Quad-Pixel AF and a global shutter coming in 2021?

[analoggrotto]

You already have that capability with DualPixel RAW, it's just nobody is processing it that way.

Yes but I'm talking about native, real time handling. That way the sensor could be rated to the true amount of pixels not half as all dual pixel sensors are now.

 

[Danglin52]

Those are not several things but 6.

R1 was not released as the surveyed bulk buyers and other 1-Series customers from 2017-2019 prefer SLRs.

That may have changed within the last 52 weeks but the question I'd pose is is Canon willing to take the financial lose of Osborning the 1D X Mark III and Series III EF 600/4 & EF 400/2.8?

Now, if Canon released a RF 600/4 & RF 400/2.8 in 2018 instead and R1 in 2020 instead then that would have been better.

I have no doubt that the MILC tech for the R1 has been there for more than 2 years already but what is needed to support a 1-Series body is the appropriate lens and accessories for it. Thus far only the three L f/2.8 zooms are there. No white long fast primes.

Nikon themselves admitted that their transition to MILC was slow. By my estimate by 2 years. Canon was late by 1.

Sony timed it perfectly. Their FE 400/2.8 came out in 2018, FE 600/4 and a9 II came out in 2019.

The R1 is not really a follow on for the 1dx III, it is a transition platform to new technology direction. Canon released the R5/R6 to establish the beachhead and prove the technology. Even if the bulk buyers are not clamoring for a mirrorless, any company with an eye to the future is already evaluating the mirrorless alternatives. If Nikon releases a flagship mirrorless and Sony updates the A9 III (at least they already have an offering) Canon can't afford to miss the early evaluation cycle. Canon seems to keep things pretty close, so they would not want a lot of evaluation units in the field. My guess is that most of the bulk orders for the EF 400 & 600 III have already been placed and release of an R1 won't have that much impact. I believe there was a rumor that some of the RF mount big whites would be out this year, so that should take care of your concerns about RF lenses. If not, expect a weather sealed RF - EF adaptor to allow a transition.

I don't think Canon will wait until 2024 to release an R1, although it may not start shipping until 2022 based on the factory delays.

I would love to have one for my Namibia trip this year!

 

[dolina]

The R1 is not really a follow on for the 1dx III, it is a transition platform to new technology direction. Canon released the R5/R6 to establish the beachhead and prove the technology. Even if the bulk buyers are not clamoring for a mirrorless, any company with an eye to the future is already evaluating the mirrorless alternatives. If Nikon releases a flagship mirrorless and Sony updates the A9 III (at least they already have an offering) Canon can't afford to miss the early evaluation cycle. Canon seems to keep things pretty close, so they would not want a lot of evaluation units in the field. My guess is that most of the bulk orders for the EF 400 & 600 III have already been placed and release of an R1 won't have that much impact. I believe there was a rumor that some of the RF mount big whites would be out this year, so that should take care of your concerns about RF lenses. If not, expect a weather sealed RF - EF adaptor to allow a transition.

I don't think Canon will wait until 2024 to release an R1, although it may not start shipping until 2022 based on the factory delays.

I would love to have one for my Namibia trip this year!

That's why I have strong doubts about the R1.

R5/R6 are the direct replacement of the 5D4/6D2 as they're spaced out by 4 & 3 years respectively unlike the R1 that will be spaced out by 13-24 months?

 

[TravelerNick]

I would imagine that Canon will keep SDI for their cinema cameras as a differentiator. Although SDI connector locks well, it also takes up a lot of real estate on the body. A full sized HDMI would better suit the form factor IMHO

On a camera this size is the SDI size really going to be an issue?

The problem with HDMI on this type of camera is it's not really designed for field use. A camera with extreme weather sealing and all the other hard life features this sort of market demands might/will be better off with something other than HDMI.

 

[Michael Clark]

My guess: 3 1/2 years until summer olympics Paris 2024 as Tokyo 2020/21 will be canceled because of COVID-19.

As for winter games I have hope for winter 2022 if the vaccines can be distributed fast enough...

BINGO!

