Canon EOS R5 Mark II sensor resolution likely to stick at 45mp but with new AI features [CR2]

[Del Paso]

I had hoped for a camera with at least 60 mpx.
So if it really is 45 mpx it represents, for me, a useless purchase.
It would be the trick that breaks the camel's back after the half-disappointment of the new 24-105/2.8 which requires software corrections to remedy the terrible distortion and chromatic aberration.
I'm honestly thinking of abandoning Canon to switch to Leica SL (SL3 will be released soon with 60 mpx and 15 exposure latitude apertures) with its Apo lenses.

I understand, the SL 3 is tempting.
What deterred me from buying it, is that some lenses I often use, like TS and longer-zooms, are either missing (TS), or made by Sigma or Panasonic. The choice of genuine Leica SL lenses is reduced.
And I do not buy a Leica to fit it with Sigma lenses. Leica is about lenses, much more than about bodies. The R5 is much more advanced than the SL 2, especially in its AF quality. As to the Apo lenses, which are truly exceptional, I own some in Leica M version.
But I must confess, I was tempted by the SL 2 and still am by the coming SL 3...If only they were as reliable as they should be. For Euro 6800 I expect total reliability...

 

[Walrus]

Is that for new users? I'm still paying what I paid in 2013 for the LR/PS plan.

View attachment 215124

Same here, since 2015 I'm still paying the exact same amount of 107,34€ each year.

 

[Del Paso]

Good luck. I hope Leica lives up to your standards.

There have been many issues with the Leica SL and SL 2. They are not at all, apart from pricing, on the quality level of the film era Leicaflex SL and SL2.
The R3 & R5 are much more advanced and better bodies, in almost every feature, though optically not as well finished.
If it were not for the Apo lenses...

 

[gdanmitchell]

Reading your note abut the upcoming R5ii. Sticking to a 45Mp sensor would be disappointing. I know… not everyone really needs the highest possible sensor resolution, and some favor things like fast burst rates more. But there are other cameras for that. The R5 is the ostensible successor to the 5DsR, yet it was already a (small) step backwards in sensor resolution. At a time when competitors are producing higher MP sensors on full frame (and over 100MP on miniMF), it is hard to imagine what AI, etc. developments would make up for the loss of appeal to those of us who print big and do value higher sensor resolution.

I remain hopeful that earlier rumors of higher resolution sensors will still an out in the end. I’ve been sticking with my 50MP 5DsR, waiting to the expected higher resolution R5ii… but I’ll be considering other options if it remains at 45MP.

 

[David - Sydney]

They're all wrong. The correct number of megapixels is 42. Because that's the answer to the most important question.

And it has been roughly 42 years since Hitchhiker's Guide to the Galaxy came out!

 

[Skux]

This thread is crazy lol, I'm just here doing paid gigs with my 20mp R6

 

[masterpix]

The Canon world is preparing for the lauch of two massive cameras, the EOS R5 Mark II and EOS R1. We expect to hear official word about both cameras by the end of May 2024. We have reported in the past that Canon would move the EOS R5 Mark II to 60mp, but we have

See full article...

45MP is enough for most things, of course modeling and architecture might need more than that. But for most photographers, 45MP is far more than enough. I wish the ned R5 will have faster FPS, wider dynamic range and most of all, far better AF tracking for fast moving objects. Beside that, the R5 is amazing camera.

 

[BBarn]

This thread is crazy lol, I'm just here doing paid gigs with my 20mp R6

And I'm sitting here as a hobbyist liking the R6 better than the R5, and thinking I'd likely buy an R6II over spending considerably more for an R5II or R1.

 

[EOS 4 Life]

But the camera was entirely incapable of "learning." That was the deceptive part of the marketing jargon.

"Learning" in this case the ability to be trained.
Machine Learning does not mean the ability to learn on its own.

 

[EOS 4 Life]

Indeed. On a tangential note, I shudder to think that we could reach a point where it is difficult to distinguish between machine-generated and photographer('s skills) generated photo. Imagine a camera that can automatically recompose, crop, adjust the colours (they already do somewhat) according to the genre, remove noise, add/substrate DoF, etc based on personal preferences (that the AI learns about the user). I do not know whether to welcome this, toss out photography, or go back to film.

We are already at the point where a human can make deep fakes.
Generative AI is not really necessary to achieve that.

