Update: The Canon EOS R3 will be officially announced on June 29th

[Skyscraperfan]

Isn't there a kind of noise that gets overproportionally worse when pixels get smaller? Otherwise it is strange that sensors optimized for low noise often have a low resolution. I am thinking about that sensor by Canon for example that can do images with ISO 3,276,800 or so and only has a very low resolution. Of course if noise is random at each pixel, you can average it out, but there also is noise with other frequencies that effects a hundred or so pixels in the same way, Averaging it out will not work there and it could get worse on small pixel, because on small pixels tha signal has to be amplified much more than on a large pixel.

 

[EOS 4 Life]

Isn't there a kind of noise that gets overproportionally worse when pixels get smaller? Otherwise it is strange that sensors optimized for low noise often have a low resolution. I am thinking about that sensor by Canon for example that can do images with ISO 3,276,800 or so and only has a very low resolution. Of course if noise is random at each pixel, you can average it out, but there also is noise with other frequencies that effects a hundred or so pixels in the same way, Averaging it out will not work there and it could get worse on small pixel, because on small pixels tha signal has to be amplified much more than on a large pixel.

Cameras focused on high ISO have lower resolution.
I guess maybe they also have less noise at low ISO but they capture a lot less information.

 

[lethiferous]

Yes, at the current rate of market change that should only take a decade or so. Did you know that the best-selling ILC in Japan for the last two months was a Canon DSLR?

Pretty cure cost of entry and low price ef glass plays a huge role in that. A rebel with access to old cheap EF lenses is a lot cheaper.

 

[Loswr]

Pretty cure cost of entry and low price ef glass plays a huge role in that. A rebel with access to old cheap EF lenses is a lot cheaper.

The cost of the entry level DSLRs and MILCs (e.g., SL3/Kiss X10 and M50 MkII/Kiss M2) are essentially the same (within a few 10’s of dollars and different ones are slightly cheaper in different markets, so no real difference there).

Most APS-C body sales are 1- and 2-lens kits, and given the 1.4:1 lens:body market ratio, most consumers never buy another lens, so ‘old, cheap EF lenses aren’t a factor, and regardless EF-M lenses are similarly-priced with EF-S counterparts.

The reality is that a large swath of the market still prefers DSLRs over MILCs. Do you honestly think it’s a good idea to ‘slowly kill off’ camera lines that comprise 45% of the ILC market?

 

[Del Paso]

My dream has just come true!
The R 3 is for me the ideal camera for macro, with its eye-control AF, class-leading IBIS, and the RF 100mm, since 99% of my macro shots are handheld. Bank-account, get ready for a savage attack!
Edit: I forgot the integrated grip, perfect for vertical pictures.

 

[GMCPhotographics]

Fingers crossed for a reasonable selling price. After forking out for the 1dx mkiii last year the chances of persuading the boss that this is a must have could be difficult especially if its over £5k.

Invest in camera glass...not the camera bodies. The bodies worth drop like a hot brick over 3-5 years. Most lenses only take a 20% hit over their life span of 10-20 years.

 

[blackcoffee17]

I am guessing you meant "only 30MP".

No, i meant 4K video at 30fps.

 

[Skyscraperfan]

Many of your posts are nonsensical, but this one makes you look more foolish than most.

Downsampling means removing information from an image, a RAW image contains the original data from each pixel on the sensor.

Perhaps you're confusing downsampling with compression, they are not the same.

When you downsample a RAW image, it is no longer RAW. Period.

I don't think they will evev give use the real RAW images. For example there is one quite extreme transformation they apply before creating the RAW: The Bayer Filter only lets through either red, green or blue light for each pixel. However the RAW file already seems to contain an interpolated version of that real raw data that comes from the camera. Each colour is interpolated to neighbouring pixels to create the final image. So combining two green, one red and one blue pixel into a single pixel of an sRAW image would even be more accurate than the big RAW image that contains a lot of interpolated colour information. Even some noise reduction that can't be disabled is used to create the RAW file. So it is quite far away from real raw data.

 

[EOS 4 Life]

No, i meant 4K video at 30fps.

OK, thanks. My mistake. Now when I see FPS I think of still photos.

 

[koenkooi]

I don't think they will evev give use the real RAW images. For example there is one quite extreme transformation they apply before creating the RAW: The Bayer Filter only lets through either red, green or blue light for each pixel. However the RAW file already seems to contain an interpolated version of that real raw data that comes from the camera.[..]

