Canon announces development of the EOS R5 full-frame mirrorless camera

slclick

EOS 3
Dec 17, 2013
4,634
3,040
A really wide angle L prime like a 14 or a 24 would be a sweet little bone to throw the astro guys.

Like me
I would think it would be any of those 14,24,35 but not the 135. While the EF version is my favorite lens, it just doesn't seem to be a sales favorite what with stellar zooms that cover that focal length with almost or as good as primes IQ.
I say wide...14 and 24
 
Upvote 0
Last edited:
Upvote 0
I do not understand the point.
The point is it is very unlikely that 8K will have any crop or a large crop because it just isn't possible for smaller resolutions to create an 8K output. The reason being is that the width of 8K is either 7680 or 8192 (depending on the 8K standard we're talking about). A 40MP sensor would have ~7680 pixels on the long end, meaning it would need the whole sensor to do one standard of 8K, and the other standard of 8K would require a whole 45MP sensor. So if they cropped into that sensor at all, the crop wouldn't have enough pixels to output 8K.

In other words, the only way this camera can have a crop on 8K will be for the sensor to be way way bigger than 45MP. As an example, if this R5 had the same resolution as the a7RIV (>60MP), even that wouldn't have enough resolution to create much of a crop - it would still need ~80% of the sensor width just to get enough pixels to output 8K - the format is just that big.

Because of all that, we can be pretty confident that there will be no crop or near no crop, unless Canon makes this by far the highest resolution full frame sensor ever.
 
  • Like
Reactions: 1 users
Upvote 0
Jan 22, 2012
4,484
1,350
  • Like
Reactions: 1 user
Upvote 0
Jan 22, 2012
4,484
1,350
The point is it is very unlikely that 8K will have any crop or a large crop because it just isn't possible for smaller resolutions to create an 8K output. The reason being is that the width of 8K is either 7680 or 8192 (depending on the 8K standard we're talking about). A 40MP sensor would have ~7680 pixels on the long end, meaning it would need the whole sensor to do one standard of 8K, and the other standard of 8K would require a whole 45MP sensor. So if they cropped into that sensor at all, the crop wouldn't have enough pixels to output 8K.

In other words, the only way this camera can have a crop on 8K will be for the sensor to be way way bigger than 45MP. As an example, if this R5 had the same resolution as the a7RIV (>60MP), even that wouldn't have enough resolution to create much of a crop - it would still need ~80% of the sensor width just to get enough pixels to output 8K - the format is just that big.

Because of all that, we can be pretty confident that there will be no crop or near no crop, unless Canon makes this by far the highest resolution full frame sensor ever.
Thank you for this!!!!! I am delighted that it will be full-frame 4k. Let's see if it has RAW, DPAF, etc. I keep wondering how they will differentiate this from 1dx3. If it is on par, it will most likely be goodbye DSLR's for me.
 
  • Like
Reactions: 1 user
Upvote 0

davidhfe

CR Pro
Sep 9, 2015
346
518
According to your math you seem to know this. It is not the internet that made me feel Canon cameras will have a crop, it is how it has been. Is 1dx3 not the 1st camera without a crop? This is why I was wondering. :)

I appreciate your faith in my math (I can assure you my freshman calculus professor did not share it), but there are a dozens of places where engineering constraints might override some back of the envelope calculations.
 
  • Like
Reactions: 1 user
Upvote 0
Apr 25, 2011
2,519
1,898
At the possible expense of write speed as the card becomes fragmented (due to varying file size of each image depending on scene contents).
Unless you constantly keep writing onto an almost full card (which is an alternative to not writing onto a full card at all), wear-leveling logic of the card controller should practically eliminate the difference.
 
Upvote 0
Thank you for this!!!!! I am delighted that it will be full-frame 4k. Let's see if it has RAW, DPAF, etc. I keep wondering how they will differentiate this from 1dx3. If it is on par, it will most likely be goodbye DSLR's for me.
No worries!