 

[Michael Clark]

With a quad pixel sensor, a 20 MP sensor is sending 80 MP worth of data to the processor, a 30 MP would be possible, but quad pixel is not a variant of a R5 sensor unless it has quad pixel already there. A global shutter requires memory for each pixel, so a layered sensor would be likely to do all of that. It would bump up the cost. The heat generated from all that is a issue as well, so 8K seems unlikely.

You are assuming one of several possible interpretations of quad pixel.

What if "quad pixel" were to be actually much like dual pixel except that a significant number of the the dual pixels used for AF are vertically oriented instead of horizontal? In the same way "cross-type AF points" in DSLRs are actually a horizontal set and a vertical set of AF lines superimposed over one another? One set of vertical and one set of horizontal dual pixels would give the appearance of quad pixel AF capability. This is then greatly simplified if the raw files only record to composite values for each dual pixel group (the way most DP cameras record raw files) instead of recording the values of each subpixel the way the cameras capable of dual pixel raw can. Since the 1-series has been significantly lower resolution compared to the 5-series since the 1DX II/5D IV were introduced in 2016, I don't think it would be that big a deal to most potential buyers.

 

[Michael Clark]

Reason why I do not believe why a 1-Series MILC will come out in 2021 as of this writing

• Canon Inc is a for profit company that will produce products in line with customers needs/wants so long as it is profitable
• Customers of the 1-Series like bulk buying photo news agencies surveyed between 2017-2020 wanted an SLR and not a MILC
  
  It's not early 2020 any more. The disruption caused by the SARS-CoV-2 virus and the release of the R5 have totally changed the game from what it was in early 2020.
  
  
• EF 600mm & 400mm were updated to Series III in 5-Sep-2018
  
  But the EF 300mm f/2.8 L IS II and EF 500mm f/2.8 L IS ii were not. I wonder why...
  
  
• 1D X Mark III was announced 378 days ago and was initially released in some markets on February 2020 while others got it as late as July 2020
  
  Everything was disrupted by worldwide events in 2020. Not releasing the 1D X Mark III in some markets until five months after it was released in others was not a marketing strategy Canon preferred, it was forced out of necessity due to events far beyond Canon's control.
  
  
• Market for flagship bodies is too small to support more than 1 "new" one at the same time
  
  The market for flagship bodies is larger than a single manufacturer, particularly when talking about moving from one mount system (EF) to another (RF). Canon manufactured the FD flagship 'New F1' until 1992, sold it until 1994, and supported it with factory service until 2004. The first EOS flagship, the EOS-1 was introduced in 1989 and replaced by the EOS-1N in 1994, the same year the FD mount 'New F1' was discontinued. Sales of ILCs in the period between 1987 and 1994 were not any higher than the current market.
  
  
• Having 2 "new" flagships at the same time would halve that market, doubles the R&D cost and would not pay for the R&D of last year's model
  
  Much of the R&D cost incurred in the development of the 1D Mark III has been applied to the R5/R6 as well and will presumably also be applicable to the R1. Claiming "double" R&D costs for using the same technologies in more than one model camera is either totally ignorant or willfully dishonest.
• 1D X Series sees a replacement every ~4 years to coincide with the Summer Olympics that occurs months later
  
  Both the camera industry and the Olympic games have had their timelines totally disrupted by the emergence of the SARS-CoV-2 virus in 2020.
  
  
• Not enough RF mount lenses
  
  Yet, but it is getting pretty close with the exception of beyond 200mm.
  
  
• No 300mm, 400mm, 500mm, 600mm or even 800mm are available now
  
  Now is not the end of 2021, it's currently the beginning of 2021.
  
  
• This year's Olympics is 6 months away and there is not much lead time for the R1 to come out.
  
  That's assuming they will go on as rescheduled. There's already a LOT of doubt that will happen. In which case the next Olympic games will be the Winter 2022 games if they are able to be held.
  
  
• The wants/wishes of outliers like those on a photo forum does not reflect the totality of the photographic market
  
  Neither do your prognostications.
  
  
• Rumors of flagship models gets a lot of clicks. How many times has CR published rumors of a Series II 800/5.6 or 200/2.0?
• Back orders for the R5 & R6 are high and are produced in batches on demand.
  
  Canon knows far better than you or I how close they are getting to catching up with demand of the R5/R6 bodies..
  