Ehm, no. RAW images from Canon cameras haven't been debayered or interpolated yet, the software on you computer needs to do that. In the past you could use 'mRAW' and 'sRAW', which weren't actually RAW. They were debayered, downscaled TIFFs with enough metadata to be able to change the whitebalance in post.

As for the alledged noise reduction on the R5 sensor, we don't know where and why it's being done. It could very well be done in hardware by the sensor.

 

[Sporgon]

Cameras focused on high ISO have lower resolution.
I guess maybe they also have less noise at low ISO but they capture a lot less information.

I was disappointed when I compared the 12mp Sony A7s against my 5DS at relatively high ISO, ie practical for my situation, about 6,400.
The Sony is meant to be the one of the best quality at high ISO, yet when I downsampled the 5DS to the same 12mp it wasn’t that different.

 

[EOS 4 Life]

Ehm, no. RAW images from Canon cameras haven't been debayered or interpolated yet, the software on you computer needs to do that. In the past you could use 'mRAW' and 'sRAW', which weren't actually RAW. They were debayered, downscaled TIFFs with enough metadata to be able to change the whitebalance in post.

As for the alledged noise reduction on the R5 sensor, we don't know where and why it's being done. It could very well be done in hardware by the sensor.

Not completely.
RAW files are not a direct readout from the sensor.
They have still gone through an image processor.
The amount of processing varies from camera to camera which is one of the reasons Photoshop has to do custom RAW processing for every camera.
Adobe came out with DNG (Digital Negative) RAW format to try to avoid this but that became a clusterbomb.
Now each vendor has a custom RAW format that is different for every camera.

 

[Loswr]

I don't think they will evev give use the real RAW images. For example there is one quite extreme transformation they apply before creating the RAW: The Bayer Filter only lets through either red, green or blue light for each pixel. However the RAW file already seems to contain an interpolated version of that real raw data that comes from the camera. Each colour is interpolated to neighbouring pixels to create the final image. So combining two green, one red and one blue pixel into a single pixel of an sRAW image would even be more accurate than the big RAW image that contains a lot of interpolated colour information. Even some noise reduction that can't be disabled is used to create the RAW file. So it is quite far away from real raw data.

Sorry, no. You need to do a little more research on what comprises a RAW image.

 

[Loswr]

Not completely.
RAW files are not a direct readout from the sensor.
They have still gone through an image processor.
The amount of processing varies from camera to camera which is one of the reasons Photoshop has to do custom RAW processing for every camera.
Adobe came out with DNG (Digital Negative) RAW format to try to avoid this but that became a clusterbomb.
Now each vendor has a custom RAW format that is different for every camera.

Not quite. The ‘custom RAW format’ is a set of corrections applied by the RAW conversion software that supports the camera, and the manufacturer supplies those instructions. For example, the RF 24-240 is actually wider than 24mm and has mechanical vignetting at the corners, but when cropped to a 24mm FoV the vignetting is outside that frame.

When you view the RAW image in software that supports that camera, that cropping is applied by the viewing software. When you use an agnostic viewer like RawTherapee, you see the wider-than-24mm FoV of the 24-240 and the vignetted corners.

 

[usern4cr]

I don't think they will evev give use the real RAW images. For example there is one quite extreme transformation they apply before creating the RAW: The Bayer Filter only lets through either red, green or blue light for each pixel. However the RAW file already seems to contain an interpolated version of that real raw data that comes from the camera. Each colour is interpolated to neighbouring pixels to create the final image. So combining two green, one red and one blue pixel into a single pixel of an sRAW image would even be more accurate than the big RAW image that contains a lot of interpolated colour information. Even some noise reduction that can't be disabled is used to create the RAW file. So it is quite far away from real raw data.

Are you saying the raw file (as stored in memory, NOT as viewed) contains 3 colors per pixel after Bayer interpolation? I very strongly doubt that, but would welcome you to provide evidence of it. You might be assuming this because when you "view" a raw file (by using some software program), that program will do the Bayer interpolation of the (non-interpolated) file so that you can see it on a monitor which is expecting 3 colors per pixel format.

Now, some manufacturers are reported to apply filtering on the stored raw file itself (which is tragic IMHO) so that they have better results when their camera & sensors are tested by companies like DXO. (Remember the "star eater" firmware versions from Sony?) I don't know if this is true for Canon or not, and would be interested to know if it was.

