Canon Interview: EOS R1 is the true flagship

[Quack]

It is peak of their technology so for that company that is a flagship device it is simple as that.

Yes, so simple. But most folks want to make their own definition, so they can complain about a camera they would never buy. And all those that complain that it does not have the MPs of the A1 or Z9 conveniently forget that the R5 mark II gives them exactly what they want. Or are they afraid to mention the R5 mark II, or buy it, since it is not the "flagship" and they think their camera buddies will laugh at them? Grow up. It's amazing to me that people on forums don't mind giving others the impression that they are 11 year old crybabies.

 

[MikeGalos]

Ok

haha.

Yep, my first internet experience was on old Digital DEC stations using gopher protocol that predated HTTP.

If we are dating ourselves

Bah. Another youngster. HTTP wasn't invented until 1989 and wasn't standardized until 1997. Gopher dates back to 1991. I had my first Internet account in the mid 1980s.

To be fair, though, a friend of mine makes me look like a newbie. He worked on the initial hardware for the ARPAnet - the predecessor of the Internet - back in 1969

And there's nothing wrong with dating yourself. At least you're dating someone who likes you.

 

[Richard CN]

So Canon builds a flagship camera for the contract sports photographer what is the problem?

Has any 1-series camera ever been a desirable camera for the average photographer? How does one define average in this context?

If one has a demanding project that requires the "highest resolution possible" a photographer might want to consider a Fuji GFX MF with 100 MP sensor which will provide more pixels than any FF camera.

and more importantly a higher quality pixels at that.

 

[Berowne]

and more importantly a higher quality pixels at that.

BTW I always wondered, whether a Dual-Pixel-AF means two Pixel and so a 20 Mpx Sensor with Dual-Pixel-AF has indeed 40 Mpx. Directly asking, does the 24 Mpx-Sensor of the R1 with Cross-Type-AF has indeed 96 Mpx? And can you measure the Pixelpitch by looking at the sensor with a microscope? Another question i was thinking about was, whether the R1-Pixel are too comlex to shrink them.

 

[Richard CN]

BTW I always wondered, whether a Dual-Pixel-AF means two Pixel and so a 20 Mpx Sensor with Dual-Pixel-AF has indeed 40 Mpx. Directly asking, does the 24 Mpx-Sensor of the R1 with Cross-Type-AF has indeed 96 Mpx? And can you measure the Pixelpitch by looking at the sensor with a microscope? Another question i was thinking about was, whether the R1-Pixel are too comlex to shrink them.

a 20MP sensor, would be 40 million photodiodes, yes.

cross type is still 40 million, they just change the orientation of some.

 

[mxwphoto]

They already have a product here (https://www.dpreview.com/news/8482927209/canon-announces-its-new-cr-s700r-robotic-camera-system) though I don’t know if it is fast enough to track motorsports.

I saw that too but the cr-s700r is simply a motorized mount with connectors for the camera. It lacks the automatic AI functions of subject recognition, tracking, and opportune moment picture taking. Though I am sure it won't be rocket science to integrate an automated control unit to such a rig in a future development.

 

[Omn]

Inverse correlation? If that statement were correct, then a 1 pixel or 100px or 1000px or 1 mpx sensor would have better image quality than a 24 Mpx or 45 Mpx sensor.

Reductio ad absurdum​

Look it up.

 

[Omn]

The problem is that when people ding the R5, they never mention the most important reason to buy an R5 is the sensor megapixels. Those who want high FPS typically will give up a high-quality image. That's fine if that's what you want, but if the other vendors are offering 60mp on their high-end cameras, why haven't they gotten the message and dumbed down their sensors?

"High quality image" is not defined by sensor megapixels. This was common knowledge in the times of top end 12Mp FF cameras.

I posted a few links from dpreview comparisons showing directly how R3 excells in shadow performance compared to R5, A1, Z9 and even 1DXIII does. And Nikon ZF compared to Z9. This is just one part.

No one has "dumbed down" their sensors.

 

[AlanF]

Reductio ad absurdum​

Look it up.

Reductio ad absurdum is a process I use regularly when teaching or analysing problems. You are using the term in a popular parlance of pushing to extremes. But, it has a real meaning in logical arguments for testing whether a proposition is true or false: "In philosophy, reductio ad absurdum often reveals contradictions in theories by assuming a position is true and showing its absurdity. " You can read more here: https://simple.wikipedia.org/wiki/Reductio_ad_absurdum


You made the statement:

6) Image quality and pixel resolution are not synonims, and the correlation is inverse

I showed that that your proposition that the "correlation between image quality and pixel resolution is inverse" is false by the simple reductio ad absurdum argument:

Inverse correlation? If that statement were correct, then a 1 pixel or 100px or 1000px or 1 mpx sensor would have better image quality than a 24 Mpx or 45 Mpx sensor.

Your proposition is wrong because it predicts that all pixel resolutions below the one you start with give better images, which is nicely illustrated by the above. If your proposition were correct, a 24 Mpx sensor would have poorer IQ than a 20 Mpx, which in turn would have poorer IQ than a 12 Mpx, which would in turn be poorer than a 6 Mpx sensor, which would be poorer than a 3 Mpx sensor etc etc.

 

[Omn]

I showed that that your proposition that the "correlation between image quality and pixel resolution is inverse" is false by the simple reductio ad absurdum argument:

There is nothing false there, no you didn't show anything false and logical fallacy didn't help you.

