There are still surprises in store for the Canon EOS R5 announcement [CR2]

[mattamaxwell]

This is my first comment but I wonder if Canon could be introducing computational photography tricks that we have seen on our phones over the past few years. Does anyone have thoughts on that? I keep waiting for a big camera manufacturer to attempt what Apple and Google are doing with their tiny little sensors.

 

[Cryhavoc]

I thought exactly the same ! let's hope that is something more than this, but is already fine

and several other mirrorless camera's.
My bet is it's pixel shift as others have mentioned. Only because not one word has come out about pixel shift.

I think Canon takes it to 11 and gives the shooter the ability to determine how many shifts they want the camera to make.

 

[privatebydesign]

This is my first comment but I wonder if Canon could be introducing computational photography tricks that we have seen on our phones over the past few years. Does anyone have thoughts on that? I keep waiting for a big camera manufacturer to attempt what Apple and Google are doing with their tiny little sensors.

God I hope not! I’m sure I’ll get told off for being behind the times or traditionalist, or old, or something less kind, but we already have ‘tricks’ like cooking the RAW files to adjust for wild distortion in lenses and clumsy high iso noise reduction that can’t be turned off. I want to be in control of any computation, how can we realize our vision if ‘artificial’ stuff keeps getting put in our way?

 

[mattamaxwell]

God I hope not! I’m sure I’ll get told off for being behind the times or traditionalist, or old, or something less kind, but we already have ‘tricks’ like cooking the RAW files to adjust for wild distortion in lenses and clumsy high iso noise reduction that can’t be turned off. I want to be in control of any computation, how can we realize our vision if ‘artificial’ stuff keeps getting put in our way?

I hear you and agree but I am also curious to see what could be on the horizon for the future of photography. IF they introduced something like computational photography, surely they would give you the option to turn on and off.
Ive been very impressed with what Apple and Google accomplished with their cameras and, as I said before, what could canon accomplish with some of this cutting edge technology?

 

[SteveC]

At the risk of beating a freshly dead horse, I can see four dials in Fv mode quite readily. One for ISO, one for aperture, one for shutter speed, and one for exposure compensation. Why have an apparently redundant one for exposure compensation, when you can get the same result twiddling one or more of the other three parameters?

It's because in Fv mode there's a tendency to leave at least one of the other three in auto. So, say, you fix the shutter speed, and let the other two be in auto. But you find yourself with a dark exposure at some point, and rather than override an auto setting, you just crank the exposure comp. You can then set it back to zero and your other settings are unchanged, with aperture and ISO still on auto. This works with any combination of manually set/auto (as long as something remains auto). Yes, technically you could do it with fewer dials...for example some cameras when you're in Av a particular dial controls aperture, and a different dial controls exposure comp, when in Tv that first dial now controls shutter speed, and the second dial is EC, when in M, now suddenly the first dial controls shutter speed and the second dial controls aperture, or is it the other way around--the point being the two dials have different meanings depending on what mode you're in. And ISO is controlled yet somewhere different (hopefully not in a menu!!!) Having four dials that ALWAYS have the same meaning does mean one of the dials will be useless at any given time, but at least you don't have to think about "Now which dial is it that controls EC now? Whoops, never mind, I'm in manual mode, none of them do." or "Hmm, which one controls shutter speed? The answer differs between M and Tv."

 

[privatebydesign]

At the risk of beating a freshly dead horse, I can see four dials in Fv mode quite readily. One for ISO, one for aperture, one for shutter speed, and one for exposure compensation. Why have an apparently redundant one for exposure compensation, when you can get the same result twiddling one or more of the other three parameters?

