Big Megapixel Camera in 2014

[LetTheRightLensIn]

Well D800 already does 4fps at 36MP in 2012 so two years later it is not crazy to think you can get a 6fps 39MP 5D4.

This is the pixelrate discussion that I wanted us to steer into.

If we assume that the fundamental technology improves yearly (e.g. Moores law), then one might predict at which speed the pixelrate will improve in the future. This is probably the case for the digital image processing (raw->jpeg development): ARM processors (according to wikipedia, DIGIC is based on OMAP, using ARM) are progressing on a steady pace wrgt speed, heat, etc.

Canon have reportedly used the same process technology for years in their large image sensors. This means that they cannot take advantage of Moores law if I am right.

There may be other factors that does not improve at anywhere near Moores law. You want a shutter mechanism that flaps a substantial mass 12 times a second while not shaking the camera enough that any increase in MP count is effectively negated unless the camera is bolted to solid rock? You want to comply with RF requirements (and avoid cellphones killing image quality)? You want to have optical path tolerances that allows AF to focus the lense accurately at 40 MP (I have just spent a few hours AFMA-ing my 7D with lenses).

-h

Moore's doesn't apply to sensors but it does to digics and the memory buffer chips and all. Yeah at some point the shutter becomes an issue but a 6fps shutter is hardly a big deal for FF, not like the 5D3 doesn't do it already or that tones of other FF cams haven't done 6fps easily. WHo is saying more than 12fps for the 5D4?? And for 1 series the 1DX already does that.

40MP FF is still LESS density than the 7D so it hardly needs AF to be even more accurate (and the 5D3/1DX and the newest lenses DO have more precise AF anyway).

[jrista]

Well D800 already does 4fps at 36MP in 2012 so two years later it is not crazy to think you can get a 6fps 39MP 5D4.

This is the pixelrate discussion that I wanted us to steer into.

If we assume that the fundamental technology improves yearly (e.g. Moores law), then one might predict at which speed the pixelrate will improve in the future. This is probably the case for the digital image processing (raw->jpeg development): ARM processors (according to wikipedia, DIGIC is based on OMAP, using ARM) are progressing on a steady pace wrgt speed, heat, etc.

Canon have reportedly used the same process technology for years in their large image sensors. This means that they cannot take advantage of Moores law if I am right.

There may be other factors that does not improve at anywhere near Moores law. You want a shutter mechanism that flaps a substantial mass 12 times a second while not shaking the camera enough that any increase in MP count is effectively negated unless the camera is bolted to solid rock? You want to comply with RF requirements (and avoid cellphones killing image quality)? You want to have optical path tolerances that allows AF to focus the lense accurately at 40 MP (I have just spent a few hours AFMA-ing my 7D with lenses).

-h

Moore's doesn't apply to sensors but it does to digics and the memory buffer chips and all. Yeah at some point the shutter becomes an issue but a 6fps shutter is hardly a big deal for FF, not like the 5D3 doesn't do it already or that tones of other FF cams haven't done 6fps easily. WHo is saying more than 12fps for the 5D4?? And for 1 series the 1DX already does that.

I don't see the 5D Next getting 12fps. It just doesn't really make sense, and unless bandwidth from the sensor to the digics can be improved, there would either need to be a third digic and a broader bus, or maybe digic 5++.

The 12fps frame rate of the 1D X did require a redesigned mirror-box and mirror actuation mechanism. The 14fps bonus rate requires mirror lockup. Lot of fairly cutting edge technology there, which again seems to diminish the prospects of a 5D Next getting an ultra high frame rate.

[Rick]

Canon would be wise to release a hi-res 5 series alongside any 1 series release, you know, so that they can sell some meaningful number of units.

[neuroanatomist]

Another??

L versions of the TS-E 45mm and 90mm.

[jrista]

Another??

L versions of the TS-E 45mm and 90mm.

Of all the lenses out there, I really think these puppies need an update, along with becoming L-series glass. They are decent lenses for what they are, but really out dated at this point.

