Canon EOS R5 Specifications

[Michael Clark]

You don't have to reply to my 'for humors sake' posts you know...or do you?

(please don't reply)

OK.

 

[Michael Clark]

But Canon already has thrown in the towel on the EF mount - they are only working on R glass and have said so. So if I'm a photographer looking for a new camera, why would I buy into obsolecense?

I expect a few more R lenses this year including smaller F4 variants. So say there are a total of 15 R lenses by end of year. Yet if I buy a DSLR, I have no access to any of them. No chance of ever buying the compact 70-200 2.8.

Yet if I have an extensive EF lens collection, I can still use them seamlessly with a very compact adapter on an R mount camera.

This is obviously what Canon's game plan is. That is why to me it makes no sense to make a new 5D when they are about to announce the R5. I predict that the R5 is the new 5D.

It makes no sense to you because you don't use your camera to do the things for which an OVF camera still beats an EVF camera hands down. There are plenty of other 5-series users that do.

 

[Michael Clark]

This argument, to me, is the same as throwing up your hands and giving up.

Language is about precision. If you can't define the term, the term has no meaning.



Oh good.

There are varying levels of precision. You can't use a standard deviation of .28 units between two sample sets to argue for a significant difference if your smallest unit of measurement is 1.0 units.

Symbols used in mathematics or physics are much more precise than most words we use.

Expecting words such as "art" or "happiness" or "acceptable" to have the same degree of precision as "pi" or "48°C" is not only unrealistic, it is impossible.

Words and other symbols have no meaning apart from what their users assign to them. Some symbols have broader ranges of meaning assigned to them by those who use them than others do.

 

[Michael Clark]

An exception would be if Canon updated the 5D with an RF mount. All the EF glass could still be used and the new 5D would keep the OVF and DSLR handling and ergonomics that many seem to desire. Then Canon let’s the buyers decide, but still moves to exclusive RF Mount. That lets 5D users begin investing in RF glass moving forward.

Where's the mirror for a FF OVF going to fit in the RF's 20mm registration distance?

 

[Michael Clark]

Honestly, I do think a front to back sliding sensor is possible - but all that extra movement would make me a bit nervous on life of the sensor moving system, or introducing more opportunity for vibration in the body. But if I were to imagine a camera that could use both EF and RF without an adapter, that's how it'd look.

I don't think either is likely, but wouldn't it make more sense from an engineering point of view to move the mount flange forward and back, rather than the sensor? There would be a barrel (similar to a zoom lens that extends - but maybe "squarer") with a just large enough to fit around the light box. The 4mm difference between the minimum 20mm registration for RF and maximum 44mm registration should not be a deal breaker.

 

[snappy604]

yeah ... might have been, recall the naming scheme was different. Luckily I still have it.. so I can check when I get home. My only problem is the dates don't match.. it was bought for a wedding in 2001 and that doc indicates 2002.

D60 and 2003. Was looking at an incorrect date stamp from my 3 megapixel minolta I was also using at that time

 

[Kit.]

as a 14 bit sensor can't deliver more than 14 stops.

A 1-bit sensor can deliver more than 1 stop, with the help of dithering. Why a 14-bit sensor cannot do the same?

 

[Kit.]

Just for the heck of it I went over and peeked at the Sony Alpha Rumors site for talk of the R5. They did have this article posted and the rumored specs. I was expecting a lot of mocking " yea...right!...suuuuuuure" replies but Surprisingly everything I read in the comments was positive on Canon basically. Most wished the specs were real and that the R5 would get Sony to stop slacking off and give them something competitive with some features they have been yearning for.

Are there any clearly Canon trolls, though?

 

[Quarkcharmed]

A 1-bit sensor can deliver more than 1 stop, with the help of dithering. Why a 14-bit sensor cannot do the same?

By capturing multiple images? In that sense any HDR technique will improve the DR. But how stacking/blending/HDR is related to DXO measurements? You think they use HDR mode?

 

[Bahrd]

A 1-bit sensor can deliver more than 1 stop, with the help of dithering. Why a 14-bit sensor cannot do the same?

Alternatively: you can have 15 ounces of water in a single 15oz bucket or in two 14oz ones, can't you?

