Canon EOS R7 Mark II to Have Stacked 40MP Sensor?

[EricN]

Thanks. Still I don't understand that for the R6 iii the difference is much smaller. It seems they primarily try to get extra money from the professionals.

I'm guessing they decided based on analysis of how much the target consumers have been willing to pay for similar products and purchase rates in the past.

 

[tareqphoto]

Because of R1 up to R5 latest models/versions i am putting my hope on both, R6 Mark III and R7 Mark II, i don't think i have to go very high specs, i came from 1DX and 1 series so i am sure many mirrorless nowadays can match quality or even surpass for sure, and that is all what i need, i don't care much about body performance, i am planning to return back to sports photography, i won't think too much about high flagship body because i don't do anything professionally and not very serious or not getting paid, so i won't spend for R1/A1/Z9 or even Z8/R3/A9III level to justify it, i saw enough results from R5 and R6 and R7 and even R8 that showing me that quality is there, struggling with focus or settings is another story.

 

[Michael Clark]

I'm still a bit concerned because the original R7's noise performance was quite bad above 3200, and with the greater pixel density it could be even worse if they don't do something else. Also, is this the highest density pixel sensor that Canon has(or will) released?

Well, the original 18 MP 7D was a lot noisier than the 20MP 7D Mark II. Or at least my 7D was a lot noisier than my 7D Mark II.

My 22.3MP 5D Mark III is also a little noisier than my 30.4MP 5D Mark IV.

Moderate increases in resolution can be offset by improvements in sensor technology over a 4-5 year period.

 

[Michael Clark]

... the noise is primarily due to statistical fluctuations in the number of photons at high iso.

It's actually the larger influence of Poisson distribution (i.e. statistical fluctuations) due to the reduced amount of light allowed into the camera, which is what requires those higher ISOs, that causes the increased noise. If you shoot at high ISO but let enough light into the camera to fully saturate the highlights there's a lot less noise than if you still don't get enough light to saturate anything in the frame, even with the higher analog amplification due to the higher ISO.

 

[Michael Clark]

I hope it is 40mp just to see another wave of upset 7D DSLR fans. I want to see a post in 2030 of someon explaining why they're still waiting to switch over to mirrorless.

What makes you think 40MP would upset any more 7D DSLR fans than 32MP would? Either way there will be complainers.

If they make it 40MP there will be those who moan, "They should have left it at 32MP and improved the high ISO/low light performance!"

If they leave it at 32 MP others will cry, "They didn't increase the resolution at all!"

 

[Michael Clark]

I'm in if two wishes come true. Put it in a R6 body I need bigger buttons and please add the voice note. As a news photographer it is so helpful to add an ID or "this was a two run double in the sixth". When you are sending pictures on deadline it is so helpful.

Canon has never put that capability in anything other than 1-Series bodies. I highly doubt they ever will.

Maybe if their competitors begin to give it to lower tier bodies it may eventually happen?

 

[Planterz]

What makes you think 40MP would upset any more 7D DSLR fans than 32MP would. Either way there will be complainers.

If they make it 40MP there will be those who moan, "They should have left it at 32MP and improved the high ISO/low light performance!"

If they leave it at 32 MP others will cry, "They didn't increase the resolution at all!"

I'd say that it's pointless to argue about theoretical specs before we even know what they are, or assume that Canon hasn't improved both noise performance and resolution. I guess that's the point of these forums though.

 

[Michael Clark]

Based on my experience with a 7D vs 5D3 and R7 vs R6-2, I think it's been this way since about 2012. My maximum auto ISO for the APS-C cameras is 6400 and for the FF cameras, 12800.

The 2009 7D should be compared to the 2008 5D Mark II.

The 2014 7D Mark II is the one that should be compared to either the 2012 5D Mark III or the 2016 5D Mark IV. The 7D Mark II was released almost at the end of 2014 while the 5D Mark III and 5D Mark IV were released early in 2012 and mid-year 2016, respectively, so the 7D Mark II release date was closer to the 5D Mark IV than to the 5D Mark III.