[pedro]

@jirista: Thank you so much for taking your time to elaborate your explanation for me. Now I got a faint idea of what it is all about. I highly appreciate that! Has  there leaked any info (patents) that Canon are changing to the CP-ADC approach sometime soon? or let's say within the 5DIIIs or 6Ds body cycle? Could the rumored 5DX contain at least some first components towards this system? Looking forward to read about CP-ADC approaches related to future canon sensor designs....Cheers, Pedro

[tcmatthews]

There is one issue I have not seen here that could be holding up Canons announcement.  That is at the write speeds discuses here it is unlikely that CompactFlash or SD cards can handle the write speed.  Canon is a very conservative company they could be exploring the successor technologies for CompactFlash.  CFast which is based on serial ATA, or the other one which is based on PCIe buss. One problem is both of these technologies are incompatible with CompactFlash pin layout.  What to do create a camera with 3 card inputs? CFast, CompactFlash, SD?

They are probably using this time to see which technology comes out on top and update a number of lenses to handle increased sensor density.   It is only a mater of time before CompactFlash is completely discontinued.  It is already becoming more and more a niche product.

[meli]

24mp x 10fps would require 1dx's pipeline so its highly unlikely, even 24x8 would be kinda farfetched but it might be forced if the competition steps up. That would drive the whole APS pro category to 2K pricerange across the brands and the signs sofar from both camps point to a more conservative generation, i guess we'll see..

Why?? 7D did the same as the 1 series throughput?
Did it make it cost $6000? No. Digics cost dollars.

that 6k must be 2k i suppose? Eitherway keep in mind that msrps were 1.7k for 7d & 5k for 1d4. Its 7k for the 1d5 1dX, plus consider Canon's pricing trend for the last 2 years

My point was that it is not crazy to expect that sort of throughput from a 7D2 when even the 7D did it and no it did force the 7D to cost anything close to 1 series prices.

Ditto. I don't really see any difference between the 7D's likeness to the 1D IV and the potential for the 7D II's likeness to the 1D X. The use of dual digics and high frame rate will not make the 7D II cost as much as a 1D X. The 7D II may indeed cost around $2100, but that is the price we pay for the continual advancement of technology...newly released products rarely start out at the ending street price of their predecessors.

Agree, thats what im saying in my original post; The 7dmk2 costing as much as a 1D was a misunderstanding from LetTheRightLensIn

[LetTheRightLensIn]

I don't see the 5D Next getting 12fps. It just doesn't really make sense, and unless bandwidth from the sensor to the digics can be improved, there would either need to be a third digic and a broader bus, or maybe digic 5++.

Neither do I. My point was that it wasn't crazy to expect 6fps with a lot of MP though from a 5D4.

[jrista]

I don't see the 5D Next getting 12fps. It just doesn't really make sense, and unless bandwidth from the sensor to the digics can be improved, there would either need to be a third digic and a broader bus, or maybe digic 5++.

Neither do I. My point was that it wasn't crazy to expect 6fps with a lot of MP though from a 5D4.

I know. I was agreeing with you, and reinforcing your point. ;P

[neuroanatomist]

Can't we all just agree to agree? 

[jrista]

@jirista: Thank you so much for taking your time to elaborate your explanation for me. Now I got a faint idea of what it is all about. I highly appreciate that! Has  there leaked any info (patents) that Canon are changing to the CP-ADC approach sometime soon? or let's say within the 5DIIIs or 6Ds body cycle? Could the rumored 5DX contain at least some first components towards this system? Looking forward to read about CP-ADC approaches related to future canon sensor designs....Cheers, Pedro

Glad I could help. I think the reason Exmor is so good is rather misunderstood...I even misunderstood it for a long time. Then I read a PDF article on it, and it all made sense. On-die parallel digital processing is really the way of the future, I think.

Regarding what Canon has, I really can't say. A lot of their patents are a bit over my head. I have a decent understanding of image sensor technology and electronics, so I can generally understand the bulk of their CIS patents, but when it comes right down to the exact mechanisms, I am not an expert, and I cannot say exactly 100% what Canon has up their sleeve. Also, I've only read the patents that have been granded and are publicly available...I'm sure Canon has patents they have filed over the last year or two that are pending grant, and not yet available to the public. What they may have pending could be anything...some form of CP-ADC is a possibility, CIS active thermal cooling is definitely another...but I have not seen explicit patents about them so far.

