Do you expect an EOS 5D Mark V anytime EVER?

davesrose wrote in post #18993277
Quite a few cell phones now have that with HDR mode that does that automatically....but of course overall quality won't be as good as if you are experienced and using a dedicated camera with settings you know.

As for the 5D4, keep in mind that it's DR is improved and isn't as dependent with ETTR. I think most users are experienced enough to know to judge such in the histogram. It doesn't have the noise or banding as previous models when raising shadows. Also, it seems folks either don't know or are overlooking the fact that the 5D4 has a feature for judging brightness after the fact the RAW was shot (RAW image processing).

interesting. Does the ‘raw image processing generate the after-shot LCD histogram and blinkies?

Pippan wrote in post #18993312
interesting. Does the ‘raw image processing generate the after-shot LCD histogram and blinkies?

It does generate a new preview that interactively shows less blinkies (so I know I'm completely safe without any blinkies taking brightness down).

I did not think Canon would release a 1Dx3- I was wrong. They did and its a souped up version of a 1Dx2

I just dont see a 5D5 in the line up. They are too invested in the R line up for glass and I think they will release a EOS R pro version that will effectively kill off the 5D line up for good

Just dont see a 5D5 at all with the amount of RF glass they have introduced. I think they introduce something that competes for the A9 and A7iii markets

umphotography wrote in post #18996033
I did not think Canon would release a 1Dx3- I was wrong. They did and its a souped up version of a 1Dx2

I just dont see a 5D5 in the line up. They are too invested in the R line up for glass and I think they will release a EOS R pro version that will effectively kill off the 5D line up for good

Just dont see a 5D5 at all with the amount of RF glass they have introduced. I think they introduce something that competes for the A9 and A7iii markets

I don't know Mike, I respectfully disagree.

Canon is a huge company with a great deal of production capacity. Sure they've have put a lot into RF glass but they also have a huge following of 5D users since the first one came out in 2005. That was a breakthrough camera in many ways and has since captured substantial market share. Many chose that time to leave medium format film and finally go pro digital. A lot of people with older gear are waiting for a 5D5 - I'm one of them .

Canon is building a following with the R and also continue to enjoy sales with the 5D line. I really believe they can do both over a period of time. You just won't see new models out every other month. Of course this is all speculation.

What is fundamentally NOT possible about this occurring (not abandonment of EF line as so many predict)

1. putting the emphasis (100% of the corporate effort) at more fully completing the R line of lenses PLUS cameras (remember they still have not launched a 'pro R' body, while
2. simply milking the EF mount business with the existing full line-up of EF lenses
3. and at some point, when the complement of R mount accessories and lenses is broad enough to sustain the line
4. go back to developing key new EF mount bodies and selectively replacing some of the older EF mount lens designs, where significant improvements can be had

According to the 'obsoleting EF' bunch, BMW and Audi would have abandoned their sedans and convertibles for SUVs and built no M series sedan ever again (not an SUV)?! yet both coexist quite well...I would never buy an SUV, even though Mom with 5 kids could not transport her brood in a sedan. Same for dSLR and mirrorless...something for each buyer.
In the past few years there has been little 'revolutionary' technology...mirrorles​s is 'yet another way of taking pictures'...it has some strengths and it has some weaknesses compared to dSLR, after all; not the clear winner in all situations.

It is all related with autofocus an buffer capacity. 5D4 and 7D2 owners will never jump on mirrorless systems until they cannot match (at least) the current performance for these feautures.

BTW 1DX2 owner have got the 1DX3.

Tom Reichner wrote in post #18989579
.
I keep waiting for a foolproof exposure mode, and I think that perhaps a 5D V would be the camera that introduces it.

My foolproof exposure mode would be an expose to the right mode, in which the user sets an adjustment that determines how many "blinkies" are acceptable. . This, of course, would be based on the jpeg generated from the RAW file.

It would give as much exposure as possible, without exceeding the blinky limits that you set.

You could set it by gross number of blinkies - set it to no blinkies at all, or that a maximum of 500 pixels could be blinkies, or that a maximum of 1,000 pixels could be blinkies, or 1500, or ..... all the way up, in increments of 500 pixels.

Or you could set it on number of adjacent pixels that are blinkies. . In this mode, it wouldn't matter how many total blinkies are in the frame, but rather how many adjacent pixels are allowed to be blinkies. . You could set it so that no two adjacent pixels can be blinkies, or that no more than 2 adjacent pixels can be blinkies, or no more than 3, or 5, or 7, or 10, or 12, or 15, or 20, and so on.

Yes, I've desired this type of control for years. Manufacturers seem only interested in replacing film, though, as if ETTR did not exist or make sense.