I think a lot of people expected a crop because there have been crops in the past. Make no mistake though - 8K requires 4 times the data as 4K. That is an eye-popping amount of data, and that's why some people just don't believe this is going to happen without caveats. It is entirely possible that 8K is only possible without AF or only using an external recorder, but they should be able to do 8K at 30fps because the 1DXIII is doing 5.5K raw at 60fps, which (so I'm told) is moving about the same amount of data as 8K at 30fps. In other words, Canon has already shown that they can move that much data. With that said, I don't think the auto focus works at 60fps in 5.5k raw on the 1DXIII, so there is fair reason to think that AF may not be possible at 8K. Hard to tell what the other modes will look like - for context, if it has 4K120, that should be moving the same amount of data as 8K, and the 1DXIII can't do that so it may not be in this either.

At the end of the day though, not having AF at 8K or requiring an external recorder would be a pain, but the truth of the matter is this will likely still be the only full frame camera capable of doing 8K at all for some time.

But then again, it's all hypothetical until the camera is released, and the fact that they're claiming 8K in any form is ground breaking no matter how you slice it.
 
  • Like
Reactions: 1 user
Upvote 0
I do not understand the point.
You need a certain amount of megapixels (MP) for 8K (33-44 MP depending on pixels on width and aspect ratio). That means, if you have 1.6X crop on 8K, then:
1.6X crop 8K = ~33-44 MP
The diagonal of a full frame 35 mm sensor (36*24) is ~43 mm (please apply Pythagoras theorem to calculate that).
So, 1.6 crop on diagonal will yield ~26.8 mm.
Now, 26.8 mm diagonal at 3:2 aspect ratio will give dimensions of ~ 22.2*14.8 = 328.56 mm sq
If a 328.56 mm sq sensor has 8K resolution (assuming 33 MP for 8K) and a full frame sensor is the size of 864 mm sq,
then, no crop sensor (approximately, ) = (33*864)/328.56 = 86.77 MP
 
  • Like
  • Wow
Reactions: 1 users
Upvote 0

digigal

Traveling the world one step at a time.
CR Pro
Aug 26, 2014
283
586
The world cannot hear your over the singing angels they claim to hear while shooting 20 fps FF cameras. ;)

Battery power matters, I don't deny. But we're at the point that people are frothing at the mouth for what mirrorless can do in spectacular disproportion to their concern over batteries. People are just kind of dealing with the battery by packing an extra one or two, or they are pleasantly surprised with how long these mirrorless cameras actually perform in the field (with their preferred use-case, workflow, etc.) vs. estimates and specs.

- A
I got 1800 shots on one battery at Bosque this year using my R. I don't do video. Used my R almost exclusively there to try to see how the different focus tracking settings would perform because I want to use it as my b/u camera for the new R5--my main "squeeze" has been the 7DMII
Catherine
 
  • Like
Reactions: 1 users
Upvote 0

Michael Clark

Now we see through a glass, darkly...
Apr 5, 2016
4,722
2,655
Unless you constantly keep writing onto an almost full card (which is an alternative to not writing onto a full card at all), wear-leveling logic of the card controller should practically eliminate the difference.

In the scenario being discussed, aren't we talking about deleting images to make more room on a card? Doesn't that infer that the card is at near capacity?
 
Upvote 0

Michael Clark

Now we see through a glass, darkly...
Apr 5, 2016
4,722
2,655
Well, technically---they COULD create 4K on a 45 MP sensor by doing a 2.0 crop, since it only needs 1/4 the number of pixels. But if they can record the entire sensor as a movie at full resolution, why would they have to do that? There'd be no engineering reason to do so, that I can see.

To double the frame rate without increasing the processing load.
 
  • Like
Reactions: 1 user
Upvote 0

ahsanford

Particular Member
Aug 16, 2012
8,620
1,651
I would think it would be any of those 14,24,35 but not the 135. While the EF version is my favorite lens, it just doesn't seem to be a sales favorite what with stellar zooms that cover that focal length with almost or as good as primes IQ.
I say wide...14 and 24


The 135L doesn't sell well because:
  • It's 24 years old and what was once famously sharp is no longer
  • It does not have IS
  • It's only a stop faster than a 70-200 2.8
  • Canon's 70-200 2.8s are pretty damn legendary
Only the last bullet point may be true on RF. If they made one -- I'm no champion of wanting this, I'm just saying -- it would either have IS, be faster than f/2 or both.

