[btnapa]

It would have more, but the sensor's physical size is part of the difference in tonality rendering IMO, which I think would still give the edge to the Z for single shots.

I would think that the high ISO noise would be better on 645Z as well. TBH, I stayed with K1 instead of the 645Z and never felt I was missing much. Resolution was more than adequate and the dynamic range, especially in low ISO settings was more than I could ask for. I have 40"x60" prints from K1 files that look ridiculously good.

[repaap]

It would have more, but the sensor's physical size is part of the difference in tonality rendering IMO, which I think would still give the edge to the Z for single shots. Pixel Shift? Now, that's a different thing altogether---but of course PS isn't possible for lots of work. But apropos of PS, here's where I'd love to see AI tech deal with the artifacts....

I might be fine with a 45mp K1mkIII if it had improved Pixel Shift and greatly improved video---and the video might indeed be more viable with the 45mp sensor. Don't need 8K! But good 4K and I'd be very interested.

We shall see about that video. And yes, 8K is too much and you'll need good codecs. Oversampled 4K is very nice, but as we are spaking of photo first company, dunno if it is going to happen/how it would be done if it will.

45 MP as it is already is very nice, let alone PS.

[Rondec]

I'm starting to change my mind on super high pixel count sensors since I took delivery of a Fuji X-T5. It's an APS-C 40MP sensor, which if upsized to full frame would be close to 90MP.
What I am finding is that excellent technique and very high lens quality are absolutely required to get the most out of the sensor. Handheld shots have about the same level of fine detail as my old K3 had under similar conditions. To get the most out of the sensor I have to be on a solid tripod, my aperture choices are very limited, I have to use electronic shutter, and some of my older Fuji lenses are simply not up to the demands of the sensor.
A 60MP full frame sensor is probably exceeding the limits of how much effort most people are willing to put into technique lens quality and aperture limitations.

I don't particularly want to bump resolution a bunch. I do imagine that with a solid tripod, DFA * lens, and electronic shutter, you could get more resolution than what the K-1 offers. The megapixel race seems more about one-upping the last sensor than providing a feature that most photographers are needing.

The big benefit with higher megapixels is increasing print size (and I suppose cropping ability), but most of us aren't printing huge prints any more and so that is sort of wasted.

[Wheatfield]

I don't particularly want to bump resolution a bunch. I do imagine that with a solid tripod, DFA * lens, and electronic shutter, you could get more resolution than what the K-1 offers. The megapixel race seems more about one-upping the last sensor than providing a feature that most photographers are needing.

The big benefit with higher megapixels is increasing print size (and I suppose cropping ability), but most of us aren't printing huge prints any more and so that is sort of wasted.

The problem is that extra cropability is lost because the increase in resolution that one would expect from the higher pixel count is not there because the resolution of the captured image is no better than what one gets from a 24mp (in the example I gave) sensor.
If you want that cropability, everything has to be right to support it.
You may have the potentially higher resolution from the higher pixel count, but you don't have the fine detail to support a deep dive into the image for cropping.
40mp on APS-C is hard to exploit from a resolution perspective. 60mp on full frame will be almost as hard.

[biz-engineer]

The big benefit with higher megapixels is increasing print size (and I suppose cropping ability), but most of us aren't printing huge prints any more and so that is sort of wasted.

Contrary to popular opinion, higher pixel density benefits image quality on two things:
1) absence of aliasing (false colors and irregular sharp edges)
2) noise management, noise can be reduced with less impact on image detail
I am not concerned with noise management.

But there is something that is really problematic with sensors without optical low pass filter and insufficient pixel density (such as Z7, A7RIII, GFX50 and to less extent K1), the impact when up-scaling e.g with Gigapixel AI or Sharpen AI, the magnification of pixel artifacts.
This may seem counter intuitive, but for example, GFX50S files don't scale well due to the presence of aliasing at pixel level in the original file. AI sharpening applied to Z7 files look atrocious. The 40Mpixels of the X-H2 on the other hand, scale up and sharpen very well with Sharpen AI. After up-scaling and sharpening, A7RV files look better than GFX50. Simply, when re-sampling images, software doesn't quite know what to do with edges and color artifacts, those are either interpolated with adjacent pixels in the case of down-sampling, or used as texture when up-sampling.
It is counter intuitive, but the best image fidelity is achieved when the sensors out-resolve lenses, although most people will find that images are soft when zooming to 100%, but that's precisely when images are soft that they are free from false pixels.
Pentax are very aware of that, they provide two ways to correct this problem: AA simulator for shutter speed slower than 1/1000, or pixel shift for static subject matter. With over 60Mpixels on FF, there is no need to use AA simulation or pixel shift anymore and absence of motion artifacts in every single image taken.