A lot of the things that we wish for in cameras, but are absent from them, are not absent because they are difficult, but rather, because the manufacturers are being conservative and difficult. Take limits to shutter speeds in Av or P mode, or limits to f-number In Tv or P mode, or auto-ISO in M mode; look how long they took to get into cameras, but they are extremely easy to program. The latter is as simple as re-ordering a very simple equation; as soon as I realized that for a given exposure, my EOS 10D gave the least noise with the highest practical ISO (regarding highlight clipping), I envisioned auto-ISO in M mode, to take advantage of this. The equation just gets re-ordered so that ISO is alone on the left of the equation, instead of shutter speed, or f-number.

All you have to do in this specific request we have here is examine a histogram that isn't clipped too much already, and derive a new clipping point that meets your requirements, and take the next shot with the exposure that would put that at the actual clipping point of the RAW. I would guess that the most difficult thing about implementing new exposure paradigms is not the coding, but working the feature into the corporate roadmap and file metadata standards.

I would call that "optimal" exposure; "correct" implies meeting a specific exposure index target, like the ISO exposure index implied in the setting, or some pre-determined offset from it, like always shooting "ISO 100" at an ISO 64 exposure index that would give a good Kodachrome 64 slide (possible, actually, with almost all recent digital cameras, some of which can handle ISO 40 or even ISO 32 exposure, with even lighting and matte subject matter, if they have dual conversion gain, which increases full-well capacity at base ISO). With what we are wishing for here, the ISO exposure index becomes variable in the result. With the camera set to use ISO 100's gain, for example, the specular highlights in a chrome sculpture might result in an ISO 458 exposure, and a grey cat on a black couch might result in ISO 23.

sploo wrote in post #18989636
I do wonder though; how many pro 1Dx shooters are complaining? For sports and reportage I suspect the large resolution isn't a big pull, so maybe this is more of an issue for non-1D shooters who were thinking of taking the plunge?

Canon could be thinking more in terms of name brand visibility at things like the Olympics, where professionals might be concerned more with not missing moments with a throttled buffer, slow burst speed, or AF tracking that falls off target, or takes too long to acquire.

This is a "don't miss the shot" camera; not a state-of-the-art fine art camera, even if it should have fine IQ for a 20MP camera.

Wilt wrote in post #18992568
As sploo conjectured earlier, there is always gonna be a situation which is missed by the 'foolproof auto'. And that is the challenge of providing 'foolproof auto'...it isn't foolproof. Maybe 'foolproof auto' used with EC?!

People who are cynical about progressive paradigms often are quick to find potential problems, all the time ignoring the possibly greater problems and risks of older paradigms that they have just learned to accept. Take auto-ISO in M mode for example; people who are experienced with full-manual often ask "but what happens when your EC setting is not perfect when you suddenly have to take a shot", a problem which is only likely to have ISO off from ideal by a stop or less, while their beloved full-manual could be several stops off, switching from shade to sunlight, or visa-versa.

A control that tells what percentage of pixels can clip take the place of EC, and use the same dial, and the same way as always, you use experience as feedback to bias the way that you dial it in. Why do people assume that anything new is hopeless, takes over the photography process, and can't be worked with, and worked around, if necessary?

sapearl wrote in post #18996185
I don't know Mike, I respectfully disagree.

Canon is a huge company with a great deal of production capacity. Sure they've have put a lot into RF glass but they also have a huge following of 5D users since the first one came out in 2005. That was a breakthrough camera in many ways and has since captured substantial market share. Many chose that time to leave medium format film and finally go pro digital. A lot of people with older gear are waiting for a 5D5 - I'm one of them .

Canon is building a following with the R and also continue to enjoy sales with the 5D line. I really believe they can do both over a period of time. You just won't see new models out every other month. Of course this is all speculation.

Agreed. I was reading an interesting article on the 1DxIII from the Canon CPS website. They're claiming a 1 stop improvement in noise over the 1DxII, and prioritising the ISO3200-6400 range (where they believe their sports customers are using the camera most).

Also claimed is that the new low pass filter gives an effective 24MP (from a 20MP sensor); in the sense that the reduction in softening gives more detail, whilst still preventing moire.

If some part of the above were to trickle down to a 5D5, along with (an obviously slightly detuned) version of the new AF system, it would be a worthwhile step forward.

Hopefully they'd also give it at least one CFexpress slot (maybe CFexpress + UHSII SD); as that would greatly aid buffer clearing and max burst lengths. With luck, the new AF-ON joystick/mouse control would be present, and maybe another one or two fps max shooting speed.

I probably wouldn't jump on it immediately (like I did when the 5D4 was launched, over my 5D3), but it'd certainly be on the cards later.

John Sheehy wrote in post #18996457
People who are cynical about progressive paradigms often are quick to find potential problems, all the time ignoring the possibly greater problems and risks of older paradigms that they have just learned to accept. Take auto-ISO in M mode for example; people who are experienced with full-manual often ask "but what happens when your EC setting is not perfect when you suddenly have to take a shot", a problem which is only likely to have ISO off from ideal by a stop or less, while their beloved full-manual could be several stops off, switching from shade to sunlight, or visa-versa.

