Canon EOS R3 sensor resolution to be 45mp?

[FrenchFry]

But could they be downsizing because they've signed an agreement not to post any images larger than 24MP? I admit 24MP is looking more and more likely.

This is the possibility that I was pointing out. It can't be ruled out, based on the information that we have. It's a more appealing narrative to me as someone interested in small wildlife than the alternative, a camera maxed out at 24MP. I know that won't make it true, but I am still hoping for the best.
I think the evidence is compelling, but I'm not quite ready to give up the 45MP or 30MP dream yet.

Also, what is the likelihood that the R3 is exactly 24MP? Certainly possible (R5 is 45MP), but less common (5DS R is 50.6, 90D is 32.5, 5D m4 is 30.4, 1 DX iii and R6 are 20.1, R is 30.3, RP is 26.2, etc.)

 

[neuroanatomist]

How many Getty photographers shoot RAW? How many sports photographers shoot RAW? Even if there are some with R3 who do, they’d be limited to some unreleased version of DPP for them, that Canon would have needed to provide.

Far more likely they’re all shooting JPG, and quite probable that the cameras Canon puts in their hands can only output JPG files. Thus, Canon could simply set the firmware in those cameras to output only 24 MP files.

 

[privatebydesign]

I commented earlier that Canon normally work to DCI standards for video resolution, 6,000 pixels doesn't downscale to 4k logically or easily, 6,144 does. But that puts it at 25.1mp.

 

[Bdbtoys]

How many Getty photographers shoot RAW? How many sports photographers shoot RAW? Even if there are some with R3 who do, they’d be limited to some unreleased version of DPP for them, that Canon would have needed to provide.

Far more likely they’re all shooting JPG, and quite probable that the cameras Canon puts in their hands can only output JPG files. Thus, Canon could simply set the firmware in those cameras to output only 24 MP files.

Although with other images at native, I would think the 24mp is the most likely. However @neuroanatomist made a good point. An R5 is 44.8 @ L and 22.5 @ M jpeg settings. There is nothing stating they were not limited to using M.

 

[neuroanatomist]

Although with other images at native, I would think the 24mp is the most likely. However @neuroanatomist made a good point. An R5 is 44.8 @ L and 22.5 @ M jpeg settings. There is nothing stating they were not limited to using M.

It doesn’t even need to be M jpg size. Would be simple for Canon to code a fixed jpg output of 6000x4000 pixels regardless of the sensor MP count.

 

[unfocused]

I'm about 90% sure that the R3 is 24 megapixels at this point looking through everything.

What are you looking at? I see that Getty offers rights to use images in three sizes – small, medium and large. Are you referring to the sizes quoted there? If so, that means nothing.

Every time I shoot any sports event, after I'm done processing the files I send the final files through the Photoshop script that resizes them and converts them to jpgs (From PSDs) I have one setting for photos that will end up on the web and another larger size in case the photos are needed later for print. It takes hardly any time at all.

These photographer are not posting or processing anything themselves. Their files are transferred to editors who select the best frames and to techs who process those frames. No doubt, Getty has custom scripts that resize the processed file so they can offer them in three sizes for anyone to purchase the usage rights.

I apologize if you are looking at something else, but if you are looking at the sizes Getty is offering the images in, that means zero, zip, nada.

 

[Bdbtoys]

It doesn’t even need to be M jpg size. Would be simple for Canon to code a fixed jpg output of 6000x4000 pixels regardless of the sensor MP count.

True, but that was just an example. However, from a programmer perspective... it would be far easier to use/remove a setting that already exists.

 

[HotPixels]

Al Bello, one of Getty's staff photographers, uploaded a bunch of photos from swimming today where a large chunk of the photos are *exactly* 24 megapixels, with some being slight crops around 23 megapixels. There's other ones taken with the exact resolution of the 1DX mark III, so these aren't upsized 1DX images, and there were some R5 images that were used as a remote which are kept around full 45 resolution. I doubt it's a coincidence that there's dozens of exactly 24 megapixel images in the set. Even cropping you wouldn't get that perfect.

Getty has a contract with Canon to only use Canon gear, so these also aren't A9II images or anything like that.

Pictures and Photos - Getty Images

View and license pictures & news photos from Getty Images.

www.gettyimages.com

I don't think the assumption that all photographers shooting for Getty are automatically using Canon holds true.