What I do know Canon has are regarding on-die hyperparallel readout logic. This does NOT include ADC, it is just column-wise parallel read. One of the readout patents described per-column  (rather than per-pixel) amplifiers, which I thought was interesting. Canon also has some interesting noise reduction patents. I don't understand them all (still need to finish reading and researching most of them). One that I did understand involved a power source disconnect feature in readout logic. The concept seemed to boil down to the idea that noise-generating dark current exists only when an active power source is flowing current through a circuit. Disconnecting the active power source, and using capacitance and existing charge to actually read out the pixel value, eliminated dark current as a source of noise.

The majority of the patents I have read seem to involve on-die readout logic. I am not sure that the stuff that occurs on-die (at least in Canon designs) is really the primary source of noise. Canon uses off-die ADCs housed in their DIGIC chips, across a high speed bus. I think the source of the most hated noise in Canon sensors...horizontal and vertical banding noise (HBVN, as I term it), is largely caused by the fact that read pixel data must then travel along a bus, after which they must be converted by high frequency ADCs. Assuming Canon has no on-die digital readout technology, I think their best option to reducing noise would be to eliminate that bus and move the ADCs on-die, in a much more parallel fashion than they have now. Even if it is not fully column-parallel, greatly increasing the parallelism beyond 16 and putting ADC on-die should help to reduce noise considerably.

There is also that ever-present die shrink. As pixel sizes get smaller, Canon is losing out in terms of Q.E. more and more. A 500nm process was quite good for APS-C and FF until a few years ago, however as pixel densities continue to increase, and as more and more logic is put on the same die as the pixels themselves, a 500nm process is going to really start hurting Canon's ability to compete on the sensor front. We've already seen it in the difference in IQ between the 5D III and D800, by an order of two magnitudes. A 180nm process is a three-fold shrink in transistor size. That means a few things. One, they can put more circuitry on-die without needing to dramatically increase the amount of power used (which, being the prime contributor to dark current noise, would have a negative impact on IQ).

Second, the usable area per pixel would increase at all sensor sizes and pixel densities. A 500nm transistor consumes a lot of space, which reduces the total photo-diode area that can actually receive photons. Microlensing helps, but it is far from perfect, and stronger, more precise microlenses are required when larger readout wiring and transistors are used. A 180nm process would allow more photodiode area to be dedicated to "pixel", and less to "transistor and wiring". Combined with better microlens technology, and possibly even better color filtering technology¥, Canon could bring their Q.E. up to competitive levels (55-65%), which would intrinsically reduce the impact of read noise. Add in their active thermal cooling, and Canon may even be able to turn the tables on their competition, surpassing D800 IQ (even without digital readout technology).

That is a LOT of technology Canon would need to bring to bear in the next cycle. They may have it, but perfecting such technology and making it viable for mass production takes time. If they have only developed some of that technology in the last year or so, it may still be another year or two before it is all actually ready for mass commercial consumption. I think the gambit for Canon is to figure out a way to "go digital". Analog signal processing has its advantages, is an extremely well known and well understood concept, but clearly digital signal processing, in the context of CIS, is and has surpassed it. And as time goes on, hardware DSP will only get better, and I suspect more and more DSP logic will shift on-die. If Canon cannot figure out a way to patent their own on-die CIS DSP technology, they will never really catch up to their primary competition (Sony). I can't say about other competition from other CIS manufacturers...not sure if any of them are actually doing the kind of digital readout that Sony CP-ADC does.