A control that tells what percentage of pixels can clip take the place of EC, and use the same dial, and the same way as always, you use experience as feedback to bias the way that you dial it in. Why do people assume that anything new is hopeless, takes over the photography process, and can't be worked with, and worked around, if necessary?

Not cynical, realistic! Early in my work career there was talk about voice input for control of computer program. It took over 35 years for voice recognition to become usable enough. I tried Dragon in the early 2000s, and it was not good enough for me. What we have, nearly 20 years later, works pretty darned good, but there are still a lot of manual corrections to computer dictation of documents and even text messages. And I don't have a heavy accent for the program to get past.
It took Auto ISO in Canon cameras how many years to become usable rather than overlooked 'feature'?!

John Sheehy wrote in post #18996442
Canon could be thinking more in terms of name brand visibility at things like the Olympics, where professionals might be concerned more with not missing moments with a throttled buffer, slow burst speed, or AF tracking that falls off target, or takes too long to acquire.

This is a "don't miss the shot" camera; not a state-of-the-art fine art camera, even if it should have fine IQ for a 20MP camera.

A fair point. 20MP of "the shot" is better than 30MP of a missed shot.

John Sheehy wrote in post #18996457
People who are cynical about progressive paradigms often are quick to find potential problems, all the time ignoring the possibly greater problems and risks of older paradigms that they have just learned to accept. Take auto-ISO in M mode for example; people who are experienced with full-manual often ask "but what happens when your EC setting is not perfect when you suddenly have to take a shot", a problem which is only likely to have ISO off from ideal by a stop or less, while their beloved full-manual could be several stops off, switching from shade to sunlight, or visa-versa.

A control that tells what percentage of pixels can clip take the place of EC, and use the same dial, and the same way as always, you use experience as feedback to bias the way that you dial it in. Why do people assume that anything new is hopeless, takes over the photography process, and can't be worked with, and worked around, if necessary?

My caution regarding the "percentage of clipped pixels" system is just that as soon as you offer such a feature, people will complain when it gets it wrong. Given that the camera can't know for sure what percentage of the capture will be clipped until it actually takes the image it can only take an (educated) guess, based on whatever level of detail is available with the current shooting mode's metering sensor.

Personally I'd love something like this - indeed there should be little point in having to select an ISO setting; the camera should just shoot with my chosen aperture and shutter speed and configure the sensor for whatever settings the manufacturer knows will get the best quality image in those conditions.

However, having worked for a major Japanese tech company; I've found the Japanese mentality to be very cautious when it comes to big changes. Maybe the Magic Lantern guys could do something - though I haven't kept up to date with what they've been doing.

sploo wrote in post #18996802
Agreed. I was reading an interesting article on the 1DxIII from the Canon CPS website. They're claiming a 1 stop improvement in noise over the 1DxII, and prioritising the ISO3200-6400 range (where they believe their sports customers are using the camera most).

That could be mostly software noise reduction, no better than what you might get in a 3rd-party software. I don't believe anything a manufacturer says about anything with as much play as this terminology has. We'll see when RAWs are available.

Sounds like corporate weasel-speak. What would they say if it had no AA filter at all? "Effective 28MP"? At the end of the day, there are only 20.2 million samples, perhaps with a little more contrast between neighbors in the RAW images.

sploo wrote in post #18996806
Personally I'd love something like this - indeed there should be little point in having to select an ISO setting; the camera should just shoot with my chosen aperture and shutter speed and configure the sensor for whatever settings the manufacturer knows will get the best quality image in those conditions.

However, having worked for a major Japanese tech company; I've found the Japanese mentality to be very cautious when it comes to big changes. Maybe the Magic Lantern guys could do something - though I haven't kept up to date with what they've been doing.

But, how does the camera choose the best ISO to use? How does it know if it is on a tripod? For example, in a given amount of light that results in these choices...

1. It might choose 1/60 f/8 ISO 1600
   
2. It might choose 1/15 f/8 ISO 400
   
3. It might choose 1/250 f/8 ISO 6400
   
4. It might choose 1/125 f/8 ISO 3200
   

• I might find ISO 1600 to be sufficiently low in noise to select that combination
  
• I might think that ISO 1600 is too noisy, and since my camera is on a tripod I do not care about camera movement at 1/15
  
• I might think that ISO 6400 is plenty satisfactory for noise in the 'relatively bright' conditions and I need a speed fast enough to handhold
  
• I might think that ISO 3200 with 1/125 f/8 is the OPTIMUM choice, given its somewhat conservative ISO with a sufficiently fast speed to handhold acceptably.
  

John Sheehy wrote in post #18996862
That could be mostly software noise reduction, no better than what you might get in a 3rd-party software. I don't believe anything a manufacturer says about anything with as much play as this terminology has. We'll see when RAWs are available.

Sounds like corporate weasel-speak. What would they say if it had no AA filter at all? "Effective 28MP"? At the end of the day, there are only 20.2 million samples, perhaps with a little more contrast between neighbors in the RAW images.

I think you hit the weasel on the head...how to get effective 24MP out of a 20MP sensor!