I never shot beyond ISO 3200 with the 7D. I'll go ISO 5000 or if desperately needed ISO 6400 with the 7D Mark II. But if the light is that dim I'm usually swapping my 70-200/2.8 from the 7D to the 5D Mark III to use with the 5D Mark IV as my "wide" body in a two-body setup. Either that or forgetting f/2.8 zooms altogether and going with fast primes, typically 35/2, 85/1.8, and 135/2

 

[Michael Clark]

I said a 70-200mm equivalent f2.8 APS-C, not a 70-200mm f2.8 equivalent APS-C.

There's very little difference in size, weight, and cost between a 70-200/2.8 with a FF image circle and a 70-200/2.8 with an APS-C image circle. The front group, which is where a significant portion of the money is spent, has to be the same size either way to allow 200mm f/2.8. The most expensive piece of glass in most 70-200mm f/2.8 is an aspherical high index (UD in Canon speak) element which is usually the second or third element in the front group.

To be f/2.8 "equivalent" for APS-C , it would need to be an f/1.8 lens, rather than f/2.8. That makes it even more expensive than a FF 70-200/2.8.

 

[AlanF]

It's actually the larger influence of Poisson distribution (i.e. statistical fluctuations) due to the reduced amount of light allowed into the camera, which is what requires those higher ISOs, that causes the increased noise. If you shoot at high ISO but let enough light into the camera to fully saturate the highlights there's a lot less noise than if you still don't get enough light to saturate anything in the frame, even with the higher analog amplification due to the higher ISO.

The iso doesn't affect the amount of light let into to the camera - it's the same amount at all isos, and there is the same noise (Poisson) distribution if you keep the aperture and speed constant and fiddle with the iso up and down. The iso just sets the ceiling for the amplification in the read out, it doesn't increase the amplification. If you fully saturate the highlights, you will bleach them.

 

[Michael Clark]

What cameras and lenses were used here? Since no camera you list has more than 24MP, for all we know the degraded left-hand image was an APS-C crop from from the center of a 24MP full-frame image - and is thus barely more than 9MP, not to mention the lower resolving power of a shorter lens - particularly if it's shot at a smaller aperture incurring more shot noise? Hardly fair. Doesn't "spark joy" to see such a post.

Not to mention the possibility of differences in processing pipeline. AI NR could get the left example a lot closer to the right example than what is shown.

 

[Michael Clark]

Thanks for the examples, Alan!
I was looking at RAWs SOOC with Canons original DPP.
And there the results were not as good as yours look here.

There's no such thing as "RAWs SOOC". The data is highly processed to give you the image you see on your screen. If you haven't modified any adjustments after opening the images, then whatever default settings are selected when DPP opens raw files can have an effect. You may even be looking at the JPEG preview image attached to the raw file if you have some of your default display settings in DPP set to "fast" rather than "high quality".

Not to mention that Digital Photo Professional has been on version 4 since around 2012. I'm surprised the "original" version of DPP even supports those two cameras, as version 3 certainly does not.

 

[Michael Clark]

Mistakes won't show more easily on higher MP sensors if you view the image at the same magnification. It's only if you insist on viewing at 100% or cropping more on the higher pixel sensor that eg diffraction or camera shake will be more clearly resolved.

What many pixel peepers fail to realize is that viewing at 100% from a 22MP sensor is a much lower enlargement ratio than viewing at 100% from a 50MP sensor.

On a standard FHD 24" monitor with 96ppi, at 100% you're looking at a piece of something like a 60x40 inch enlargement for the 22MP image. On the same monitor at 100% you're looking at a piece of something like a 90x60 inch enlargement for the 50 MP image. The more you enlarge, the more you magnify everything, including blur and noise.

 

[Michael Clark]

Total amount of light gathered is relevant to what? Exposure and signal to noise ratio both depend only on the intensity of light hitting each photosite and the exposure time. What does total amount of light gathered across the whole sensor affect? Answer: Nothing, unless you're trying to use your camera to sunbathe.

