All new 24mp sensor coming to the next Canon ILC’s [CR1]

Yup. “All new” is now meaningless when stated by Canon. They do sometimes release generational leaps, but all releases will be termed “all new.”
Why, it's not meaningless. It now translates to 'the same but slightly tweaked'. An actually new sensor will be called 'revolutionary' or something like that, so we will know.
 
  • Like
Reactions: 1 user
Upvote 0
Jan 12, 2011
760
103
yes, you may as well forget about that.

it's not just the sensor it's the processors that have to handle all that. Canon's only done that on their C700. it's not even in their CINI lineup that much, and they have cooling fans.

So, I guess you're saying Canon doesn't have processors that can handle all that inside a DSLR body?
 
Upvote 0
So, I guess you're saying Canon doesn't have processors that can handle all that inside a DSLR body?

It's either that or sensor readout speed that's the bottle neck. The C700 is able to do a full readout fast enough because it's a low resolution sensor. The 5.9k sensor crops for 4k, but it can shoot 5.9k with no crop.
 
  • Like
Reactions: 1 users
Upvote 0
I also don't want to see aliasing artifacts in my photos. And I only very rarely see them with my 5DSR, which is why I would like more sensors without an AA-filter as they give sharper images.
Please explain the theory of how you can have AA-filters that kill the artifacts and have very little impact on actual fine detail.

I am with you here, I got rid of the AA filter when switching to the 5DSr and and never looked back. Now my backup camera is a Sony A7R and my next camera will likely be the next Canon that doesn't have an AA filter. I do mostly architecture and interieur for a living, and I still have to try to find a single picture that shows any artifacts. Actually when I picture fabrics, I have less (if any) moiré, when my 5D2 with AA was a moiré beast. High Res cameras do not need AA filters, period. I prefer to have 1 picture out of 1000 with a vague artifact in the background than all my pictures with a soft filter. IMO Canon is too conservative on the matter.
 
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,351
22,524
I am with you here, I got rid of the AA filter when switching to the 5DSr and and never looked back. Now my backup camera is a Sony A7R and my next camera will likely be the next Canon that doesn't have an AA filter. I do mostly architecture and interieur for a living, and I still have to try to find a single picture that shows any artifacts. Actually when I picture fabrics, I have less (if any) moiré, when my 5D2 with AA was a moiré beast. High Res cameras do not need AA filters, period. I prefer to have 1 picture out of 1000 with a vague artifact in the background than all my pictures with a soft filter. IMO Canon is too conservative on the matter.
Yes, especially the higher the resolution, the less the need for AA-filters as the artifacts are pushed to higher and higher frequencies.
 
Upvote 0
Mar 20, 2015
428
372
I am with you here, I got rid of the AA filter when switching to the 5DSr and and never looked back.

Your 5DSR has exactly the same AA filter as the 5DS, it just has a de-AA filter behind it too... it restores most of the original optical path but doesn't replicate non-AA to 100%.

We have never had a Canon camera delivered without an AA filter.
 
  • Like
Reactions: 1 user
Upvote 0

dtaylor

Canon 5Ds
Jul 26, 2011
1,805
1,433
I also don't want to see aliasing artifacts in my photos. And I only very rarely see them with my 5DSR, which is why I would like more sensors without an AA-filter as they give sharper images.

One of my very first shoots with the 5Ds showed moire in the groom's shirt. I would have hated to see it or clean it up without the AA filter attenuating it. While pixel peeping I could also find sections of his jacket and the bride's dress that would have broken out into full moire on a 5Dsr (i.e. they were showing hints of it).

Before buying I went cross-eyed looking at test shots for sharpness and fine detail differences. On sharpness I found the difference amounted to small changes in post sharpening. I sharpen each image to taste any way so I would rather deal with that than moire. On fine detail/resolved detail I could not find a convincing example where one actually resolved more than the other. (I did find a couple examples where the 5Ds shot seemed to suffer from some other factor which the reviewer pinned on the AA filter.)

That said it seems the 5Ds filter is weak which would make sense as the 5Dsr has to cancel that filter. It's not actually removed. A lot of other Canon cameras seem to have AA filters which are too strong. I'm in the camp that would like to see Canon continue to use AA filters, just weak ones.
 
  • Like
Reactions: 1 users
Upvote 0
Aug 26, 2015
1,380
1,042
This camera won't have an AA filter (not cancelled like the 5DsR) and it will be fairly weak for video and speed, simple as that. It will be interesting to see it against the GFX and other cameras.

If the highest resolution/IQ at reasonable ISOs is not a priority, it is not the ideal camera for those, so no point in complaining about things that it wasn't intended for. With the CR3 it will have crop modes and compressed Raw, not sure about smaller Raw options.

Of course most people would prefer to just have an EOS R Mark II or something with its missing features added in, but that will have wait for quite a while, and a higher-end, much more expensive model will get those 'fixed' first and other cameras from other manufactures will be released in the meantime. So the banter can continue forever.
 
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,351
22,524
One of my very first shoots with the 5Ds showed moire in the groom's shirt. I would have hated to see it or clean it up without the AA filter attenuating it. While pixel peeping I could also find sections of his jacket and the bride's dress that would have broken out into full moire on a 5Dsr (i.e. they were showing hints of it).

Before buying I went cross-eyed looking at test shots for sharpness and fine detail differences. On sharpness I found the difference amounted to small changes in post sharpening. I sharpen each image to taste any way so I would rather deal with that than moire. On fine detail/resolved detail I could not find a convincing example where one actually resolved more than the other. (I did find a couple examples where the 5Ds shot seemed to suffer from some other factor which the reviewer pinned on the AA filter.)