[Rondec]

Contrary to popular opinion, higher pixel density benefits image quality on two things:
1) absence of aliasing (false colors and irregular sharp edges)
2) noise management, noise can be reduced with less impact on image detail
I am not concerned with noise management.

But there is something that is really problematic with sensors without optical low pass filter and insufficient pixel density (such as Z7, A7RIII, GFX50 and to less extent K1), the impact when up-scaling e.g with Gigapixel AI or Sharpen AI, the magnification of pixel artifacts.
This may seem counter intuitive, but for example, GFX50S files don't scale well due to the presence of aliasing at pixel level in the original file. AI sharpening applied to Z7 files look atrocious. The 40Mpixels of the X-H2 on the other hand, scale up and sharpen very well with Sharpen AI. After up-scaling and sharpening, A7RV files look better than GFX50. Simply, when re-sampling images, software doesn't quite know what to do with edges and color artifacts, those are either interpolated with adjacent pixels in the case of down-sampling, or used as texture when up-sampling.
It is counter intuitive, but the best image fidelity is achieved when the sensors out-resolve lenses, although most people will find that images are soft when zooming to 100%, but that's precisely when images are soft that they are free from false pixels.
Pentax are very aware of that, they provide two ways to correct this problem: AA simulator for shutter speed slower than 1/1000, or pixel shift for static subject matter. With over 60Mpixels on FF, there is no need to use AA simulation or pixel shift anymore and absence of motion artifacts in every single image taken.

I use a lot of pixel shift and that really takes care of most of the issue. I would think that pixel shift would still give color benefit on a 60 megapixel image, although you might struggle with scene movement more with that level of pixel density.

[JensE]

I use a lot of pixel shift and that really takes care of most of the issue. I would think that pixel shift would still give color benefit on a 60 megapixel image, although you might struggle with scene movement more with that level of pixel density.

The linear ratio of the pixel pitch is not that high: sqrt(36MP/60MP) = 0.77, which means that a potential 60MP pixel shift exposure still allows for 77% of the movement that the K-1 resolution allows. Noticeable, but a night and day difference.

[Wheatfield]

Contrary to popular opinion, higher pixel density benefits image quality on two things:
1) absence of aliasing (false colors and irregular sharp edges)
2) noise management, noise can be reduced with less impact on image detail
I am not concerned with noise management.

Aliasing is pretty much a non issue by 24mp in APS-C and 36mp in full frame. More MPs isn't going to hurt, but other factors will negate the advantage, and may well take care of aliasing on their own.
Regarding noise management and image detail, as I said earlier, unless all the photographer's ducks are lined up in a row, the detail isn't going to be there to be affected.
Until one has shot with a really high pixel count sensor, one won't know if their ancillary equipment such as lenses and tripods or technique is up to the sensor.
If the captured detail is no better than 36mp because one part of the imaging chain isn't up to snuff, photographers really won't see much benefit from the extra pixels.
Getting the most out of the X-T5 is a rigid exercise in technique combined with using a very stable and fairly heavy tripod (forget about multi section travel tripods), the very best glass available at their best apertures, fully electronic shutter, turn off the beeps too, and near perfect shooting conditions. Even a slight breeze is enough to nullify the extra pixels, and handholding? Forget about it, shake reduction or not.
Also, how stable is the sensor support? Is it stable enough to not let the sensor move at all? I don't know how accurate the X-T5 sensor positioner is.
The one in the K1 seems able to keep up, but I don't know if it would still do so with a 60mp sensor.
A 60mp full frame sensor would be more rigorous than I think a lot of people realize.

