Canon 7D Mark II - Finally using Canon's newer fab?

[jrista]

So, I was poking around on sensorgen.info, and noticed the 7D II was listed. I clicked into it, and saw this:

A larger format Canon sensor with 59% Q.E.!! I've only seen that on their small sensors, the ones that come from their other fab. It hasn't topped 60% yet, but overall it's quite a jump from the 40% range that the vast majority of Canon's prior DSLRs fell into.


Given the fact that DR hasn't improved despite the increase in Q.E., I am more certain than ever that Canon's problem is in their ADC units. I think this correlates well with the fact that as the readout frequency increases, read noise increases (1D X, 70D, 7D II...all have higher read noise.) That gives me hope that if Canon does move in the future to an on-die hyper-parallel readout system that operates at a lower frequency, that they should be able to reduce their read noise.


Read noise aside, seeing a Canon APS-C sensor with 59% Q.E. is a good sign. Certainly nothing conclusive, however it does indicate that they very likely have move at least 7D II sensor production to a better fab. It's a stepping stone, and hopefully one of very few to get to a better place on the sensor technology front. Here's to hoping the 5D IV sensor tops 60% Q.E. (and gets to around 65%...that would be another nice jump), and also gets an on-die readout system.

 

[2n10]

Good catch Jon. I wonder if the 5DIV or the 1DXII will be the sweepstakes winner.

 

[Don Haines]

So, I was poking around on sensorgen.info, and noticed the 7D II was listed. I clicked into it, and saw this:

A larger format Canon sensor with 59% Q.E.!! I've only seen that on their small sensors, the ones that come from their other fab. It hasn't topped 60% yet, but overall it's quite a jump from the 40% range that the vast majority of Canon's prior DSLRs fell into.


Given the fact that DR hasn't improved despite the increase in Q.E., I am more certain than ever that Canon's problem is in their ADC units. I think this correlates well with the fact that as the readout frequency increases, read noise increases (1D X, 70D, 7D II...all have higher read noise.) That gives me hope that if Canon does move in the future to an on-die hyper-parallel readout system that operates at a lower frequency, that they should be able to reduce their read noise.


Read noise aside, seeing a Canon APS-C sensor with 59% Q.E. is a good sign. Certainly nothing conclusive, however it does indicate that they very likely have move at least 7D II sensor production to a better fab. It's a stepping stone, and hopefully one of very few to get to a better place on the sensor technology front. Here's to hoping the 5D IV sensor tops 60% Q.E. (and gets to around 65%...that would be another nice jump), and also gets an on-die readout system.

we knew it was just a matter of time.....

 

[Dylan777]

@ 59% Q.E would that apply to JPEG or RAW?

We have seen great high ISO photos from 7D II. However, most(all) of photos were shot in JPEG.

 

[jrista]

@ 59% Q.E would that apply to JPEG or RAW?

We have seen great high ISO photos from 7D II. However, most(all) of photos were shot in JPEG.

The data on sensorgen.info is based on the raw-derived data from DXO, before DXO mucks with it. DXO gets a LOT of raw data, from countless individual RAW images, so I'd say it's pretty accurate.

 

[swampler]

DR may not have improved at base ISO, but it has improved at higher ISOs where most sports and wildlife shots take place (high shutter speed and low light). Improvements where it's needed.

 

[LetTheRightLensIn]

If they went to a new fab then why on earth would they not have implemented any of their column ADC or dual ISO read out patents?

Maybe it's just from the CFA filter having been made even more color blind.

 

[Don Haines]

If they went to a new fab then why on earth would they not have implemented any of their column ADC or dual ISO read out patents?

Maybe it's just from the CFA filter having been made even more color blind.

They are very conservative... this might be a case of step one - get the photodiodes/microlenses/support circuitry transferred over to the new fabrication run, and step 2 being to implement the A/D on chip. Don't forget that they are currently using A/D chips from a third party and that might complicate the move as they have no experience designing them... it could be a licencing of the design problem... or countless other reasons. It will come, but with Canon expect it to take time and when it does arrive, to be good.

 

[Marauder]

So, I was poking around on sensorgen.info, and noticed the 7D II was listed. I clicked into it, and saw this:

A larger format Canon sensor with 59% Q.E.!! I've only seen that on their small sensors, the ones that come from their other fab. It hasn't topped 60% yet, but overall it's quite a jump from the 40% range that the vast majority of Canon's prior DSLRs fell into.


Given the fact that DR hasn't improved despite the increase in Q.E., I am more certain than ever that Canon's problem is in their ADC units. I think this correlates well with the fact that as the readout frequency increases, read noise increases (1D X, 70D, 7D II...all have higher read noise.) That gives me hope that if Canon does move in the future to an on-die hyper-parallel readout system that operates at a lower frequency, that they should be able to reduce their read noise.


Read noise aside, seeing a Canon APS-C sensor with 59% Q.E. is a good sign. Certainly nothing conclusive, however it does indicate that they very likely have move at least 7D II sensor production to a better fab. It's a stepping stone, and hopefully one of very few to get to a better place on the sensor technology front. Here's to hoping the 5D IV sensor tops 60% Q.E. (and gets to around 65%...that would be another nice jump), and also gets an on-die readout system.

Great catch Jon!

 

[jrista]

If they went to a new fab then why on earth would they not have implemented any of their column ADC or dual ISO read out patents?

Maybe it's just from the CFA filter having been made even more color blind.

Highly, highly doubtful it has anything to do with the CFA. The CFA "color blindness" might have a minor impact on color noise, but overall the primary source of read noise is the off-die stuff.


