce quantization error, which exhibits as a minor amount of noise.

Thermal noise is different than electronic noise. It can be a problem, but when people talk about thermal noise its usually the kind that only occurs at temperatures a fair bit higher than room temperature...such as when the sensor has been exposing for a long time (say during long exposure with an ND filter or for night sky/milky way photography.)

There are other forms of electronic noise. Dark current noise is caused by charge buildup due to the small amount of continuous current that runs through any circuit. There are also forms of noise caused by slight differences in response for each photodiode, etc.

Sony Exmor mitigates noise via more complex electronics. Their technology adds circuitry to each pixel to mitigate dark current noise, transistor response differential noise, etc. Exmor also moved the ADC on-die, and hyper-parallelized it by having one for each column of pixels. That all had the effect of greatly reducing (but not quite eliminating) electronic noise.

Prior to Exmor, the most common way to reduce or eliminate electronic noise was to cool the sensor. Silicon devices tend to improve in efficiency at colder temperatures. Around -80°C, noise in a circuit caused by the electronics themselves is nearly eliminated (approximately 200 times less than the same circuit at room temperature.) I highly doubt that Canon has attempted to cool their sensors to sub-freezing temperatures, so they won't be gaining that much of a lead over Sony and Nikon (and technically speaking, Exmor is still the better technology, as it mitigates noise without the need for added cooling, which to really be effective in Canon's next camera might need to be thermoelectric, requiring additional power.) I am guessing Canon has found a way to moderately cool their sensor, probably with an efficient heat pipe cooler or something along those lines. They won't be able to get the sensor below "room temperature", but they should be able to keep it cooler than it would otherwise be. If they add some form of active cooling...fans, external heat plates to exhaust heat from an internal heat pipe cooler, etc. they might be able to get the sensor below room temp. If they add a peltier that operated at a lower current load, they could cool the sensor quite a bit...but again that would require an additional power draw (proportional to how much they actually cool the sensor.)

[Leon]

How long until we hear about people who switched to Nikon wanting to switch back to Canon..... 

Did somebody just call my name?

No seriously, I'd say it'll take a while. For me, and I think for many other people, the Sony sensor got Nikon attention, but actually Canons recent price policy had me buy the D800 instead of the 5D3.
The 5D3, good as it is, is priced at the top end of what they could take. So is the Speedlite 600XT, the 24-70 L II (gee, look what Tamron gets you for half the price), the 1D-C and, most recently, the 6D. DPReview put it very nicely: "Whereas Nikon seems to have taken the approach of taking away as little as possible from D800 when creating the D600, Canon appears almost to have gone the other way, removing as much as it thinks it can get away with at the price."

- I'm not saying Canon is stupid to do so, and I'm definitely not a Nikon fanboy, but this is a big part of what had me switch to Nikon, and since I don't see that changing soon, my bet is that system switchers like me will not reconsider as quickly.

For the big MP body: I don't know why you wouldn't want it, all Canon has to make sure is to include various downsized RAW outputs (and yes I find it annoying that Nikon didn't have that idea for the D800). They might need a pretty fast CPU in the camera, but in principle that shouldn't be a problem. Offer RAW output @46, 23 and 11.5 MP and no one can say it has too many MP. Hardly anyone NEEDS 46MP, but having it in the bag for landscapes or some commercial shoots is definitely a selling point.

However, not only the resolution, but also the aim to go for higher DR sounds like they're back on track with their sensor development. Someone here mentioned some special technology in the Sony sensors, could you go into more detail on that? I think it is very apparent that Sony has got some magic going on for the shadow noise/ DR. It's ridiculous how those Sony sensors can save your day after you screw things up in manual mode, so Ive been wondering why Canon and Nikon sensors can't put up with that. (ok this was already answered while I was typing)

[elbeasto]

Newer, bigger and better specs are all well and good but I'd be happy if a new body was released from either company without any design flaws and having only average specs.

Both the 5DIII and D800 were released with problems.
In this day and age, with the technology available and the approximate price tag of $4k (upon release), glaring design flaws are unacceptable, period.