 

[EOS 4 Life]

Why do we have to give up the current model where we paid for our camera and it will work until its circuit boards fail? Why do we have to pay for something, then continue to pay for it to be able to use it?

Brian

A lot of professionals will not use a camera that is no longer supported by the manufacturer.
Paying for extended support would be a form of subscription.
It was reasonable for Canon to stop supporting DSLRs with mechanical shutters after a limited period of time but mirrorless cameras with electronic shutters could potentially last forever.

 

[EOS 4 Life]

I guess my musing is whether AI in photography will get to the point where the photographer's input becomes miniscule or to the extent that it is largely irrelevant. On the other hand, photography might evolve into a form that cannot be imagined today, and we evolve with it in tandem. Meanwhile, I will enjoy its present form or something close enough where the photographer's input matters.

I am not sure that would be a bad thing.
There are still a ton of skills that a photographer needs to use.
Operating a camera makes one a camera operator not a photographer.

 

[EOS 4 Life]

I guess the other piece that would be interesting to consider if the R5ii moved exclusively to a subscription model, is it would likely change the economics of how some people buy cameras - you wouldn't be able to sell your body to recoup some of that investment when you buy a new body. It could have long-lasting impacts on the cost of ownership over time depending on how you manage camera purchases.

There would be less of a reason to buy new bodies if old ones had the latest features and upgrading to a new body could be cheaper or even free.

 

[EOS 4 Life]

For me, it was straight forward: buy a new LR each year for €99 or pay €96 a year and get photoshop as a bonus.

I might prefer that as well but I would have preferred they game me the choice.
Instead, I left Adobe entirely.

 

[Del Paso]

I am not sure that would be a bad thing.
There are still a ton of skills that a photographer needs to use.
Operating a camera makes one a camera operator not a photographer.

"Operating a camera makes one a camera operator not a photographer."
So true, so true!

 

[danfaz]

"Operating a camera makes one a camera operator not a photographer."
So true, so true!

True, indeed!
I made a similar comment on IG a while ago:

 

[Fischer]

This thread is crazy lol, I'm just here doing paid gigs with my 20mp R6

And clearly your specific needs is everyone else's'...so much for craziness

 

[David_B]

5D MarkII killed the 1Ds Mark III. Canon won't make that mistake again.

 

[David_B]

What about if tethered shooting was a subscription service? Or being able to do in-camera edits was a subscriptiion fee? Or Generative AI was a subscription add-on to EOS Utility?'Think how does Canon's camera package compete with Google Pixel and magic erase? How does Canon sell more cameras to people when someone with a Google Pixel can do all of these AI feature things to make their pictures look like they want but your expensive Canon camera can't do any of that out of the box? If AI-AF (heh, AiAF!) was a subscription service then you can bet that the people they're targeting are those at the Olympics and other sporting events. Soccer mums and weekender BIFs will just ... IDK... maybe if you could program your Canon camera to prefer to focus on pictures of certain people (ie your kid, so that when you're at their soccer game, it preferentially focuses on your kid and not someone else's) then maybe parents would see added value in THAT. Now it might be that if you can train the camera to discver your kid then you can also train the camera to more easily pick out lions, gorillas, or birds. If the camera knows whats in front of it then it might also make better suggestions in auto-mode, for example if the camera can work out the subject is a person then it might choose an f-stop that'll give at least 6-7 inches for DoF, or more, so that your subject is properly in focus while standing up at the wedding photo shoot.

AI could bring lots of interesting features to a camera, I think the mindset here is very 20th century about what it oculd offer.

I can't see Canon doing $1/1000 photos...

 

[David_B]

I think Adobe is the reason a lot of companies are trying the subscription model. It has worked for them, but I’d argue that the reason it works for them seems to be missed by a lot of companies trying to replicate the model. I’m no expert, but I’ve always thought a lot of people stick with adobe because they have numerous products which especially graphic designers may need.

Adobe are the big name in town, with mature products that people like to use. That is why people flock to them.

But there's another side to the subscription story that Disney+ and others are finding out: people only pay a subscription for as long as they use it.

What does that mean for my Canon camera? Well if it is $1000 less to buy but there are subscription things I can buy per-month, then I might only buy a 1 month subscription once or twice a year if I only use my camera whien I go on vacation. That Adobe subscription, I might use that all year round becuase I can sit at home and edit photos at almost any time, but my camera, why would I buy a year's subsrciption if I only pull it out to use a couple of times a year?