That said it seems the 5Ds filter is weak which would make sense as the 5Dsr has to cancel that filter. It's not actually removed. A lot of other Canon cameras seem to have AA filters which are too strong. I'm in the camp that would like to see Canon continue to use AA filters, just weak ones.
Optyczne and lensrentals have measured the MTFs with sharp lenses on the 5DSR and 5DS and found the 5DSR to be significantly sharper - see
https://www.optyczne.pl/324.4-Test_aparatu-Canon_EOS_5Ds_Rozdzielczość.html
https://www.lensrentals.com/blog/2015/06/canon-5ds-and-5ds-r-initial-resolution-tests/
Here are the Optyczne results

183937
 
Upvote 0

dtaylor

Canon 5Ds
Jul 26, 2011
1,805
1,433
Optyczne and lensrentals have measured the MTFs with sharp lenses on the 5DSR and 5DS and found the 5DSR to be significantly sharper - see

I don't print graphs. Here's the sharpness difference at full size. This is extreme hair splitting that would be invisible in any print with no post processing.

183938

Here it is with 35%/2px applied to the left.
183939

If the three were unlabeled and shuffled I don't know if I could accurately, repeatedly tell them apart while pixel peeping. No human has the eyesight to tell them apart in a print. The difference is less than the difference between fine art papers on an Epson Pro printer.
 
Last edited:
Upvote 0
Your 5DSR has exactly the same AA filter as the 5DS, it just has a de-AA filter behind it too... it restores most of the original optical path but doesn't replicate non-AA to 100%.

We have never had a Canon camera delivered without an AA filter.

I know it's just a "canceled" AA filter on the 5DSr, the same Nikon did on the D800E. The result tough is equivalent., though my A7R that doesn't have any AA filter, seems a micro hair sharper at pixel level.
 
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,351
22,524
I don't print graphs. Here's the sharpness difference at full size. This is extreme hair splitting that would be invisible in any print with no post processing.

View attachment 183938

Here it is with 35%/2px applied to the left.
View attachment 183939

If the three were unlabeled and shuffled I don't know if I could accurately, repeatedly tell them apart while pixel peeping. No human has the eyesight to tell them apart in a print. The difference is less than the difference between fine art papers on an Epson Pro printer.
You don't print graphs. I don't use prints of playing cards that lack fine detail to make judgements about resolution of fine detail.

Putting it scientifically, conventional sharpening increases edge sharpness, and so you could sharpen edges from a 5DS image to be very close to that from a 5DSR. But, when it comes to fine detail, like measured by lensrentals or optyczne, the 5DSR has higher resolution.
 
Last edited:
Upvote 0

dtaylor

Canon 5Ds
Jul 26, 2011
1,805
1,433
You don't print graphs. I don't use prints of playing cards that lack fine detail to make judgements about resolution of fine detail.

Considering the size of the cards in the full dpreview test scene they're a valid example. If you want to go 'finer' look at the fabric samples in the Imaging Resource test shots, or the fine print in the dpreview shot. Or hair, eyelashes, and skin pores in model shots. Still no difference that's visible outside of pixel peeping.

Putting it scientifically, conventional sharpening increases edge sharpness, and so you could sharpen edges from a 5DS image to be very close to that from a 5DSR.

You can manipulate either shot to produce a much higher MTF50 value than LensRentals recorded. However, the level of sharpening that produces the maximum MTF50 value in a test isn't going to look very pleasing to the eye. Such tests can be informative, but understand how they work, what their limitations are, and how they map to the real world. Especially in the age of digital processing.

But, when it comes to fine detail, like measured by lensrentals or optyczne, the 5DSR has higher resolution.

They didn't test fine detail or resolution, they tested sharpness.
 
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,351
22,524
Considering the size of the cards in the full dpreview test scene they're a valid example. If you want to go 'finer' look at the fabric samples in the Imaging Resource test shots, or the fine print in the dpreview shot. Or hair, eyelashes, and skin pores in model shots. Still no difference that's visible outside of pixel peeping.



You can manipulate either shot to produce a much higher MTF50 value than LensRentals recorded. However, the level of sharpening that produces the maximum MTF50 value in a test isn't going to look very pleasing to the eye. Such tests can be informative, but understand how they work, what their limitations are, and how they map to the real world. Especially in the age of digital processing.



They didn't test fine detail or resolution, they tested sharpness.
 
Upvote 0

dtaylor

Canon 5Ds
Jul 26, 2011
1,805
1,433
Read the y-axis of the graph: the units are lpmm = line pairs per mm ie resolution. The MTFs reported are resolution.

The MTF50 point is not a measure of "resolution" or "fine detail", i.e. extinction resolution. It's the traditional point chosen to measure perception of sharpness.

Edit: lpmm at MTF10 is traditionally the measurement point for total or extinction resolution.

Also: comparing numbers at one point is ultimately a poor way to judge how two optical elements (sensors; lenses) compare. We've always done it because people have an insatiable need for clear, simple scores. But seeing a full MTF curve is always better.

Let's say that Lens A tests to 80 lpmm MTF50 and Lens B to 60 lpmm MTF50. Is that because Lens B contrast at 80 lpmm is 49% or 11%? 80/60 may indicate very little perceptible difference if the graph for Lens B shows a shallow descent that just happened to dip below the test threshold sooner, or a large difference if the graph is nose diving. Every image comparison I've reviewed or made suggests the former for canceled AA vs. AA sensors (both 5Ds/sr and D800/E).

Again: I think Canon's AA filters in general are too strong. The 5Ds probably has a weak one as a consequence of needing to cancel it for the 5Dsr. But if you shoot weddings/fashion/urban landscape you really don't want to loose it completely. If Canon's next high resolution system is an R with no AA I'll certainly still be happy with it. But if they continue to give the choice I would likely make the same choice again.
 
Last edited:
Upvote 0