Yes, larger pixels have multiple.advantages over smaller. In dynamic range, affecting highlight and shadow performance and saturation properties across rhe range.

No one needs 1 pixel or 1000 pixel camera so some rational middle ground for photography between spatial sampling density and overall image quality must be made. Like 24Mp and 6 microns, giving large enough pixels for quality with sufficient resolution for many.

Not interested in debating belief systems originating in pixel race marketing.

 

[AlanF]

There is nothing false there, no you didn't show anything false and logical fallacy didn't help you.

Yes, larger pixels have multiple.advantages over smaller. In dynamic range, affecting highlight and shadow performance and saturation properties across rhe range.

No one needs 1 pixel or 1000 pixel camera so some rational middle ground for photography between spatial sampling density and overall image quality must be made. Like 24Mp and 6 microns, giving large enough pixels for quality with sufficient resolution for many.

Not interested in debating belief systems originating in pixel race marketing.

It's nothing to do with belief in pixel race marketing, it's the basic fundamentals. Larger pixels do not necessarily give larger dynamic range or shadow recovery when images are viewed at the same size, and in theory should be very similar all else being equal. Here, for example, are the actual measurements by Bill Claff for the R3 and R5.
https://www.photonstophotos.net/Charts/PDR.htm#Canon EOS R3,Canon EOS R5
https://www.photonstophotos.net/Charts/PDR_Shadow.htm

 

[neuroanatomist]

There is nothing false there, no you didn't show anything false and logical fallacy didn't help you.

Yes, larger pixels have multiple.advantages over smaller. In dynamic range, affecting highlight and shadow performance and saturation properties across rhe range.

Yes, at the level of the individual pixel. That’s great if you look at pixels. Personally, I look at pictures, not one pixel but the output of a sensor composed of them. At the level of sensor output, larger pixels offer little to no advantage in terms of IQ, as shown by the Bill Claff plots that @AlanF posted.

Not interested in debating belief systems originating in pixel race marketing.

Are you interested in debating facts? If so, you’re not doing it very effectively.

 

[Omn]

It's nothing to do with belief in pixel race marketing, it's the basic fundamentals.

Basic fundamentals in consumer tech propaganda.

Larger pixels do not necessarily give larger dynamic range or shadow recovery when images are viewed at the same size,

No, they necessarily give/allow larger DR, shadow recovery, highlight performance and overall saturation performance due to combination of factors related to well size for capture and processing crammed into each pixel.

Viewing size has nothing to do whatsoever with camera performance.

and in theory should be very similar

They are similar, but 6 micron size bucket is superior to 4 micron size bucket in the start so all goes from there so it depends on what you see or care about.

all else being equal. Here, for example, are the actual measurements

I have better measurements.

But you don't need any to confirm this by yourself. You have a set of eyes and few posted links to dpreview raw high iso comparisons and clear insight into how 1DXIII and Nikon ZF demolish every new high pixel count camera. So a set of eyes and accurate logic will do. No need to rely onto someone's claims.

You can also just take cameras like R3 and R5 and compare by yourself real world examples in high contrast shots where more advantages can be seen and experienced directly.

 

[Omn]

Are you interested in debating facts? If so, you’re not doing it very effectively.

Not interested in debating this at all.
Who understands, understands. Who doesn't but can and wants to, will.

Who prefers drinking the marketing koolade and parroting the media, ignoring own eyes and common sense is best left at it, any effort in disenchanting is a waste of time.

 

[OldDudePhotog]

I have better measurements.

Please provide these measurements, I would be interested in seeing them for myself as I am interested in the data to support your statements. (this is not any sort of dig at you, I am genuinely interested in seeing your measurements and the environment you collected them in)

 

[AlanF]

Basic fundamentals in imaging tech propaganda.



No, they necessarily give/allow larger DR, shadow recovery, highlight performance and overall saturation performance due to combination of factors related to well size for capture and processing crammed into each pixel.

Viewing size has nothing to do whatsoever with camera performance.



They are similar, and if you want to nitpick and aim for the best performance possible, 6 micron size bucket is superior to 4 micron size bucket in the start so all goes from there.



I have better measurements.

But you don't need any to confirm this by yourself. You have a set of eyes and few posted links to depreview raw high iso comparisons and clear insight into how 1DXIII and Nikon ZF demolish every new high pixel count camera. So a set of eyes and accurate logic will do.

Or you can just take cameras like R3 and R5 and compare by yourself real world examples in high contrast shots where other advantages can be seen.

As @neuroanatomist has pointed out you are confusing looking at individual pixels rather than the actual image. A simple illustration, for example, is that a square of 4 pixels each of 3x3 microns^2 has the same area as 1 pixel of 6x6 microns^2 and the 4 small combined have the same well depth as the single larger so images from both viewed at the same size are looking at he same well depth despite one sensor having 4x the number of pixels. Viewing size is crucial to comparing the performance of cameras. For example, APS-C cameras are said to have more noise and lower DR than FF precisely because the images from each get examined at the same viewing size. If you viewed the FF image at 1.6x1.6 the size of the APS-C image, then both images would appear to have the same DR and noise per unit area.

 

[josephandrews222]

 

[josephandrews222]

 

[josephandrews222]

One more...all from

Photons to Photos

 

[AlanF]

View attachment 218963

View attachment 218964

Thanks for the heads up - they appeared today. The points for the R5ii are all downward pointing triangles, which means that noise reduction is being applied to raw throughout the whole iso range.