Canon EOS R5 Mark II going to 60mp? [CR1]

[Sporgon]

Where did you get the MTF50 of 3200 LW/PH for the G1XIII? Optyczne has measured it to be 1700 LW/PH at best in the centre at 45mm? https://www.optyczne.pl/389.4-Test_aparatu-Canon_PowerShot_G1_X_Mark_III_Optyka.html

https://www.photoreview.com.au/reviews/advanced-compact-cameras/fixed-lens/canon-powershot-g1-x-mark-iii/

 

[neuroanatomist]

From my experience of nearly seven years with 5DS cameras I would say that if you want more than 30 or so mp in FF you'd be better off looking at a larger format rather than cramming more pixels into FF.

Unless you're cropping, I agree.

One notable difference in that regard is the new OLPF (anti-aliasing filter) that Canon is now using delivers more resolution than the older one. Though there's likely not much difference between the new OLPF and the 'self-canceling' AA filter on the 5DsR, the new filter means that the current 24 MP sensors used in the R3, R8 and R6II outresolve the 30 MP sensors used in the 5DIV and EOS R.

 

[Sporgon]

Where did you get the MTF50 of 3200 LW/PH for the G1XIII? Optyczne has measured it to be 1700 LW/PH at best in the centre at 45mm? https://www.optyczne.pl/389.4-Test_aparatu-Canon_PowerShot_G1_X_Mark_III_Optyka.html

There is a bizarre difference in the results of those two sites. I don't believe that the Australian site is quoting MTF20. Could the difference be in the RAW vs (sharpened) Jpeg output I wonder ?
The original G1X was notorious for having soft RAW compared with the camera's own Jpeg output.

 

[Sporgon]

the new filter means that the current 24 MP sensors used in the R3, R8 and R6II outresolve the 30 MP sensors used in the 5DIV and EOS R.

I can believe it.

 

[AlanF]

There is a bizarre difference in the results of those two sites. I don't believe that the Australian site is quoting MTF20. Could the difference be in the RAW vs (sharpened) Jpeg output I wonder ?
The original G1X was notorious for having soft RAW compared with the camera's own Jpeg output.

You could well be right. They measure the MTF by IMATEST using slanting edge. 3200 LW/PH does seem rather high for a zoom lens and 4000 pixel high sensor. I like the Polish site as they show complete plots and seem to do the measurements carefully.

 

[neuroanatomist]

I can believe it.

I haven't tested it personally, but Canon made that statement and I presume they'd know.

 

[AlanF]

I haven't tested it personally, but Canon made that statement and I presume they'd know.

Optyczne confirm the 5DIV is outresolved by them

Test Canon EOS 5D Mark IV - Rozdzielczość - Test aparatu

Test aparatu Canon EOS 5D Mark IV - Rozdzielczość ...

www.optyczne.pl

Test Canon EOS R8 - Rozdzielczość - Test aparatu

Test aparatu Canon EOS R8 - Rozdzielczość ...

www.optyczne.pl

They show also 45 Mpx R5 outresolves the 50 Mpx 5DS and even the 5DSR that lacks the AA-filter.

 

[Sporgon]

Optyczne confirm the 5DIV is outresolved by them

Test Canon EOS 5D Mark IV - Rozdzielczość - Test aparatu

Test aparatu Canon EOS 5D Mark IV - Rozdzielczość ...

www.optyczne.pl

Test Canon EOS R8 - Rozdzielczość - Test aparatu

Test aparatu Canon EOS R8 - Rozdzielczość ...

www.optyczne.pl

They show also 45 Mpx R5 outresolves the 50 Mpx 5DS and even the 5DSR that lacks the AA-filter.

Having seen for myself the differences in resolved detail (or lack of it) between 20 to 50 mp FF sensors I can well believe that an improvement in design could result in 24 out resolving 30.

 

[neuroanatomist]

…the 5DSR that lacks the AA-filter.

It doesn’t, actually. It still has a pair of lithium niobate filters, they are just oriented such that the AA effect is canceled out.

 

[AlanF]

It doesn’t, actually. It still has a pair of lithium niobate filters, they are just oriented such that the AA effect is canceled out.

Yes - they didn't remove the filter but self cancelled in order to maintain optical thickness.