Just consider an exotic tele prime that Canon might offer. Mitakon pulled off a limited run 135 f/1.4 for Sony, Nikon still makes the 105 f/1.4 for F mount, and Canon's 200 f/2L IS (and f/1.8L before it) is pretty damn sweet even if it does cost a mint:

Screen Shot 2020-02-14 at 1.41.47 PM.png

I think there's *a* prime lens in a 100-200 range there in RF's future. Why not the 135?

- A
 
  • Like
Reactions: 1 users
Upvote 0
Jan 22, 2012
4,484
1,350
No worries!

I think a lot of people expected a crop because there have been crops in the past. Make no mistake though - 8K requires 4 times the data as 4K. That is an eye-popping amount of data, and that's why some people just don't believe this is going to happen without caveats. It is entirely possible that 8K is only possible without AF or only using an external recorder, but they should be able to do 8K at 30fps because the 1DXIII is doing 5.5K raw at 60fps, which (so I'm told) is moving about the same amount of data as 8K at 30fps. In other words, Canon has already shown that they can move that much data. With that said, I don't think the auto focus works at 60fps in 5.5k raw on the 1DXIII, so there is fair reason to think that AF may not be possible at 8K. Hard to tell what the other modes will look like - for context, if it has 4K120, that should be moving the same amount of data as 8K, and the 1DXIII can't do that so it may not be in this either.

At the end of the day though, not having AF at 8K or requiring an external recorder would be a pain, but the truth of the matter is this will likely still be the only full frame camera capable of doing 8K at all for some time.

But then again, it's all hypothetical until the camera is released, and the fact that they're claiming 8K in any form is ground breaking no matter how you slice it.
I do not have any use for 8k just yet. As long as it has good 4k - good bit rate, DPAF, RAW (or log) and full-frame, I am a happy customer.
 
Upvote 0
Jan 22, 2012
4,484
1,350
The 135L doesn't sell well because:
  • It's 24 years old and what was once famously sharp is no longer
  • It does not have IS
  • It's only a stop faster than a 70-200 2.8
  • Canon's 70-200 2.8s are pretty damn legendary
Only the last bullet point may be true on RF. If they made one -- I'm no champion of wanting this, I'm just saying -- it would either have IS, be faster than f/2 or both.

Just consider an exotic tele prime that Canon might offer. Mitakon pulled off a limited run 135 f/1.4 for Sony, Nikon still makes the 105 f/1.4 for F mount, and Canon's 200 f/2L IS (and f/1.8L before it) is pretty damn sweet even if it does cost a mint:


I think there's *a* prime lens in a 100-200 range there in RF's future. Why not the 135?

- A
Love this photo. Really do! The blacks and the hint of lilac in the background. Favor requested: Pls pet the dog for me!
 
  • Love
Reactions: 1 user
Upvote 0
I keep scratching my head with something about the specs...

Before go on let me just say that for me it´s the same that this "spec" is there or not, I don´t care, I will buy the camera anyway. But let´s talk about the 8k in EOS R5. Something that Canon wrote in the development announcement got my attention. They said that: "...still images from video footage as well as process 8K video into higher-quality 4K video". So what I understand is that Canon EOS r5 process 8k image into a high quality 4k? So....This is an 8k image oversampled to 4k, like the Sony cameras that have 6k video oversampled to 4k? Is that it?? So if it is that, there will be NOT 8k25fps (pal region), but instead an 8k image oversampled to 4k video. So it´s not quite 8k...:)

Probably I am wrong but that´s my understanding about statement!

But what bothers me THE MOST is: "still images from video footage!" DAMN I totally HOPE they dont use that crappy MJPEG codec again!!!!

Any thoughts on this?
 
Upvote 0
I would think it would be any of those 14,24,35 but not the 135. While the EF version is my favorite lens, it just doesn't seem to be a sales favorite what with stellar zooms that cover that focal length with almost or as good as primes IQ.
I say wide...14 and 24
14 and 24 make a lot of sense, but part of me wonders if there was any truth to the thought of creating an f/2 trinity of zooms and the 28-70 will see a wider and a longer sibling. 14-28 f/2? Count me in!
 
Last edited:
  • Like
Reactions: 1 user
Upvote 0