I think this is just a baby step. Don said it already, Canon is a very conservative company. Who knows why they did not implement CP-ADC or dual-scale ADC or anything like that. (As far as I know, Canon does not have a dual ISO readout patent...they have a dual-scale ADC, which is used to switch to a slower readout when possible (i.e. longer exposures/slower frame rates) which allows for lower frequency ADC operation, which does reduce noise, but not like a dual-ISO readout would.)

 

[Woody]

I don't think sensor QE tells us much. Consider the following Nikon camera models:

Camera QE
D3 90%
D3X 42%
D3s 57%
D4 52%
D4s 52%
D610 51%
D800 56%
D810 47%

 

[Coldhands]

I would like nothing more that for Canon to finally move on from their 500nm process, but I can't help but think that if they went to all the trouble to implement a new fab, that they would use it to create an all-new sensor that provides market separation from the 70D. In other words, why spend so much on R&D for your top-of-the-range APS-C body only for people to think "oh, its just the same as the old sensor with a few tweaks."

Then again my knowledge of CMOS production is elementary, and there could be a perfectly good reason to use a new process on existing architecture. Here's hoping that Chipworks analyses one so we can know for sure.

 

[Harv]

You lost me at.... A larger format Canon sensor with 59% Q.E.!!

 

[AlanF]

@ 59% Q.E would that apply to JPEG or RAW?

We have seen great high ISO photos from 7D II. However, most(all) of photos were shot in JPEG.

The data on sensorgen.info is based on the raw-derived data from DXO, before DXO mucks with it. DXO gets a LOT of raw data, from countless individual RAW images, so I'd say it's pretty accurate.

Quantum efficiency (QE) is the efficiency at which photons are converted to current, which is prior to and independent of the processing of data to RAW or JPEG. Right?

 

[Marsu42]

So, I was poking around on sensorgen.info, and noticed the 7D II was listed. I clicked into it, and saw this

Interesting, changing the fab probably is the reason why Canon calls it another sensor than the 70d when at first sight they're very similar. Is the qe also the reason why it has nearly as good (for Nikon trolls: worse) dynamic range as the 6d (iso200: 7d2: 11.2ev, 6d:11.4ev and ... 70d: 10.8ev )?

It would be interesting to know what sensorgen-like values would result if this sensor-qe would be upscaled to ff size.

 

[racebit]

I would like nothing more that for Canon to finally move on from their 500nm process, but I can't help but think that if they went to all the trouble to implement a new fab, that they would use it to create an all-new sensor that provides market separation from the 70D. In other words, why spend so much on R&D for your top-of-the-range APS-C body only for people to think "oh, its just the same as the old sensor with a few tweaks."

Then again my knowledge of CMOS production is elementary, and there could be a perfectly good reason to use a new process on existing architecture. Here's hoping that Chipworks analyses one so we can know for sure.

Intel, the leading CPU manufacturer, uses a TIC TOC method for their CPUs, where one changes the the manufacturing process and the other changes the architecture, but Intel does not change both at same time, too difficult / risky. They prefer to consolidate the manufacturing process first, and only on next iteration change the architecture.
Let's hope the same is happening with Canon, but I still find too good to be true.

 

[privatebydesign]

I don't think sensor QE tells us much. Consider the following Nikon camera models:

Camera QE
D3 90%
D3X 42%
D3s 57%
D4 52%
D4s 52%
D610 51%
D800 56%
D810 47%

Just goes to show, the further the measurement moves away from actual best processed images, the less interesting and relevant the metric becomes. D810 9% less than the D800 and 12% less the the 7D MkII.

 

[tayassu]

I don't think sensor QE tells us much. Consider the following Nikon camera models:

Camera QE
D3 90%
D3X 42%
D3s 57%
D4 52%
D4s 52%
D610 51%
D800 56%
D810 47%

90% with the D3?! That can't be true...

But otherwise, a nice improvement!

 

[Sporgon]

I don't think sensor QE tells us much. Consider the following Nikon camera models:

Camera QE
D3 90%
D3X 42%
D3s 57%
D4 52%
D4s 52%
D610 51%
D800 56%
D810 47%

Just goes to show, the further the measurement moves away from actual best processed images, the less interesting and relevant the metric becomes. D810 9% less than the D800 and 12% less the the 7D MkII.

That's nothing: we should all be still using Nikon D70s, QE two hundred and something percent !

I'm not convinced on how this quantum efficiency is calculated using SNR. I have both the 5DII ( QE 31 %) and 6D ( 49%) and am struggling to see how this is relevant in practice. Perhaps in low light there is an advantage in over exposing the 5DII slightly compared with the 6D, but it's unclear. I presume that if enough photons have been retained by the pixel to record a 100% accuracy and converted to an appropriate charge, that is all that matters: once it's 100% accurate it doesn't matter what percentage it was of the total available.

I guess that as QE goes up ( in the way it is calculated) the sensor can withstand less exposure, but this doesn't necessarily relate to pushing under exposure. This then also allows more acceptable results at higher ISOs. That also matches the characteristics of the 5DII and 6D.

However I do often look back at images from the Nikon D70 that I had - replaced it with a D200 - and think 'that D70 was a damn fine camera'.

Just a bit of additional information: I believe the 'equivalent QE' of film was in the region of 10%.

 

[Lee Jay]

@ 59% Q.E would that apply to JPEG or RAW?

We have seen great high ISO photos from 7D II. However, most(all) of photos were shot in JPEG.

The data on sensorgen.info is based on the raw-derived data from DXO, before DXO mucks with it. DXO gets a LOT of raw data, from countless individual RAW images, so I'd say it's pretty accurate.

The fact that the D3 is listed at 90% QE would seem to indicate otherwise.

I suspect the problem is that DxO doesn't take into account any fiddling companies are doing to their raw data.