 

[JayLauz]

According to your logic, the fuzzy underwater image with lots of backscatter is a higher quality image, because it has more MP. Or maybe you will acknowledge that factors other than MP are more important for image quality. But maybe not...some people can't admit when they're wrong.

You totally warped my definition of "higher quality." I admit when I'm wrong; otherwise, I do not learn. Still, you can't convince me that a 24mp FF image is as clear as a 45mp image. As I mentioned, RAW photo resolution goes hand-in-hand with screens. Of course, if you're not using a 4K monitor, you may be unable to tell a difference. Regardless, thanks for your input.

 

[AlanF]

Having seen for myself the differences in resolved detail (or lack of it) between 20 to 50 mp FF sensors I can well believe that an improvement in design could result in 24 out resolving 30.

My 5DSR with the EF 100-400mm II significantly out-resolved my 5DIV with EF 400mm f/4 DO II + 1.4xTC and got better images in the field.

 

[Sporgon]

My 5DSR with the EF 100-400mm II significantly out-resolved my 5DIV with EF 400mm f/4 DO II + 1.4xTC and got better images in the field.

I don’t doubt it, but the 5DIV has quite a strong AA filter, certainly much more influential than that of the 5DS.

 

[AlanF]

I don’t doubt it, but the 5DIV has quite a strong AA filter, certainly much more influential than that of the 5DS.

That's why the R8/R6II outresolve the 5DIV as they have weaker ones, with very little filtering in the vertical dimension.

 

[calfoto]

Except that those files are not really RAW.

I’d used the reduced size Raw option occasionally and they always seemed to offer the same options as a full size Raw albeit at a lower resolution - so what was I working with if they weren’t really Raw?

Then again I’ve never used the full sized Raw on my R5 or R3, it’s always been the cRaw

 

[neuroanatomist]

You totally warped my definition of "higher quality."

Did I? You stated:

I think the high megapixel count is just a marketing tactic for people who want the absolute best quality photos.

You equate high MP count with ‘absolute best quality photos’. Pretty clear…but wrong.

Still, you can't convince me that a 24mp FF image is as clear as a 45mp image. As I mentioned, RAW photo resolution goes hand-in-hand with screens. Of course, if you're not using a 4K monitor, you may be unable to tell a difference. Regardless, thanks for your input.

4K is 8.3 MP. If you downsample a 24 MP and a 45 MP image to 8 MP, any ‘increased clarity’ you see in the downsampled 45 MP image is your imagination.

Incidentally, I use a 5K:2K ultrawide monitor, same vertical pixel count as 4K, but 33% more horizontal pixels than 4K. I like the extra screen real estate.

 

[David - Sydney]

This rumor is really old. The rumor sounds like Canon R5 II getting the old Sony sensor from the a7r IV... not good. If Canon is now taking Sony's hand me down's then Canon is no longer the camera company for professionals.

Hard to say that Nikon doesn't sell cameras for professionals (or Sony for that matter) just because they use Sony sensors.

 

[neuroanatomist]

I’d used the reduced size Raw option occasionally and they always seemed to offer the same options as a full size Raw albeit at a lower resolution - so what was I working with if they weren’t really Raw?

Then again I’ve never used the full sized Raw on my R5 or R3, it’s always been the cRaw

As @koenkooi stated, the older mRAW and sRAW options (for the .CR2 format) are already demosaiced. RAW converters that handle mRAW and sRAW must first interpolate the demosaiced pixel values. It’s not the end of the world, but technically they’re not RAW. Gory details from Doug Kerr on those formats are here.

The new cRAW option (for the .CR3 format) is RAW (albeit with some of the original data lost).

 

[David - Sydney]

Canon RAW is never binned (except Magic Lantern which is not official).
Last time it was written that an R5s would be a separate model alongside the R5 II which would make way more sense.

For video, the A1 does not (cannot?) record 8k raw and only 8k via codecs. One of the reasons is that it doesn't use CFe type B cards. It hasn't seemed to be a blocking issue for Sony sales.
The R5's 4kHQ video is pixel binned from 8k raw.

Let's not forget cRaw which Canon has no issues to call it a raw variant.

Whether 60mp is sufficient enough in a R5s to differentiate from a 45mp is debatable.

 

[David - Sydney]

8K displays are 33Mp, but for 3:2 aspect ratio sensor to cover it, it has to have at least 39Mp. With 4K DCI, it's 44Mp.

Do you mean "8k DCI"?