The Canon EOS R6 Mark II will be announced on November 2, 2022

I mean, if it does end up with the R3's stacked sensor, we're in bargain of the decade territory, aren't we, even if Canon nerf some (or a lot) of the R3 features? Here's hoping. I think, more likely for the rumoured price, it'll be a simplified version of the R3 sensor. But fascinating to see what drops.
More likely than not, if they keep to dual SD cards, you can totally forget 195 fps raw :)
 
  • Haha
Reactions: 1 user
Upvote 0

Bahrd

Red herrings...
Jun 30, 2013
252
186
Sony claims to have dramatically improved the IBIS, but I have not seen anyone who actually tested it.
These guys are a quite reliable source (in Polish): Sony A7R V hands-on. They say:
Jeśli chodzi o nasze obserwacje, to uzyskaliśmy taki sam rezultat jak w teście A7R IV – 3 i 2/3 EV.
That is, there is no noticeable improvement.
 
  • Like
Reactions: 1 user
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,575
4,111
The Netherlands
More likely than not, if they keep to dual SD cards, you can totally forget 195 fps raw :)
On the R3, it buffers to memory and then writes it out to the card, it is limited to 50 full size raws in each burst. If the R6II gets the same amount of ram as the R3, it could do the same. The 195 fps mode isn’t limited by the cards.
 
Upvote 0
I found that moving from a 7d2 to an R6 gave me cleaner images in all but the best light conditions despite the massive drop in resolution. In average conditions I can crop in to match the FOV of the 7d2 and the image is better than the 7d2. The difference in all the other aspects of the sensor and AF more than make up for a lack of pixels most of the time
The resolution of both sensors is ~20mp. If you are referring to a 1.6x crop from ff to APS-c in apparent focal length then there is a difference.
 
Upvote 0

Aussie shooter

https://brettguyphotography.picfair.com/
Dec 6, 2016
1,183
1,817
brettguyphotography.picfair.com
The resolution of both sensors is ~20mp. If you are referring to a 1.6x crop from ff to APS-c in apparent focal length then there is a difference.
Yes. I am referring to pixels on target. Not total sensor resolution. As a wildlife photographer cropping is a given much of the time so that is how I compare my cameras
 
  • Like
Reactions: 1 user
Upvote 0
The price of the camera is not 80% sensor and 20% everything else. IBIS unit, EVF, Digic, memory, readers, LCD and materials for the body, projected volume (economy of scale) - return on investment ...
Thare is a reason for R6 and not wrongly projected supply of sensors. In first year you sell the sensor at the premium with the premium package (body, readers, EVF) all the while perfecting any problems in production (it is easear with a smaller production volume). With the R6 Canon may have also bridged a year of problems with supply chains and scaling of production BSI stacked sensors while making a big profit selling a very competitive body. And maybe good sale numbers of R6 just opened a new strategy resulting in a similar move for R6 II.
So the R6 II price is not than unusual - they can reuse the R6 body, maybe older DIGIC from the R6 and not the advanced one from R3 with less cache memory resulting in 20fps (not 30fps of R3), lower quality EVF (R6 could use an update - in low light it lacks dynamic range), dual SD is cheaper ... and you have a very, very capable body that is very competitive in the market and is a viable choice for all the people who cannot afford or justified the price of R3 or as a second body with similar specs. And you have R3 as a premium product for all the people who needs it.
 
  • Like
Reactions: 3 users
Upvote 0

koenkooi

CR Pro
Feb 25, 2015
3,575
4,111
The Netherlands
So how is that possible? Is it because the R has an OLPF and the R3 a weaker one or none?
Yes, starting with the 1Dx3 Canon has used a 'high res 16-point' OLPF. Shooting the 1dx3 and RP side by side, the pictures from the 1dx3 had a lot more detail, even if it had 6 fewer megapixels. This was shooting on a tripod using the 180L and manual focus, so the 1dx3 couldn't flex its AF muscles :)
 
  • Like
Reactions: 2 users
Upvote 0

Aussie shooter

https://brettguyphotography.picfair.com/
Dec 6, 2016
1,183
1,817
brettguyphotography.picfair.com
The 1Dx3 doesn’t use a fantastically expensive stacked sensor and all the electronics that come with it.
You say 'fantastically expensive ' sensor. Can you inform us what the price difference would be to reuse the r3 sensor as opposed to designing, retooling and producing and all new 24mp CMOS FF sensor? I literally have no idea so really need someone to fill me in
 
Upvote 0
Sep 20, 2020
3,066
2,395
You say 'fantastically expensive ' sensor. Can you inform us what the price difference would be to reuse the r3 sensor as opposed to designing, retooling and producing and all new 24mp CMOS FF sensor? I literally have no idea so really need someone to fill me in
All that we really know for sure is what Canon charges to replace the R3 image sensor.
It is quite a lot more expensive than they charge for replacing the sensor in the 1DX III.
 
  • Like
Reactions: 1 user
Upvote 0

Aussie shooter

https://brettguyphotography.picfair.com/
Dec 6, 2016
1,183
1,817
brettguyphotography.picfair.com
All that we really know for sure is what Canon charges to replace the R3 image sensor.
It is quite a lot more expensive than they charge for replacing the sensor in the 1DX III.
How much more? Again. I have no idea so I need a number to go off
 
Upvote 0
Jul 21, 2010
31,099
12,863
Not sure if changing the order of the layers really makes it much more expensive. Would it really cost more to print the electrical circuits at the back of the photodiodes?
View attachment 206095
Source: PC Mag
BSI etches the substrate wafer from both sides meaning a thinner, more fragile wafer. That likely leads to greater QC loss, which disproportionately affects larger sensors (with FSI, and made even worse with BSI).

Stacking requires a second die fused to the photodiode die. However, putting the bulk of the circuitry on a second die means the photodiode die is simpler and can be fabricated with larger-scale (older, cheaper) processes.

I have no idea how that breaks down in actual numbers. Certainly when they were new tech, BSI sensors cost more. But BSI enabled smaller pixel pitch, driven mainly by the smartphone and industrial sensors that are a much larger market than ILCs, and no doubt major efforts were made to reduce fab costs.

Either way, for high-end items like ILCs (all of them, even entry-level), the cost of goods is not the most important factor in setting price. It’s much more about market size and the required margin to drive ROI on the R&D.

An example from my own industry, the first Vertex drug to treat cystic fibrosis costs ~$475 per pill. One Tylenol caplet costs about the same to manufacture as one Kalydeco tablet, so that $475 for one pill would buy you >6,000 Tylenol caplets.
 
  • Like
Reactions: 1 users
Upvote 0