[biz-engineer]

I use a lot of pixel shift and that really takes care of most of the issue. I would think that pixel shift would still give color benefit on a 60 megapixel image, although you might struggle with scene movement more with that level of pixel density.

Pixel shift have no influence of overall color, only improves image resolution in areas not containing a balanced mix of color, e.g especially detail in red or blue areas. With 60Mpixels (or more) I wouldn't use pixel shift anymore, I'd stack a few frames to reduce noise if necessary, motion would appear as long exposure blur, with I much prefer over pixel shift artifacts, and no need for specific PP software to deal with brand specific pixel shift and losing camera color profile consistency.

---------- Post added 14-04-23 at 07:35 ----------

Aliasing is pretty much a non issue by 24mp in APS-C and 36mp in full frame.

You need to understand what aliasing is. Aliasing and moiré patterns are two different things. Aliasing is present in every pixel even in absence of periodic image pattern. Aliasing manifest at edge sharp edge in images, every edge of tree leaf, every sharp edge of object, building etc. Aliasing is what make people thing the image is sharp when it actually false edge! While moiré patterns are absent, aliasing is present, but incorrectly perceived as sharpness. High fidelity image should be soft when zoomed in at 100%.

[Wheatfield]

You need to understand what aliasing is. Aliasing and moiré patterns are two different things. Aliasing is present in every pixel even in absence of periodic image pattern. Aliasing manifest at edge sharp edge in images, every edge of tree leaf, every sharp edge of object, building etc. Aliasing is what make people thing the image is sharp when it actually false edge! While moiré patterns are absent, aliasing is present, but incorrectly perceived as sharpness. High fidelity image should be soft when zoomed in at 100%.

I know what aliasing is. Yes it's visual effect is lessened at higher megapixels, but there is a strong diminishment of returns as the megapixels increase, and much of what is left is negated simply because there isn't a strong increase in fine detail over 36mp for full frame unless the photographer gets everything right, the difficulty of which increases rather a lot at high megapixels.
It's just the way it is.

[biz-engineer]

Getting the most out of the X-T5 is a rigid exercise in technique combined with using a very stable and fairly heavy tripod

That's were our understanding differ. I think it's very easy to use the X-T5. Please take a moment to think about why Pentax engineering took the time to implement AA filter simulator on Pentax cameras and why Canon always kept optical low pass filters on their cameras. Two things in the camera world: commonly believed and accepted group trends, and technical realities. For example, if you look image stitching tutorials, every single video, article with tell you that you must level your tripod , rotate ballhead around tripod base, and 99% of tripod and ball head are designed accordingly, and technically that's wrong, I can write an article with image projection examples to prove that 99% of the internet is wrong (except ArcaSwiss who got it right with their P0 ballhead...). The other commonly accepted thing is gray scale camera are named monochrome cameras, everybody call it that way, but it's wrong, such cameras aren't monochrome because black , white and shades of gray don't have any color, cameras should be called "achromatic" cameras. Yet, everyone talk about monochrome. Same for aliasing, people confuse moiré patterns with aliasing, and they think aliasing is sharpness.

---------- Post added 14-04-23 at 07:49 ----------

there is a strong diminishment of returns as the megapixels increase

Yes, it's not efficient use of processing, lot of redundant pixel data. Take 8x10" B&W film shot at f/64, diffraction dominate film grain and film resolution by a fair amount, and that give the very best images, LF photographers don't shoot at f/8 in order to make film grain look sharper in images, on the contrary, they shoot 8x10" to wash out film grain. But it's also possible to shot 35mm film, f4, make film grain visible and think that images are sharp. It's a choice. You get the idea.

[Wheatfield]

That's were our understanding differ. I think it's very easy to use the X-T5. Please take a moment to think about why Pentax engineering took the time to implement AA filter simulator on Pentax cameras and why Canon always kept optical low pass filters on their cameras. Two things in the camera world: commonly believed and accepted group trends, and technical realities. For example, if you look image stitching tutorials, every single video, article with tell you that you must level your tripod , rotate ballhead around tripod base, and 99% of tripod and ball head are designed accordingly, and technically that's wrong, I can write an article with image projection examples to prove that 99% of the internet is wrong (except ArcaSwiss who got it right with their P0 ballhead...). The other commonly accepted thing is gray scale camera are named monochrome cameras, everybody call it that way, but it's wrong, such cameras aren't monochrome because black , white and shades of gray don't have any color, cameras should be called "achromatic" cameras. Yet, everyone talk about monochrome. Same for aliasing, people confuse moiré patterns with aliasing, and they think aliasing is sharpness.