 

[Memdroid]

What is a 1Dx? Has Nikon came out with a cropped body model with "1" designation? "Dx" is a Nikon designation for cropped sensors and APS-C only lenses.

Or do you mean the Canon EOS 1D X?

Cute

 

[Kit.]

By capturing multiple images?

By covering the image with multiple pixels.

How many photons are you trying to register per pixel anyway?

 

[Quarkcharmed]

By covering the image with multiple pixels.

How many photons are you trying to register per pixel anyway?

I'm not trying it. You are So it's on your side now to explain how you're going to have more than 14 stops in a 14-bit file.

 

[dtaylor]

You simply can't eliminate the DR difference by making adjustments

Of course you can when it comes to human perception. NR affects the N in SNR. There's a limit to what you can do but you absolutely can affect SNR in post. And if two sensors are close enough you can reduce the gap such that most viewers will never perceive it.

There always will be a range in shadows (or highlights) that's recoverable on D850 and not recoverable on 5DIV.

You've spent way too long looking at graphs. And those graphs have left you with the impression that there are hard limits here and there. When someone like DxO tests DR and publishes a graph they are choosing an arbitrary threshold at which they think noise passes from 'acceptable' to 'unacceptable.' They have to because otherwise how would you perform the test?

Human perception is not nearly so fine grained. You may look at two images at 3:1 in LR and say 'omg the noise on this one is terrible.' The artist may say 'the noise is a little worse.' And 1,000 viewers looking at 36" prints in a gallery may never notice that one has more shadow noise than the other. (And if you really want to turn reality upside down, try a test with and without labels, and with labels reversed.)

Now I'll agree that 2, 3, more stops difference...as measured by someone like DxO...is going to be noticeable in some situations across most print sizes. But 1? Less than 1? You would be very hard pressed to see it even with no post processing in a big print. It's just simply not a large difference.

And again we're talking about certain scenes only. Because most scenes fit within the DR of a D60. A surprising number even fit in the range of Velvia 50.

As above, there's no practical way at the moment to properly recover DPAF data. So I'm not sure if anyone can actually measure what 5DIV sensor is capable of after 'processing DPAF data properly'.

DPAF processing sometimes failing for you is not the same as never working at all or providing useful insight into the sensor design.

Also I don't trust DxO at all, as a 14 bit sensor can't deliver more than 14 stops.

Of course it can when you scale the image to a smaller print/view size. Go read about their screen and print measurements.

As to the small pixels vs big pixels I've already responded, A7RIV has a poorer performance than A7RIII,

According to Photons to Photos they're the same at ISO 100 and neck-in-neck through the rest of the ISO range.

 

[dtaylor]

I'm not trying it. You are So it's on your side now to explain how you're going to have more than 14 stops in a 14-bit file.

By scaling the image down which increases SNR. At the extreme of this technique you can get a visible light sensor to detect ionizing radiation it was never designed to detect.

 

[AlanF]

Just for the heck of it I went over and peeked at the Sony Alpha Rumors site for talk of the R5. They did have this article posted and the rumored specs. I was expecting a lot of mocking " yea...right!...suuuuuuure" replies but Surprisingly everything I read in the comments was positive on Canon basically. Most wished the specs were real and that the R5 would get Sony to stop slacking off and give them something competitive with some features they have been yearning for.

The Sony and Nikon forums are just like the CR, complaining about their shortcomings and how Sony and Nikon are ignoring them, the new releases are no better than the previous etc. The truth is they are all manufacturers trying to optimise their profits and pay their CEOs and shareholders.

 

[padam]

sorry, the point that I was trying to make was that R and RP are, in my view, a POC, MVP type of product by Canon's intent and by design.

It can be viewed like that, but they still very much serve as integral part of their lineup and they are not going to be discontinued, the naming scheme is just what it is, marketing differentiation.
The RP is by far their cheapest ever FF camera, and the R sits between the 6D and 5D line of cameras but closer to the latter with the features that it offers, some added and some taken away, it will probably be more popular than the 5D IV.

A lower entry to their FF mirrorless system, that's what they were designed for, they were never branded as 'Pro' models (and of course, originally they were the only cameras available for they system).
And they will sell in bigger numbers than these new models, which are more refined, but also quite a bit more expensive.
Over time these upcoming models will become less expensive as well, maybe a few more firmware tweaks, and after the 3-4 year product cycle they will get replaced again.
What you can have for 1300-1500$ is the EOS R, and that's fine imho. 30MP is already more than enough, and 45MP is probably unnecessary for most people.