Correct answer: The ratio of signal to Poisson distribution noise. (a/k/a "shot" noise) Light has random distribution within a field from a point source. The randomness increases as the square root of the intensity. So the more photons you collect, the more uniform (thus, less random) the intensity of the field is.

 

[Planterz]

There's very little difference in size, weight, and cost between a 70-200/2.8 with a FF image circle and a 70-200/2.8 with an APS-C image circle. The front group, which is where a significant portion of the money is spent, has to be the same size either way to allow 200mm f/2.8. The most expensive piece of glass in most 70-200mm f/2.8 is an aspherical high index (UD in Canon speak) element which is usually the second or third element in the front group.

To be f/2.8 "equivalent" for APS-C , it would need to be an f/1.8 lens, rather than f/2.8. That makes it even more expensive than a FF 70-200/2.8.

Yeah. I know. I knew this before I was "corrected" by Chunk, and you just repeated what he said. Neither of you are reading what I said, and you're ignoring the context in which it was said.

I said a 70-200mm equivalent f2.8 APS-C, not a 70-200mm f2.8 equivalent APS-C.

We were discussing an APS-C lens to match the zoom range of the commonly used 70-200mm f2.8. That'd be ~45-135mm for Sony/Nikon/Fuji with a 1.5x crop or ~42-125mm on Canon's 1.6x crop. Fuji has one, the 45-150mm f2.8, but it's old and heavy, as has already been mentioned. Neither Sony, Nikon, nor Canon have made one.

A 42-125mm lens would give APS-C Canon users the 70-200mm equivalent range, and f2.8 would give them the same shutter speeds. It'd be smaller and lighter and presumably cheaper than a 70-200mm f2.8.

I said nothing about having an equivalent depth of field or any other sort of equivalence. Let me copy/paste it for you one more time.

I said a 70-200mm equivalent f2.8 APS-C, not a 70-200mm f2.8 equivalent APS-C.

I suppose I should have used the word "effective" instead, but I seriously doubt it that would prevented the inevitable flood of people arguing equivalence for several pages even though nobody postulated a theoretical and extremely unlikely 42-125mm f1.8 APS-C lens.

For anyone who wants to continue this discussion, we were/are talking about a lens with a ~42-135mm zoom range with an f2.8 aperture. This range on an APS-C camera would roughly effectively match the zoom range and field of view of the commonly used 70-200mm lenses used by professionals for indoor sports, weddings, events, and other things. The f2.8 aperture would let you shoot at the same shutter speeds, albeit with less shallow depth of field, which everybody understands and doesn't need to be argued further. It should be lighter and cheaper than a 70-200mm f2.8, which on a Canon APS-C camera would give an effective 112-320mm full frame field of view, but might be too long/narrow, especially if the situation calls for the 70-200mm field of view.

 

[Michael Clark]

We are, after all, in a discussion thread about an anticipated APS-C body - and many of us -perhaps most - see the R7 as their main camera, not merely a telextender for their full-frame setup.

The R7 Mark II, if it turns out to be the camera many here think it might be, is probably not the best choice from among Canon's current or near future APS-C offerings for a generalist camera.

The current R7 or R10 (or R20 if that is the name of what replaces the R10 in the presumably near future) would continue to be a better body for a generalist using it as their only camera that what many think the R7 Mark II will be. The specifications in the rumor at the top of this thread might be interpreted by some to point in that direction. Personally, I'll believe it if and when Canon announces such a camera.

It's no different than the comparisons between the 7D Mark II and the 80D/90D.

The 80D/90D were better all around cameras for non-professional generalists (i.e. those not shooting hundred to thousands of frames per day several days per week every week of the year) using it as their main body. The 80D/90D had better DR at low ISO. The 80D and 90D only had shutter life ratings of 100,000 and 120,000 actuations, respectively. The 80D/90D had polycarbonate bodies.