¥ Someone posted a very interesting link to a Panasonic patent that describes a color splitting replacement for color filters, which would allow all of the light incident on the sensor to be utilized, simply by redirecting light of certain colors to pixels of the matching color...i.e. Blue, White-Red, and White+Red. Right now, around 35-40% of the light incident on any given pixel is used, with the rest being filtered out. With color splitting in place of color filtering, the light that is not "appropriate" for any given pixel is simply redirected to the neighboring pixels where it is appropriate. Per-pixel quantum efficiency could technically double, from around 30-40% to 60-80%, if such technology could be idealistically employed. Note, per-pixel Q.E. is usually different than whole-sensor Q.E...currently, several Nikon cameras have greater than 50% Q.E., with one over 60%. That is for the entire sensor. With color splitting technology, theoretically I think Q.E. could reach 90% or more at room temperature. Combine that with some active cooling technology to eliminate thermal sources of noise, and the IQ of a future-generation image sensor with color splitting could blow even the D800 away, especially in extremely low light situations. Imagine taking a photo of an Aurora at a native ISO 204800 with IQ as good as or better than ISO 51200 today.

[TheSuede]

Every camera applies CDS, or Correlated Double Sampling. The point of CDS is to determine the reset dark current noise present in the sensor, before an exposure is taken. The charge of each pixel is registered and saved, and when the actual exposure is read out, the registered reset charge is subtracted from the exposure charge. That, effectively, eliminates noise caused by dark current present in the sensor at reset time. That is a moderate mount of noise, and exists in all sensors. It is not, actually, the worst form of electronic (read) noise, so CDS only does a partial job of removing noise in hardware.
.../cut a bucketload of text/...

There are some very dubious statements blended in there.

1) the patents regarding the exmor PRINCIPLE are highly questionable, and Sony knows this very well. Due to publicly available prior art, the general patents would never stand in court. This is why companies as Panasonic, Aptina and most notably maybe Cmosis in the latest Leica CMOS sensor use on-chip layouts very similar to the original "Exmor" patents. I doubt there's much difference between the Cmosis Leica M sensor and an exmor except in general detailing, the principles are all the same.

2) The more specified patents are in fact more of a copyright statement. "We implement this by doing this, and we do it in this way with this layout". That's normal behavior when you don't want others to copy your works by putting your stuff (basically) in a photocopier. It stops others from doing exact copies, it does not stop them from using the same basic flow principles.

In the end this means that the statement about "Sony having patended the way to get rid of banding and read noise" is pure BS. And when I say pure, I mean 100.00% pure.

That's kind of like saying that Ford has patented a car that does not swerve off the road and kill you every 50 miles or so. The general principle in itself (not killing you every 50 miles) is well known beforehand, and can be worked towards in many different and equally efficient ways.

There are thousands of possibilities for Canon's interleave readout to be balanced and banding-free. Literally thousands of ways to make it happen. Ten or fifteen of them would possibly be in conflict with a presently valid patent, but that should not stop them.

Regarding CDS, Canons version of CDS is actually more effective than the way Sony does it in some of their sensors (not all Exmor marked sensors have digital CDS!). This can be seen when you use the cameras at higher ISO - which is where the particular components of noise that CDS is meant to be working against is most visible.

Where Canon falls flat on their face is the path where the signal is taken FROM the sensor (after the CDS has already been applied) and VIA the ISO amplifier INTO the AD converters.
-That path is seriously unbalanced, and also quite noisy. In fact - at the lowest two full-step ISOs in a Canon camera, the sensor is nowhere near being the largest electronic noise-source. The stuff Canon use outside the sensor to digitize the signal adds four times more noise than the sensor does at base ISO.

[jrista]

Every camera applies CDS, or Correlated Double Sampling. The point of CDS is to determine the reset dark current noise present in the sensor, before an exposure is taken. The charge of each pixel is registered and saved, and when the actual exposure is read out, the registered reset charge is subtracted from the exposure charge. That, effectively, eliminates noise caused by dark current present in the sensor at reset time. That is a moderate mount of noise, and exists in all sensors. It is not, actually, the worst form of electronic (read) noise, so CDS only does a partial job of removing noise in hardware.
.../cut a bucketload of text/...

There are some very dubious statements blended in there.

1) the patents regarding the exmor PRINCIPLE are highly questionable, and Sony knows this very well. Due to publicly available prior art, the general patents would never stand in court. This is why companies as Panasonic, Aptina and most notably maybe Cmosis in the latest Leica CMOS sensor use on-chip layouts very similar to the original "Exmor" patents. I doubt there's much difference between the Cmosis Leica M sensor and an exmor except in general detailing, the principles are all the same.