  
• Rescheduled 2020 Summer Olympics occurring in July 2021 may even happen
  
  It's looking less and less likely by the week as many of the governing bodies in many countries for the various sports to be represented at the Summer Games are not going ahead with plans to hold qualifying meets to establish teams from each country for each sport.

Will the R1 be priced at

• $4,499 to compete with the Sony a9 II?
• $6,499 to compete with Nikon and Canon's 1D X Mark III?
• $7,149 to cover any development cost projected reduction of economies of scale?
  
  How is price speculation applicable to release date? Canon will price whatever products they release based on market conditions at the time theose products are released, not based on market conditions before they are announced/released.

As courtesy to people who do not give any citation for their counter position I did a bit of snooping for you

• 549 days is the shortest product cycle between 1-Series bodies
• This occured 22-Feb-2007 to 20-Oct-2009 from 1D Mark II to 1D Mark IIn
• As applied to the 1D X Mark III the R1 would have come out 22-Jul-2022, 2 days before the Olympics
• An R1 in 2021 could happen but it is unlikely to occur until Q1 2024

I feel safer with FAANG

Idealized R1 release would go like this

• 7-Jan-2020 would announce the R1
• 2018 would announce the RF mount 600mm, 400mm, 70-200/2.8 IS, 70-200/4 IS, etc.
• 2018-2020 would announce 200/2.0, 300/2.8, 500/4 or even 800/5.6

To be honest I was surprised Canon even bothered offering the 1D X Mark III because of the RP, R and Ra bodies then it dawned on me that those buying a 1-Series body are very conservative with their mission-critical hardware so prefer SLRs over MILC.

 

[Michael Clark]

My speculation is based on precedence. Your point of view has little substance.

Perhaps grounding your position on evidence or even some data may make your words be more compelling to be taken seriously.

Most precedence went completely out the window in 2020 with emergence of the SARS-CoV-2 virus and the economic fallout from that.

 

[Michael Clark]

How come everybody salivates at global shutter for video and nobody mentions the 800lb gorilla in the room: Flash?

Sync at all shutter speeds. Fill flash across 100ft distances with fast big flash units.....

My thoughts exactly. Can't wait to use on camera fill flash on a sunny day for group portraits at 800mm.

There would still be the limitation of the length of time needed for a full power dump by the flash. The larger the flash, the longer a full power dump takes. So while sync would be much easier to manage, available power would still diminish with shorter Tv as the flash would become more like a continuous light that stays on for longer than the total exposure time.

 

[Michael Clark]

One does not downsample to RAW

Essentially that is what a standard raw file from any Canon camera with DPAF is. Each half of each dual-pixel is combined in the raw file. Only when one records the files as "DP Raw" files are the values saved separately.

 

[Michael Clark]

If its a sports camera I doubt it will have 8k (45mpixel). I bet its more like 30mpixel and 6k or something. Though I might be wrong. Since they managed to get 45mpixel with 20fps out of the R5, it may be very well possible for the R1...

Coming from the R5 I mostly hope for:
- a full size HDMI port
- 4k50 and 4k100 without overheat and with audio
- 1080p480
- more physical buttons (especialy a mode dial, dedicated ISO button and WB button)

The digital 1-series has no mode dial, all the way back to the original 1D in 2002.

 

[Michael Clark]

Another point is the lack of RF L lenses that is normally paired with a 1-Series body.

Canon disclosed years ago that 1-Series bodies make up 1% of all EOS bodies sold.

R1 body at more than $6,499 would be almost double the price of a Sony a9 II.

They were selling Rebels hand over fist when they made that less than 1% statement, too. The lower end of the ILC is drying up fast. The 1-Series will need to be more than 1% of Canon's future market for the 1-Series to survive.

 

[Michael Clark]

But, you would still be watching that video on a smart phone display, where no one can tell it was shot in 8K. Plus, in 25 years you will be divorced and the last thing your new wife would want to look at would be your wedding video with wife #1.

If he always has to be right in real life the way he does here, he ain't ever getting married.

 

[Michael Clark]

It won't be called "R1"
Name one other pro flagship camera not called EOS-1(something)...not to mention over three decades of precedent.