I just checked their site for Olympic photos...there are several by one photographer, François-Xavier Marit, of AF-P. He is a Nikon shooter. He tags a lot of his IG photos with #nikon and #nikonfr.

Nikon featured him a while back in an article on their global NPS website:
https://nps.nikonimaging.com/interview/marit/

And indeed, many of his images are exactly 24MP.

So I think that while those directly employed by Getty may only use Canon, perhaps they also contract out to others for some of their work.

But regardless, the assumption that all photos of the Olympics offered by Getty Images are taken with Canon is incorrect.

 

[HotPixels]

I don't think the assumption that all photographers shooting for Getty are automatically using Canon holds true.

I just checked their site for Olympic photos...there are several by one photographer, François-Xavier Marit, of AF-P. He is a Nikon shooter. He tags a lot of his IG photos with #nikon and #nikonfr.

Nikon featured him a while back in an article on their global NPS website:
https://nps.nikonimaging.com/interview/marit/

And indeed, many of his images are exactly 24MP.

So I think that while those directly employed by Getty may only use Canon, perhaps they also contract out to others for some of their work.

But regardless, the assumption that all photos of the Olympics offered by Getty Images are taken with Canon is incorrect.

I'm going to reply issue a correction to myself. I now get what the point that the original post was making. Al Bello is indeed a staff photog at Getty and does indeed use Canon gear. So the collection of photos that he has uploaded is of unique interest.

So while not all Getty photos are taken with Canon gear, all photos by that one photographer, Al Bello are taken with Canon gear. And that means something. Sorry for the confusion.

 

[FrenchFry]

I'm going to reply issue a correction to myself. I now get what the point that the original post was making. Al Bello is indeed a staff photog at Getty and does indeed use Canon gear. So the collection of photos that he has uploaded is of unique interest.

So while not all Getty photos are taken with Canon gear, all photos by that one photographer, Al Bello are taken with Canon gear. And that means something. Sorry for the confusion.

Or, also possible, it doesn't mean anything.

As you pointed out, lots of the images on Getty have the same max size, regardless of the body used.

 

[Franklyok]

No reason why it should, and even if it did it would be a nominal difference.

Less mpx data to dig out would mean more DR data to dig out of sensor. Processor is limited. I still hope for more DR.

 

[privatebydesign]

Less mpx data to dig out would mean more DR data to dig out of sensor. Processor is limited. I still hope for more DR.

Same generation sensors seem to be pixel density agnostic when normalized.

 

[Franklyok]

Same generation sensors seem to be pixel density agnostic when normalized.

It is new generation , Canon’s first BSI sensor.

And still I believe DR depends on processor(s), and not so much about sensor ( less mpx the better of cause ).

Each processor generation has added some 0.3 stops of DR to the same generation sensor. Check dxomark 18 mpx apsc for example.

Hopefully dual digic 10 this time.

 

[dpockett]

I don't think the assumption that all photographers shooting for Getty are automatically using Canon holds true.

I just checked their site for Olympic photos...there are several by one photographer, François-Xavier Marit, of AF-P. He is a Nikon shooter. He tags a lot of his IG photos with #nikon and #nikonfr.

Nikon featured him a while back in an article on their global NPS website:
https://nps.nikonimaging.com/interview/marit/

And indeed, many of his images are exactly 24MP.

So I think that while those directly employed by Getty may only use Canon, perhaps they also contract out to others for some of their work.

But regardless, the assumption that all photos of the Olympics offered by Getty Images are taken with Canon is incorrect.

I am a stringer for Getty and to my knowledge, they don't have a contract with Canon. I know a few staffers who shoot on Nikon, one I can think of on Panasonic (ambassador). Haven't met any on Sony, majority shoot Canon, but there is no contract. Canon and Nikon send teams to major events to support photographers with service and gear to borrow, that's where the R3's would be coming from.

What are you looking at? I see that Getty offers rights to use images in three sizes – small, medium and large. Are you referring to the sizes quoted there? If so, that means nothing.

Every time I shoot any sports event, after I'm done processing the files I send the final files through the Photoshop script that resizes them and converts them to jpgs (From PSDs) I have one setting for photos that will end up on the web and another larger size in case the photos are needed later for print. It takes hardly any time at all.