 

[UpstateNYPhotog]

Have you played with any of the AI up-scalers lately, like Gigapixel? Pretty amazing. Who REALLY needs more than 30MP now? Even the 12MP A7s becomes useable as a stills camera again (not just for video). People who absolutely must have huge files will probably migrate to medium format if they haven't already or will rely on pixel shift (for motionless subjects) if they need to create a large file. I think 30MP is ideal for most uses. I haven't had a client ask for a print in decades. Almost all viewing is done on a screen now and for magazine work, 30MP is more than enough. If not, use AI software to double the perceived resolution. I just hope the R3 is less than $6000 USD. It apparently has the specs of the 1Dx and those have all been introduced at $6500+/- so I wouldn't be surprised is this one lands in that zone, too. But maybe Canon will not be greedy and will realize they'll sell more if they keep the price as low as possible. Remember the 1Dc that started out at $10,000? I'm hoping for a pleasant price surprise (and 4K 120fps) on the 29th.

Agreed on 30 MP. I've never had a complaint about 20 MP 1Dx II files either. 45MP would just slow me down. If the R5 was 30 MP I'd have two by now.

 

[UpstateNYPhotog]

Invest in camera glass...not the camera bodies. The bodies worth drop like a hot brick over 3-5 years. Most lenses only take a 20% hit over their life span of 10-20 years.

Yes. My original model EF 300 2.8 still takes amazing images. I think it's remarkable that Canon built a system in the late 80's such that you can still mount those lenses on an R5 from 2020 and do amazing work. I can't think of any other piece of expensive high technology that is so backwards/forwards compatible. Not that I still don't miss my FD 400 2.8.

 

[Skyscraperfan]

Are you saying the raw file (as stored in memory, NOT as viewed) contains 3 colors per pixel after Bayer interpolation? I very strongly doubt that, but would welcome you to provide evidence of it. You might be assuming this because when you "view" a raw file (by using some software program), that program will do the Bayer interpolation of the (non-interpolated) file so that you can see it on a monitor which is expecting 3 colors per pixel format.

Now, some manufacturers are reported to apply filtering on the stored raw file itself (which is tragic IMHO) so that they have better results when their camera & sensors are tested by companies like DXO. (Remember the "star eater" firmware versions from Sony?) I don't know if this is true for Canon or not, and would be interested to know if it was.

If the RAW file does not contain any interpolated colours, that would mean that each pixel only contains either red, blue and green colours. Doesn't that mean that the whole information contained in one red, one blue and one of the green pixels could be saved in a single RGB pixel? I think that depends on how the interploation works. Is each colour interpolated separately or do the blue pixels for example help interploating the red pixels?

 

[EOS 4 Life]

If the RAW file does not contain any interpolated colours, that would mean that each pixel only contains either red, blue and green colours. Doesn't that mean that the whole information contained in one red, one blue and one of the green pixels could be saved in a single RGB pixel? I think that depends on how the interploation works. Is each colour interpolated separately or do the blue pixels for example help interploating the red pixels?

There are no RGB pixels in a RAW file.
Each pixel is red, green, or blue.
(BRAW files are RGBW because it was originally designed for the URSA 12K sensor. which is not a Bayer sensor.)

 

[Danglin52]

First, and this is absolutely no insult, if you knew which was which, a subconscious assumption may have biased your analysis. That's why real science is based on blind or even double-blind testing: make people rate the photos NOT knowing which camera took which. And maybe even have the person making the test not know which is which, so they don't fall into some unspoken subconscious ordering.

Second, Canon has always had sensors with the wiring on the front of the sensor (front-side sensor). The double MP sensor has twice the wires interfering with light, so it is NOT true that a hi-MP front-side sensor will pick up as many TOTAL photons as a lo-MP sensor. In contrast the R3 and one assumes all future Canons are back-side. The entire front of the sensor is sensitive to photons, whether there's 1 big pixel or a billion. In this case, while every single pixel in the hi-MP sensor will have more noise, the scene as a whole won't, so down-sampling (especially, downsampling by an integer factor like exactly 4 pixels to 1) should give an identical image with identical noise characteristics as well as identical resolution.

I actually shared the files with a couple of other folks and just labeled them A & B version. Certainly not scientific measurement, but everyone picked the R6 shots off of a iMac 5k monitor. I am not a sensor geek, but aware of the difference between the sensors and potential advantages.