---------- Post added 14-04-23 at 07:49 ----------


Yes, it's not efficient use of processing, lot of redundant pixel data. Take 8x10" B&W film shot at f/64, diffraction dominate film grain and film resolution by a fair amount, and that give the very best images, LF photographers don't shoot at f/8 in order to make film grain look sharper in images, on the contrary, they shoot 8x10" to wash out film grain. But it's also possible to shot 35mm film, f4, make film grain visible and think that images are sharp. It's a choice. You get the idea.

The X-T5 is easy to use, but getting the most out of the sensor is not. If your technique is not close to perfect all the time, you are not going to get the best the camera can give you.
Most of your post is non relevant nonsense. You are using false equivalencies and distractions that have absolutely no bearing on the subject at hand to muddy the discussion.

[Ian Stuart Forsyth]

why Canon always kept optical low pass filters on their camer

This is one of the reasons why I have 1 D810 and 2 D800 even with handheld and 600mm zoom lens it is very easy for me to get moiré and aliasing in bird feather details and use the D800 more
Using a better quality lens and the D800 can be problematic, with the D810 I really need to pay attention

[biz-engineer]

The X-T5 is easy to use, but getting the most out of the sensor is not. If your technique is not close to perfect all the time, you are not going to get the best the camera can give you.

Yes, i understand and agree. Also I guess the same difficulty would occur when shooting with 100Mpixel smartphones. It's just that camera systems are made of two parts: the optical system to yield a 2D image focused on a plane and a digital sensor to sample that image. What the higher resolution sensor does is give you a more refined representation of the same image projected by the lens. So you may have two different goals: one goal could be to get a more refined image of the same size viewed at the same distance, and another goal could be to view a larger image closer but if so be prepared to be disappointed, a bit like printing a 100Mp phone image at 30x40" and looking close.

---------- Post added 14-04-23 at 09:28 ----------

This is one of the reasons why I have 1 D810 and 2 D800 even with handheld and 600mm zoom lens it is very easy for me to get moiré and aliasing in bird feather details and use the D800 more

If you used a Pentax K1 pixel shift you wouldn't have any moiré, but you would have pixel artifacts. If you used a 5DS, then you would have had no moiré and no pixel shift artifacts. You could also have shot with a tele zoom on your D800 take two quick shots are different FL so that you tune out moiré because moiré occur as a result of pixel pitch relative to bird feather pitch, when you zoom in-out you change that ratio, so if one exposure show moiré , the second exposure at different FL will show none in the same image areas. I experimented with using two exposures workflow using a zoom, and it worked fantastic, my subject was moving a bit, but I used Hugin to align image stack with masking areas that moved, and I was very impressed with the results. I also took wide angle shot + zoomed in shot with 3x zoom, then upscale the wide angle shot in Hugin to match the zoom shot, getting whooping 150Mpixels stacked image out of two exposures only.

[biz-engineer]

The big benefit with higher megapixels is increasing print size (and I suppose cropping ability), but most of us aren't printing huge prints any more and so that is sort of wasted.

Users of proper cameras are now a minority, and within that minority only few people translate their images into prints, and a minority still get a DSLR. The trend is to use smartphones, instant post online and forget images the next day.
But, Ansel Adams said "the negative is the partition, and the print is the performance", who wants to keep the Ansel Adams spirit alive shouldn't follow the trends and trends exist so that some people do not follow them.
The very best way to not follow the trend, even better to go in opposite direction of the trend (remaining polite here), is to 1) use a DSLR, 2) go black & white with some unique toning 3) emphasize the originality further by printing big on silver photo paper 4) not displaying images online and 5) mount paper prints inside wood frame behind true glass (no plexiglass), add 5000K spot projectors on the ceiling at 30 degres angle to lighten up the prints and eliminate reflections on the glass.