They are not going to do an A7III-like, "middle-megapixel-count" camera that has dual slots IBIS FF 4k high frame rates etc. at the same level as the EOS R.
The R6 which will likely cost a bit more than the R5, and I guess they will have a similar body, same batteries with optional battery grip.
It will be competing against the A7SIII which will also move to the ~4000$ price level where the S1H currently sits.

 

[AlanF]

I'd really be surprised if an RF 400mm f/2.8 L IS is introduced before an RF 300mm f/2.8 L IS and RF 500mm f/4 L IS.

The EF 400mm f/2.8 L IS II and EF 600mm f/4 L IS II both got total redesigns to "III" versions in 2018.

The EF 300mm f/2.8 L IS II and EF 500mm f/4 L IS II are 2011 designs. They'll be the first RF great whites.

I really wonder about the 300mm f/2.8. The price of used ones has nosedived. There was a huge take up by birders and nature photographers when the II was released because it was the way to get to a hand held 600mm with a 2xTC and and f/4 420mm with good IS, and the only alternatives the antique 400/5.6 without IS, the older 100-400mm with two stops only or the dismal 400mm DO. The introduction of the 400mm DO II effectively killed it at the top end, as did the 100-400mm II at the affordable.

 

[Quarkcharmed]

Of course you can when it comes to human perception. NR affects the N in SNR. There's a limit to what you can do but you absolutely can affect SNR in post. And if two sensors are close enough you can reduce the gap such that most viewers will never perceive it.

I recall this very argument from you in one of the previous conversations. Digital post-processing manipulations are intrinsically flawed because they change the detail, NR simply decreases the resolution. But it's getting even worse when you apply NR to one image and don't apply it to another. What are you comparing exactly in this case and how it's related to the sensor performance comparison? You apply the NR to 7D's image and compare it to the unchanged A7RIII's image. You don't see a fallacy in this comparison?
And what if I apply the same amount of NR to A7III's image? Suddenly it's two stops ahead again!

You've spent way too long looking at graphs.

I'm only doing it when looking at the new cameras. In a practical sense, I can compare my current 5DIV against 70D, 550D, 650D, 100D, and also few raw files from Sony A7RIII and Fuji GFX100.

The 70D, which should be roughly the same as your 7D, is significantly worse than 5DIV, but A7RIII is a bit better and I'd be extremely happy if Canon catches up.
Fuji is just a different league.

And again we're talking about certain scenes only. Because most scenes fit within the DR of a D60.

Haven't worked with D60, but many scenes fit within the DR of 70D. But they're noisy in the shadows and the room for manipulation is quite small.

Of course it can when you scale the image to a smaller print/view size. Go read about their screen and print measurements.

I know. I still think there's some flaw in their method. 14 stops is the theoretical limit. Any measurement above means they do digital manipulations so they don't actually measure the sensor performance.

DPAF processing sometimes failing for you is not the same as never working at all or providing useful insight into the sensor design.

But in the filed I don't care about the sensor design tricks. I care if I should shoot with dual pixel enabled and adjust my exposure hoping that DPRSplit will help get 1 stop more in the highlights. Yes it helps but somewhat randomly and often fails, so I can't rely on it.

According to Photons to Photos they're the same at ISO 100 and neck-in-neck through the rest of the ISO range.

The difference is small but the IV lags behind in many ISO points at high ISO which is exactly where the pixel size difference should be more prominent.

 

[Quarkcharmed]

By scaling the image down which increases SNR. At the extreme of this technique you can get a visible light sensor to detect ionizing radiation it was never designed to detect.

That's interesting, do you have a link with more details on ionising radiation?

Regardless, when you scale the image down you no longer measure the sensor performance. Instead, you measure how good the image is for digital noise reduction [through downsampling]. All absolute values of measured DR become totally meaningless.

However they can be used for relative comparison, e.g. A7RIV is better than 5DIV if the images are downsampled and normalised. Very limited use, but it's good enough to create a popular site with the camera scoring system.