These photographer are not posting or processing anything themselves. Their files are transferred to editors who select the best frames and to techs who process those frames. No doubt, Getty has custom scripts that resize the processed file so they can offer them in three sizes for anyone to purchase the usage rights.

I apologize if you are looking at something else, but if you are looking at the sizes Getty is offering the images in, that means zero, zip, nada.

If you look in the information below the photos you can see the resolution of the full-size image. This is the res it is uploaded as, not the small/medium/large.

Traditionally, photographers for Getty/AFP/Reuters edit their own photos live, but when there are major international (and some domestic) events, editors are on shift to handle the editing. So yeah, there is a massive team of editors for Getty currently editing the Olympics. Photographers sending remotely to editors all over the planet to edit. Usually the files won't be resized, but they are often cropped and output at whatever size cropped to.

 

[Franklyok]

30 mp is enough of a jump from my 1DX mark II that I would definitely pre-order the camera day one. 24 mp just doesn't feel particularly as future proof to me, so then it would turn back into a decision for me between the R3 or just getting a second R5 with a battery grip while I wait for the R1.

24 mp also makes me feel like they could pull off way more than 30 fps.. The 1DX mark III was doing 5.5K raw at friggin' 60 frames per second. Why not do 50-60 FPS raw photos?

The only thing I could think why one would need more mpx is to be able to crop in , as if 24 mpx isn’t enough and 800 mm f11 too expensive.

 

[David_D]

I think it might be time for some deductive speculation... I have been thinking about what we know and what we would like to know. One thing Canon have told us is that the R3 can focus in low light, down to EV -7. I believe the figure for the R5 is -6.5, i.e. a half stop improvement. Larger photo-sites, i.e. lower resolution would improve the low light figure, suggesting 30mp or even 24mp could be right. However my understanding is that the BSI sensor should also improve low light ability. Would a 24mp or 30mp BSI sensor have only half a stop improvement over the R5? Could the BSI sensor alone give that improvement, i.e. the resolution is unchanged? Could anyone who actually understands these things comment?

 

[FrenchFry]

The only thing I could think why one would need more mpx is to be able to crop in , as if 24 mpx isn’t enough and 800 mm f11 too expensive.

There are lots of reasons why the 800mm F11 won't work for many shooting situations that have nothing to do with price. If shooting small animals in a forested area, for instance, F11 may not gather enough light. There can also be issues with not enough subject separation (busy backgrounds) and a really large minimum focusing distance. Furthermore the lens is not weather sealed like an L lens and thus more vulnerable to getting fogged up or damaged in humid and rainy weather typical of the tropics.
The lens does a lot of things really well for the price, but it's not a universal solution for all reach limited situations.

 

[privatebydesign]

It is new generation , Canon’s first BSI sensor.

And still I believe DR depends on processor(s), and not so much about sensor ( less mpx the better of cause ).

Each processor generation has added some 0.3 stops of DR to the same generation sensor. Check dxomark 18 mpx apsc for example.

Hopefully dual digic 10 this time.

I know it’s a new generation, my point is this new generation will have the same dynamic range if it is 24 or 30mp, or 45, or 60. Pixel density does not seem to be a big factor in dynamic range capability when the sensors are the same generation and the output is normalized.

Look at examples from the 1DX III and R5.

 

[Franklyok]

I know it’s a new generation, my point is this new generation will have the same dynamic range if it is 24 or 30mp, or 45, or 60. Pixel density does not seem to be a big factor in dynamic range capability when the sensors are the same generation and the output is normalized.

Look at examples from the 1DX III and R5.

View attachment 199222

You probably also mean same generation processor. It is possible indeed. We have no idea if Canon has thoughtfully configured ( cripled / down shifted ) to make no difference.

however in the past the same old sensor ( 18 mpx for example ) has a litle performance increase, if loaded with newer generation processor. Processor / motherboard are the botle necks in my opinion.

R3 processor has AI side tasks, so the main accomplishment may not be bigger DR diging at all.

 

[privatebydesign]

You probably also mean same generation processor. It is possible indeed. We have no idea if Canon has thoughtfully configured ( cripled / down shifted ) to make no difference.

however in the past the same old sensor ( 18 mpx for example ) has a litle performance increase, if loaded with newer generation processor. Processor / motherboard are the botle necks in my opinion.

R3 processor has AI side tasks, so the main accomplishment may not be bigger DR diging at all.

No, I said what I meant. I also showed data to support that position.