46MP, great, lets just hope it's light tight and all the basic functions work as they should.

EDIT: I'd also agree with the comment about about Canikon's pricing policies, neither their lenses or bodies quality reflect the asking price.

[neuroanatomist]

I think CRguy mis-titled the topic.  Here's the properly marked up version:

More Big Megapixel Talk [CR1]

If they add a peltier that operated at a lower current load, they could cool the sensor quite a bit...but again that would require an additional power draw (proportional to how much they actually cool the sensor.)

Indeed.  Most of my scientific imaging cameras have Peltier-cooled sensors, in many cases cooled well below 0 °C.

[verysimplejason]

I'm not worrying about this too much.  I'm sure Canon engineers are brilliant enough to find other ways.  Besides, it's their jobs that are on the line right now.  I think there's a lot of pressure going around just to be able to have a very good sensor.  The Canon marketing and management can only take that much criticism before they think of firing those engineers and hiring better ones.

If the sensor is a 16-bit sensor with some kind of active cooling (no, not necessarily a fan), and Canon doesn't completely botch the ISO 100 and 200 electronic noise, then it could stomp all over the D800. With an extra two bits of information they could push 15 stops of DR, maybe even a little more (but no more than 16.0.)

My guess is that its still probably their same old sensor tech, but with some kind of efficient cooling to keep the sensor below room temperature (thereby reducing electronic noise), and extra bit depth. Canon needs the active cooling because they are either incapable of innovating and patenting technology similar but different enough to Sony Exmor, or there simply ISN'T another way to reduce noise electronically like Exmor, and Canon either has to pay Sony royalties, or do something entirely different.

Am I wrong in thinking that the amount electronic noise stems from the placement of the image processing unit? Too close to the sensor, too much heat for say video,  far enough away to eliminate heat more electronic noise as it passe through the camera?
Maybe they should redesign the chip and do a pure imaging camera meant for the absolute best stills possible.

You are thinking of "read noise". Thats one kind of noise that can be introduced by electronics, and the most frequently talked about noise. Its usually added by the ADC, which in most cases tend to operate at a pretty high frequency (which has the tendency to introduce noise. Analog-to-Digital conversion also tends to introduce quantization error, which exhibits as a minor amount of noise.

Thermal noise is different than electronic noise. It can be a problem, but when people talk about thermal noise its usually the kind that only occurs at temperatures a fair bit higher than room temperature...such as when the sensor has been exposing for a long time (say during long exposure with an ND filter or for night sky/milky way photography.)

There are other forms of electronic noise. Dark current noise is caused by charge buildup due to the small amount of continuous current that runs through any circuit. There are also forms of noise caused by slight differences in response for each photodiode, etc.

Sony Exmor mitigates noise via more complex electronics. Their technology adds circuitry to each pixel to mitigate dark current noise, transistor response differential noise, etc. Exmor also moved the ADC on-die, and hyper-parallelized it by having one for each column of pixels. That all had the effect of greatly reducing (but not quite eliminating) electronic noise.

Prior to Exmor, the most common way to reduce or eliminate electronic noise was to cool the sensor. Silicon devices tend to improve in efficiency at colder temperatures. Around -80°C, noise in a circuit caused by the electronics themselves is nearly eliminated (approximately 200 times less than the same circuit at room temperature.) I highly doubt that Canon has attempted to cool their sensors to sub-freezing temperatures, so they won't be gaining that much of a lead over Sony and Nikon (and technically speaking, Exmor is still the better technology, as it mitigates noise without the need for added cooling, which to really be effective in Canon's next camera might need to be thermoelectric, requiring additional power.) I am guessing Canon has found a way to moderately cool their sensor, probably with an efficient heat pipe cooler or something along those lines. They won't be able to get the sensor below "room temperature", but they should be able to keep it cooler than it would otherwise be. If they add some form of active cooling...fans, external heat plates to exhaust heat from an internal heat pipe cooler, etc. they might be able to get the sensor below room temp. If they add a peltier that operated at a lower current load, they could cool the sensor quite a bit...but again that would require an additional power draw (proportional to how much they actually cool the sensor.)