It's because in Fv mode there's a tendency to leave at least one of the other three in auto. So, say, you fix the shutter speed, and let the other two be in auto. But you find yourself with a dark exposure at some point, and rather than override an auto setting, you just crank the exposure comp. You can then set it back to zero and your other settings are unchanged, with aperture and ISO still on auto. This works with any combination of manually set/auto (as long as something remains auto). Yes, technically you could do it with fewer dials...for example some cameras when you're in Av a particular dial controls aperture, and a different dial controls exposure comp, when in Tv that first dial now controls shutter speed, and the second dial is EC, when in M, now suddenly the first dial controls shutter speed and the second dial controls aperture, or is it the other way around--the point being the two dials have different meanings depending on what mode you're in. And ISO is controlled yet somewhere different (hopefully not in a menu!!!) Having four dials that ALWAYS have the same meaning does mean one of the dials will be useless at any given time, but at least you don't have to think about "Now which dial is it that controls EC now? Whoops, never mind, I'm in manual mode, none of them do." or "Hmm, which one controls shutter speed? The answer differs between M and Tv."

I understand the point, I just feel it is overly complicated as in no situation do you need more than three control inputs but generally you only need two. If you are in auto you drop one metric but gain EC, if you are in manual you don’t need EC.

I like the way when I am in P, Av or Tv my back dial becomes EC, easy muscle memory. I also like that when I am in ‘manual’ with auto iso I use the same fingers in practically the same place. I push ‘Set’ with my thumb and move the top front dial. The problem with more dials in more places is it becomes more difficult to use them universally, a top plate mounted EC dial on the left of the camera works great on the M5 with small lenses but on a full sized camera with big lenses your left hand is normally far away from the body, same with ‘control ring’ old fashioned aperture rings, they work great in some situations and really get in the way in others.

Personally I use my right index finger for the shutter button the top dial and the iso button, my right thumb for back dial ‘set’ button and focus joystick. I have seen more dexterous people than me use right index finger for iso and top dial and right middle finger for shutter button, this allows instant control of all four/three metrics instantly. Maybe that is why they don’t mess with 1 series ergonomics much.

 

[Richard Anthony]

I just read on Sony Alpha Rumours that the new R5 is going to have a 5 million dot EVF , and the new Sony A7 is going to a 9.44 million dot EVF , lets see if this is true .

 

[highdesertmesa]

diminishing returns are certainly present chasing the high DR title.

I get the fact that having the highest dynamic range isn’t the end-all be-all feature on the wishlist. But there is a lot of room for improvement for the color and contrast in lifted shadow areas. It’s not about how many stops I can lift the shadows, it’s about how close can I get to lifted shadows that look completely natural as if I exposed for them directly. That’s the holy grail of dynamic range as I see it.

 

[navastronia]

I use the R in manual without an issue and I don't even need to use the control ring on the lens to do so. Really not hard at all

Yes it is. The R doesn't handle as well as bodies with a scroll wheel.

 

[highdesertmesa]

No, the point was all manufacturers are so close in high iso because current technology is limited by laws of physics. Until there is a substantial change in technology, a revolutionary change that hasn’t been demoed or patented anywhere, or the laws of physics are negated there is practically little improvement to be had at high iso.

Demanding a stop or two of improved high iso performance is akin to wanting your next family car to go 1,000 mph, yes there might be a car or two that can go that fast but you wouldn’t want to pay for it and even if you did it isn’t practical.

Have you watched Canon’s video on dual-gain sensor tech and what it can accomplish? It sounds to me like the answer to the current limitations. Video is down on the page a ways:

Could the Canon EOS R5 Receive a Dual Gain Output Sensor?

Let's explain Canon's new Dual Gain Output sensor.

nofilmschool.com

 

[privatebydesign]

Have you watched Canon’s video on dual-gain sensor tech and what it can accomplish? It sounds to me like the answer to the current limitations. Video is down on the page a ways:

Could the Canon EOS R5 Receive a Dual Gain Output Sensor?

Let's explain Canon's new Dual Gain Output sensor.

nofilmschool.com

Of course I have, are you missing the point of it? It increases the dynamic range at low (base) iso it does nothing for higher iso readout DR. The additional readout is at 5,000 iso and is for better shadow detail when sampled at the same time as the native 800 iso.

 

[degos]

In order to minimise cost, the only supported shooting mode is BULB.

However a Canon-branded digital stopwatch is included in each kit SKU ( not in body-only SKU ).

