The Canon EOS R5 Mark II coming in Q2, 2023? [CR2]

[Deleted member 378085]

That is sheer nonsense: photocells, whatever their size, are designed to being as close together on the sensor so there are minimal gaps between them otherwise light is lost. Spacing cells further apart in the same area would lower light gathering power, make poorer iso response and lower DR.

I’m not going to go back and forth with a silly argument, believe what you want, frankly I don’t care if you understand it or not.

 

[Deleted member 378085]

You call the physics simple, but clearly you don’t understand the physics involved in determining sensor noise. The irrelevant convo here is yours.

That makes no sense

 

[Loswr]

That makes no sense

If you’re referring to your own posts, congratulations you’ve finally make a correct statement.

Read the following slowly and carefully: assuming a similar generation of sensor technology, pixel size has virtually no impact on image noise, the primary determinant of image noise is sensor size.

You seem to be confusing pixel noise with image noise. The latter matters to photographers, the former to measurebators. Which are you?

 

[Dragon]

That is sheer nonsense: photocells, whatever their size, are designed to being as close together on the sensor so there are minimal gaps between them otherwise light is lost. Spacing cells further apart in the same area would lower light gathering power, make poorer iso response and lower DR.

I think he had a point, but just didn't say it clearly. Lower MP cameras have larger pixels with wider center-to-center spacing would be a clearer statement. In the case of DPAF, each pixel is split in half with part right looking and part left looking. Larger pixels (with larger center spacing between the DPAF half pixels) would seem to have an advantage both from a light gathering perspective (less noise at any given ISO) and from an angular resolution perspective (needed for PDAF). I think this is why you are seeing faster AF acquisition on your R5 vs your R7 even though the R7 arguably has a generational advantage with its AF system derived from the R3.

 

[koenkooi]

I think he had a point, but just didn't say it clearly. Lower MP cameras have larger pixels with wider center-to-center spacing would be a clearer statement. In the case of DPAF, each pixel is split in half with part right looking and part left looking. Larger pixels (with larger center spacing between the DPAF half pixels) would seem to have an advantage both from a light gathering perspective (less noise at any given ISO) and from an angular resolution perspective (needed for PDAF). I think this is why you are seeing faster AF acquisition on your R5 vs your R7 even though the R7 arguably has a generational advantage with its AF system derived from the R3.

Since the R7 AF is also slower in good light, my personal suspicion is that the sensor readout speed matters a lot for AF. The R7 sensor is a lot slower than the R5 and R3.

 

[EOS 4 Life]

ok.. not the right forum to ask this but how can the RF800 a f11 lens get so good reviews? even in broad day light it would be impossible to shot BIF and other wildlife where you need a shutter speed of 1:1000 or even 1:2000.....

The Canon EOS R line of mirrorless cameras performs well at high ISO.
An 800 mm lens at higher apertures costs many thousands of dollars more.
Sure it has its limits, but nothing comes close for the price.

 

[Del Paso]

I have a Lumix S1R and the high-res pixel shift works really well. I use a tripod but haven't had a problem with vibration.

According to Olympus, the OM 1's pixel shift works even hand-held.
If it's true, this would be a great feature !

 

[AlanF]

Since the R7 AF is also slower in good light, my personal suspicion is that the sensor readout speed matters a lot for AF. The R7 sensor is a lot slower than the R5 and R3.

The readout time for the R5 is 15.5ms, and that for the R7, which has fewer pixels, 31.3ms. The readout time for the R3 is quicker still at 5.5ms. Canon has recycled an older, slower sensor, which is still fine for resolution and IQ but appears to have slower response for AF.

 

[AlanF]

I think he had a point, but just didn't say it clearly. Lower MP cameras have larger pixels with wider center-to-center spacing would be a clearer statement.

You would make it much clearer still by simply writing that the amount of light gathered by a pixel (sensel) depends on its area. To relate center-to-center spacing to the area of a pixel in that way assumes that the pixels abut each other. That assumption is contradicted by the second part " before they run into an neighbouring photocell on the sensor." which implies there are gaps between pixels in lower Mpx sensors.

Because of simple physics, th R3 at 24mp has larger photocells, meaning they can be spaced apart further in the same area and can be pushed to higher iso sensitive before they run into an neighbouring photocell on the sensor.

Anyway, as pointed out by @neuroanatomist, and by many others and me here, the overall noise in an image depends on the size of the sensor and not on the size of of the pixels.

 

[whothafunk]

I never understood the appeal of SD cards. The argument is that you can get SD cards everywhere, but I think people, who pay that much money on a camera, will have enough memory cards anyway.

I have R3 and R5, have CFe cards but also SD. My Macbook has an SD slot, which in a hurry (as a sports photographer and journalist) helps a lot to not have to play around with CFe card reader and cable etc. It is just SO convenient.

 

[Skyscraperfan]

That is one of the examples where Apple has a bad influence on the whole tech world. Of course other notebooks usually also have SD card slots, but they usually copy the bad standards of Apple.

In the past notebooks at least had SD card slots with enough space for the whole card. At modern notebooks the SD cards usually stick out, which is very annoying, at that makes it much more likely to break the card.

 

[arthur]

Which RAW converter are you using? If it's something else than DPP4, you need to address that complaint to the people that wrote your RAW converter.

Also, isn't c-log a video only thing?