2) The more specified patents are in fact more of a copyright statement. "We implement this by doing this, and we do it in this way with this layout". That's normal behavior when you don't want others to copy your works by putting your stuff (basically) in a photocopier. It stops others from doing exact copies, it does not stop them from using the same basic flow principles.

In the end this means that the statement about "Sony having patended the way to get rid of banding and read noise" is pure BS. And when I say pure, I mean 100.00% pure.

Sony has patented their approach to CP-ADC, and the patents involve Digital CDS as part of the same curcuitry as the column-wise ADCs. It is a fairly specific patent. I am not sure there are thousands of ways of doing what Sony does, which is extremely effective, and that was what I was addressing.

There are thousands of possibilities for Canon's interleave readout to be balanced and banding-free. Literally thousands of ways to make it happen. Ten or fifteen of them would possibly be in conflict with a presently valid patent, but that should not stop them.

I'm sure there are thousands of ways of reducing noise in general. I would dispute the notion that there are "literally thousands of ways to balance interleaved readout and make it banding-free". Even the D5200, which uses a form of highly parallel, but not specifically column-parallel, readout has banding issues. Granted, they are a lot less than canons, but also plenty visible enough for people to complain about. To be 100% completely banding free, I do NOT believe there are unlimited options. Could Canon do it in a purely analog fashion? Maybe. Would it be as good as Sony Exmor? Possibly, but I'd call it s tall mountain to climb. Could they get as good as the D5200? Certainly.

Regarding CDS, Canons version of CDS is actually more effective than the way Sony does it in some of their sensors (not all Exmor marked sensors have digital CDS!). This can be seen when you use the cameras at higher ISO - which is where the particular components of noise that CDS is meant to be working against is most visible.

Regarding the fact that not all Exmor have DCDS, sure, there are a couple variants, one is BSI and designed for much smaller form factors, and its design is (probably by necessity) simpler than the FF version of Exmor. If you want to be that nit-picky, I'm explicitly referring to the FF and APS-C Exmor design as used in the D800 and D3200, both of which, as far as I know, employ CP-ADC with DCDS. As for Canons CDS working better than Sonys at higher ISO, I have no information to confirm or dispute that. Either way, though, read noise at higher ISO levels is such a minuscule factor of noise overall, which is completely dominated by Photon noise, I am not sure that Canon having a better CDS design even matters. Dark current's contribution to noise at that level is on the order of 2-3 electrons per pixel...trivial.

Where Canon falls flat on their face is the path where the signal is taken FROM the sensor (after the CDS has already been applied) and VIA the ISO amplifier INTO the AD converters.
-That path is seriously unbalanced, and also quite noisy. In fact - at the lowest two full-step ISOs in a Canon camera, the sensor is nowhere near being the largest electronic noise-source. The stuff Canon use outside the sensor to digitize the signal adds four times more noise than the sensor does at base ISO.

I don't dispute that. If you read my entire post, you would see I not only agree, I described that problem in detail. I am not sure that the off-die stuff adds 4x more noise than the sensor does (a reference where that is explained would be nice), but it is certainly the primary contributor. Canon's noise problem is primarily off the sensor die. Hence the reason I claimed (and have claimed for a while now) that Canon at the very least, even if they do not move to a digital readout and NR system, needs to get everything on-die. Shipping digital information along a bus for further processing is a hell of a lot safer from further contamination by noise than shipping analog information along a bus.

That is also the one area where Canon seems to be lacking patents. I have not seen anything that I could call an on-die ADC patent. I do believe that so long as Canon ships a huge amount of analog signal information off the sensor, especially if they use very few higher frequency off-die ADCs, they will always have high low-ISO read noise.

[Don Haines]

I figure a BIG megapixel camera will need a big sensor..... Could this be what Canon has in store for us..... and just imagine the size of a telephoto lens!

http://www.dpreview.com/news/2010/8/31/canonlargestsensor