 

[highdesertmesa]

Of course I have, are you missing the point of it? It increases the dynamic range at low (base) iso it does nothing for higher iso readout DR. The additional readout is at 5,000 iso and is for better shadow detail when sampled at the same time as the native 800 iso.

It still works across all ISO ranges, just not as effectively as it does at base. And I would expect Canon has improved it for the R5 versus the C300 III. It may even work differently for stills, we just don’t know yet.

 

[privatebydesign]

It still works across all ISO ranges, just not as effectively as it does at base. And I would expect Canon has improved it for the R5 versus the C300 III. It may even work differently for stills, we just don’t know yet.

You are missing the point, it isn’t improving high iso performance it is combining current high iso performance and low iso performance to give a greater range at low iso. If both samples are in the high range you gain nothing.

Sure we don’t know what they are doing, but the area of high iso performance has nudged up against physics barriers and has been pretty constant for some time.

 

[Kit.]

Have you watched Canon’s video on dual-gain sensor tech and what it can accomplish?

The only thing dual-gain output sensor accomplishes is great increase in sensor readout speed at the cost of very small loss of DR.

Good for reduction of rolling shutter effect in video. Doesn't really matter for stills taken with mechanical shutter.

 

[highdesertmesa]

You are missing the point, it isn’t improving high iso performance it is combining current high iso performance and low iso performance to give a greater range at low iso. If both samples are in the high range you gain nothing.

Sure we don’t know what they are doing, but the area of high iso performance has nudged up against physics barriers and has been pretty constant for some time.

Dual Gain Output on the C300 III

• 16 stops of DR at ISO 800 (base)
• Effectiveness is lowered at higher ISOs (they mention 14 stops), but it still is effective/working

Canon's Dual Gain Output Sensor Explained - Newsshooter

Canon has been receiving a lot of emails and inquiries asking about how their Dual gain Output sensor actually works. So they made a short video to try

www.newsshooter.com

The DGO sensor produces images optimized for HDR by creating two images. One prioritizes saturation in the highlight areas and the other prioritizes lower noise in the shadows and dark areas. Essentially each photodiode is being read out at two separate gains. This creates two signals for each photodiode which are then blended to produce an image that has both high dynamic range and low levels of noise.

Canon claims that this new sensor in the C300 Mark III is capable of 16+ stops of dynamic range with reduced noise in the shadows and improved low light performance...

...Noise is also claimed to be greatly reduced because the Dual Gain Output signal is taken directly from the sensor and before any image processing takes place...

...The DGO works across all ISO ranges on the camera... [just not as effectively off base ISO]

But this is how it works now and on a video-only camera.

I was just trying to make the point in the original post that Canon is going to greatly improve high ISO capability – either that or they are telling you to go suck it by releasing f/11 lenses. Heck, maybe they are.

 

[TM]

I hope it is a new Dual Gain Output Sensor

Yes!

 

[privatebydesign]

I was just trying to make the point in the original post that Canon is going to greatly improve high ISO capability – either that or they are telling you to go suck it by releasing f/11 lenses. Heck, maybe they are.

I was just trying to make the point that at this time there is no consumer grade technology capable of "greatly improving high ISO performance".

You keep pointing to something that does not do what you think it does and by design can't work where you say it does work.

 

[highdesertmesa]

I was just trying to make the point that at this time there is no consumer grade technology capable of "greatly improving high ISO performance".

You keep pointing to something that does not do what you think it does and by design can't work where you say it does work.

And I’m saying the DGO in one form or another could get us closer, especially if optimized for stills. If you’re looking for hard evidence, maybe a rumor site on a thread about unknown “surprise” specs isn’t the best place to be asking someone for it. This still falls under “killing time until the 9th”, lol.

 

[highdesertmesa]

The only thing dual-gain output sensor accomplishes is great increase in sensor readout speed at the cost of very small loss of DR.

Good for reduction of rolling shutter effect in video. Doesn't really matter for stills taken with mechanical shutter.

On the C300 III they turn off DGO for faster sensor read. Speed is only half the story (pun).