Yes, what I mean is raw from the video side, you can see the raw dng color quality in this video that was shot with canon 5d mark iii which was filmed 7 years ago

 

[koenkooi]

[...] In the past notebooks at least had SD card slots with enough space for the whole card. At modern notebooks the SD cards usually stick out, which is very annoying, at that makes it much more likely to break the card.

Having dealt with (micro)SD sockets in product we made at $previousjob: the springs are a major cause of failure, especially with humans involved. "I Tried to insert it real gently (with a hammer)"

For 'permanent' media, a flush, spring loaded socket is recommended. But for things like readers where the card only spends a few minutes in the reader, a stick out, non spring loaded socket has my preference. And steel backed cards like Sony Tough and Hoodman Steel

 

[koenkooi]

Yes, what I mean is raw from the video side, you can see the raw dng color quality in this video that was shot with canon 5d mark iii which was filmed 7 years ago

But that's a Magic Lantern thing, not a Canon thing, right?

 

[Skyscraperfan]

Having dealt with (micro)SD sockets in product we made at $previousjob: the springs are a major cause of failure, especially with humans involved. "I Tried to insert it real gently (with a hammer)"

For 'permanent' media, a flush, spring loaded socket is recommended. But for things like readers where the card only spends a few minutes in the reader, a stick out, non spring loaded socket has my preference. And steel backed cards like Sony Tough and Hoodman Steel

I mainly use my SD card slot for a backup of my most important files. I back them up four times per day on a Samsung "Pro Endurance" Micro SD card, which is made to last very long. That card has saved me very often. The adapter sticks out though and that is a problem.

 

[thorhe]

***

How likely is it so see this camera coming with a built-in ***? I'm okay with buying an extra battery if need be.

On a more personal level, I'd very much like to see the overall size going up. Coming from a 5D and having bigger hands, the R5 camera is too small.

The Canon GP-E2 should work with Canon EOS R5.

 

[Loswr]

I think he had a point, but just didn't say it clearly.

He had a point, but not a correct one.

Lower MP cameras have larger pixels with wider center-to-center spacing would be a clearer statement. In the case of DPAF, each pixel is split in half with part right looking and part left looking. Larger pixels (with larger center spacing between the DPAF half pixels) would seem to have an advantage both from a light gathering perspective (less noise at any given ISO) and from an angular resolution perspective (needed for PDAF). I think this is why you are seeing faster AF acquisition on your R5 vs your R7 even though the R7 arguably has a generational advantage with its AF system derived from the R3.

First off, he said nothing about AF speed, but regardless that red herring has already been pickled with readout speed brine by @koenkooi and @AlanF.

Yes, a larger pixel gathers more light than a smaller pixel. Yes, more light means lower noise. But a full frame sensor filled with smaller pixels gathers the same total amount of light as a full frame sensor filled with larger pixels. Same total light means same noise.

Many years ago, that was not true. Each pixel only had a small light-sensitive area, and thus more pixels in a given area meant less light and more noise. The non-photosensitive area of pixels was in the 60-80% range, meaning a lot of light lost. But with the gapless microlenses and light guides used in modern sensors, that’s no longer true. Essentially the full surface of a pixel is photosensitive (because the microlenses collect light over the whole pixel area).
_________________________

Since this thread is also discussing pixel shift, it’s worth mentioning that the same gapless microlenses that obviate the effect of pixel size on image noise also reduce the benefit of pixel shift for spatial resolution.

Over 20 years ago, I had Zeiss cameras with pixel shift. The sensors lacked gapless microlenses, so when pixel shift was used to capture a 2x2 full-pixel array, that increased color resolution by sampling the same subject area separately in R/G/B channels. Using a 2x2 or 3x3 sub-pixel array increased spatial resolution, capturing more area in each pixel space by moving the ~1/3 of the pixel’s photosensitive area around that space to sample most of (2x2) or the entire (3x3) pixel area.

Modern sensors are already sample the entire pixel area with a single capture, thanks to the gapless microlenses. Thus, pixel shift in today’s cameras increases color resolution, but has much less effect on spatial resolution. So when you use it with a Fuji GFX100, for example, the multiple captures with the 100 MP sensor result in a 400 MP image, but the resulting image has lower spatial resolution than you’d get with a 400 MP sensor.

 

[arthur]

But that's a Magic Lantern thing, not a Canon thing, right?

Yes, it was actually an unofficial hack of Canon's operating system, however the results are amazing, Canon can now use the same format in a better, and more professional way, but unfortunately they chose the ineffective raw format ( c-log raw )

 

[entoman]

On a more personal level, I'd very much like to see the overall size going up. Coming from a 5D and having bigger hands, the R5 camera is too small.

I came from 5D series to R5, and despite having large hands I quickly got used to the smaller body - the grip is fine and the controls are all well-spaced.

It does feel less sturdy to me though, compared to the indestructible feel of 5D and 7D models. I've had no issues (apart from occasional freezes) with the R5, but if I was going to be in a really tough environment where my gear was going to get a real bashing, I'd have more faith in my 5DMkiv.

 

[Dragon]

Since the R7 AF is also slower in good light, my personal suspicion is that the sensor readout speed matters a lot for AF. The R7 sensor is a lot slower than the R5 and R3.

Absolutely. Basic control theory tells us that the higher the sample frequency, the faster the loop response can be. The catch is that an R7 with a stacked sensor would be at a price point that would be a hard sell for the masses. At the current price, it will sell very well.