It will be the EOS R1. What's so hard to understand about that?

The EOS 5D Mark IV became the EOS R5
The EOS 6D Mark II became the EOS R6

That's before one must consider that the EOS EF mount numbering system was not the same as the FD mount model numbering system.

The EOS-1 was the EF mount equivalent of the New F1 (Yes, "new" was an official part of the model name) that was built like the T90. Take your pick.

When Canon went from the FD mount to the EF mount they went from letter-number model names to number-(sometimes letter) model names. When they went from film to digital, they added the "D" after the numbers. When they went to the RF mount they placed the R before the numbers.

It's not that hard to understand. It really isn't.

 

[Michael Clark]

Why such a focus on the Olympics? Certainly the FIFA World Cup is an equally large event for sports photography and it comes around in 2022, which may work better for the timing of the R1 release.

It's not a marquee event concentrated in one locale/region over only a 2-3 week period with multiple sports having events every morning, afternoon, and night. There are nowhere near the same number of total events, much less that number of events in only a 2-3 week period.

The FIFA world cup, including the qualifying rounds, is spread over months at various locations around the world. The tournament involves 31 games spread over about a month for an average of one game per day. The Summer olympics has at least 30 events every single day for 2-3 weeks spread over every summer sport imaginable. It's not the same at all.

 

[Michael Clark]

I prefer the moniker "1R" to accentuate that the one series as being more important than RF mount.

Just like the FD mount F1, right?

 

[Michael Clark]

I have a bunch of Hensel Expert D 250 SPEED which are able to
fire 64 Ws at 16fps while at the same time maintaining very short
flash durations. 64 ws is pretty much what a 600EX-RT fires at
full power. Coupled with an inverter this is still very transportable.
The Hensel Power Max L inverter will give you more than 10,000
pops at 64 ws with this baby and one single battery charge.
No overheating, of course.

Due to the ultrashort flash durations and extremely fast recycle times
this is good for fill flash 1 f-stop below bright sunlight at 1/4000 sec / f=2.8
across 60ft distance.

Tested successsfully up to 1/2000 sec with the leaf shutter of a G5X.
Just needed to wait for a reasonable fast camera with global shutter.

The flash is there, proven to work as intended and already in my bag.
Now bring on the camera.

64 Ws? That's puny compared to the current battery powered flashes offered by ProFoto, Godox, and others. 500-600 Ws is the current standard. Even the Godox "pocket flash" AD200 series puts out 200 Ws.

 

[Michael Clark]

If I am correct a global shutter also enables flash sync at any speed without having to switch to HSS mode which is great for when shooting action and provides much clearer sharper photos than HSS mode with its pulsing method. This would be great.

Only if the flash is fast enough to keep up. The more power a flash has, usually the longer the duration of the flash to do a full power pop.

 

[Michael Clark]

Yep, that's what it's all about.

 

[Michael Clark]

Those are not several things but 6.

R1 was not released as the surveyed bulk buyers and other 1-Series customers from 2017-2019 prefer SLRs.

That may have changed within the last 52 weeks but the question I'd pose is is Canon willing to take the financial lose of Osborning the 1D X Mark III and Series III EF 600/4 & EF 400/2.8?

Now, if Canon released a RF 600/4 & RF 400/2.8 in 2018 instead and R1 in 2020 instead then that would have been better.

I have no doubt that the MILC tech for the R1 has been there for more than 2 years already but what is needed to support a 1-Series body is the appropriate lens and accessories for it. Thus far only the three L f/2.8 zooms are there. No white long fast primes.

Nikon themselves admitted that their transition to MILC was slow. By my estimate by 2 years. Canon was late by 1.

Sony timed it perfectly. Their FE 400/2.8 came out in 2018, FE 600/4 and a9 II came out in 2019.

What if the EF 600mm f/4 L IS III and EF 400mm f/2.8 L IS III are really the RF 600mm f/4 L IS and RF 400mm f/2.8 L with EF housings on the back end?

Nothing past 200mm gains any advantage from the shorter registration distance of the RF mount. None of Canon's Super Telephoto EF lenses have rear elements that come close to the flange ring.

Sigma, Tamron, etc. do this kind of thing all of the time.