The 7D Mark II was a better tool for specific use cases like birding or sports. The 7D Mark II had a shutter life rating of 200,000 actuations. The 7D Mark II had a more sophisticated and better performing AF system than the 80D/90D, particularly when using AI Servo with targets like birds in flight or fast moving athletes. It had a magnesium alloy body and was "the most thoroughly weather-sealed camera I’ve ever run across" at the time Roger Cicala took a 7D Mark II apart.

Even further back, the original 7D was the more logical successor to the 50D, while the 60D was a downgrade from the 50D in several ways. (The 50D and 7D both had AFMA, the 60D did not. That's a big difference when using fast and long telephoto lenses as many 7D users did. The 50D and 7D had magnesium alloy bodies, the 60D was a polycarbonate body. Etc. )

For what it's worth, the R7 is more of an 80D/90D type of camera than it is a 7D Mark II type of camera. It remains to be seen which way the R7 Mark II will lean.

 

[Michael Clark]

For anyone who wants to continue this discussion, we were/are talking about a lens with a ~42-135mm zoom range with an f2.8 aperture.

Why didn't you just say that? It was less than clear that is what you meant with what seems to be intentionally vague:

I said a 70-200mm equivalent f2.8 APS-C, not a 70-200mm f2.8 equivalent APS-C.

 

[Michael Clark]

Thank you. I was, in fact, a strong chess player in my youth, which was useful training for playing with pigeons. Though I now prefer to photo them. Here is one looking out of a clock taken on my R7 in a beautiful town in Tuscany this summer.

View attachment 227193

Why does the shadow cast by the "little" hour hand look like a ground squirrel?

 

[neuroanatomist]

That'd be ~45-135mm for Sony/Nikon/Fuji with a 1.5x crop or ~42-125mm on Canon's 1.6x crop. Fuji has one, the 45-150mm f2.8, but it's old and heavy, as has already been mentioned. Neither Sony, Nikon, nor Canon have made one.

You might want to consider what you wrote above more carefully. Fuji’s lens (50-140/2.8 not 45-150) hasn’t been updated in over a decade. No other major manufacturer makes a similar lens for APS-C. So the companies that collectively sell over 85% of ILCs don’t see a market for such a lens and the one niche player that thought there was such a market has let their offering languish.

There’s a logical conclusion that can be drawn from the above facts, even if you’d prefer to ignore it.

Something else to consider is that for manufacturers that offer both APS-C and FF, having their APS-C buyers switch to FF is profitable. While people may not mind ‘abandoning’ an APS-C kit lens with a format switch, I suspect many would be reluctant to do the same with a lens costing $2000 (the old Fuji 50-140/2.8 sells for $1700, a new lens like that from Canon would likely cost more).

Even if Canon sold such a lens for less than the Fuji version, say $1500, your suggestion that someone could buy that lens and an R7II for less than the cost of a 70-200/2.8 ($2500) is ludicrous.

 

[Planterz]

Why didn't you just say that? It was less than clear that is what you meant with what seems to be intentionally vague:

View attachment 227282

I did. Here's the original context, which you've ignored twice now.

Sigma needs to add a small 50-135 or 140 f/2.8

Sony users have long lamented the lack of a ~45-135mm f2.8 APS-C lens to match the common 70-200m f2.8 telephoto zoom. I think once upon a time Sigma was rumored to be working on one, but clearly that never came to be. Fuji has a 50-140mm f2.8, but they're the only APS-C maker to do so. Now that there's potentially 4 mounts (X, E, Z, RF-S), maybe it'll finally make sense for Sigma or whomever to make one.

If indeed the R7 II is more "entry-pro"/enthusiast level, a 70-200mm equivalent f2.8 APS-C makes a lot of sense to me. A high school's photography/journalism department or proud parent could feasibly get the camera and lens to shoot basketball or volleyball games for about the cost of the RF 70-200mm Z (no body).

I'm not going to make excuses on your behalf if you're going to misread things out of context.