Resolution needed for printing is pretty much a matter of proportions.

There is a consensus that 300ppi is the target to use when printing an 8x12" image that will be viewed from about 10-12" (25-30cm). That's about all that normal eyesight can resolve.

That works out about 8.44MP, maybe all you will ever really need. That's because a person with normal eyesight (or reading glasses) can't look at an 8x12" image from any closer than the normal viewing distance (about the width of the print) without developing eyestrain and a headache.

You print that same 8.44MP image at 8x12 feet and view it from 12 feet away and still see the same level of detail. Change feet to meters and it still works.

The bottom line is that any camera that can capture an image with more than 8.44MP can print any image size you need, so long as you don't crop the image below 8.44MP.

Incidentally, Harrisonburg, VA, is just north of Fort Defiance and the Augusta Military Academy. I was there for about a year in the early 1950s. We marched in a parade in Harrisonburg. I was happy to transfer to a more comfortable school in Miami. AMA now a museum, a pretty spooky place.

I have figured out what is happening in the two photo viewers. It relates to the comment above.

When I view the full image on my 2k display (23½" wide, 82ppi) with the Windows Photo Viewer or in Capture One, the complete image covers a width of about 17", about 1390 pixels out of the 1920 for the full screen. The numbers are a little different for the 4k display but the same thing happens.

In order to get 6000 image pixels down to 1390, it has to be downsized by 1390/6000 or more than four times in each dimension. Detail and sharpness is lost during this downsizing and the highlights are darkened. But the image will actually have about the same tonality and sharpness of a 17" wide print.

As I zoom into the image, the highlights appear to get brighter because there is less downsizing.

But zooming in also shows how a larger print would look. I can only make a 16x24" with my P900. A print that size would mean that anyone looking at the print will probably view it a normal viewing distance slightly greater than the distance at which I view my desktop monitor. For a larger print a normal viewing distance would render about the same character and apparent highlight brightness.

The Capture One export is different issue. During export it offers the option for additional adjustments (Lightroom might also offer this):
https://static.uglyhedgehog.com/upload/2024/4/26/450682-adjustments.jpg
But doubling down on the sharpness already baked into the image (I normally leave it at Capture One's default) may not be a good idea.

I tried several alternatives - No Output Sharpening, Output Sharpening for Screen, Output Sharpening for Print and Disable All. They all do something a little different. Output Sharpening for Screen works well when reducing the image size for viewing but Output Sharpening for Print does not work well for viewing.

It seems that the best alternative for printing is use the full image, JPEG or TIFF makes no difference. I will have to do some more prints to see if there is any difference between No Output Sharpening and Output Sharpening for Print.

I wish it were that simple, but it's not. It's not even a question of sharpening or ICC profile in the Export dialog.

If you toggle between these two images, the highlights in the leaves are brighter in the in the reduced (4096 wide) version.

What's more, when I display either of them in the Windows Photo Viewer, I can increase the magnification by small increments (rather than going directly to 100%) and at the first increase, the highlights in the leaves get brighter.

If it were a simple matter of averaging of the pixel values, as you suggest, the highlights should get duller, not brighter.

Since this happens in the Photo Viewer as well as in the C1 Export step, there is something else happening that I can't explain.

I am sure that I am not saying it well. When I say reduce noise, perhaps it is better to say that it eliminates that which it recognizes as random. And it keeps that which reinforced as being signal. Or valid data if you will. So through the miracle of stacking (math and statistics), you end up seeing deeper and more clearly than you otherwise might.

But yeah, it depends on what you are seeking to see and understand. If you desire to capture the pure moment, rapidly unfolding then the only solution is to improve the resolution of it all. To capture more light more quickly and with greater sensitivity. Stacking is but the poor man's tool to make the most of what we are given.

The cosmos as we view it, it is SLOW and obscenely massively large, and so to be able to see it more fully, the cheating of stacking works very well.

And please don't take my wild rambling about this all to the bank. Like all photography, recognizing the difference between truth and lies requires a connection between the viewer and the object viewed and somehow a mechanical processes touches the soul and becomes Art. Lies and truth, signal and noise, but words that depend on the connection between the seer and the seen.

Or to say the same thing in a more powerful way... the famous quote - "And if you gaze for long into an abyss, the abyss gazes also into you."

thanks Sonny, and
Richard, thanks for asking. Yeah the mount tracks the stars (once it is polar aligned). Without tracking you are hard pressed to have an exposure longer than a couple of seconds for a telephoto lens before stars start becoming elongated. And then there is guiding... which is a smaller lens/sensor that takes photos every couple of seconds and makes small guiding adjustments to the tracking. This allows the larger telephoto lens to stay on target for 300 or 500 seconds. In the image of M51 for example we have about 35 images taken with each exposure being at 300 seconds, and then the software stacks (blends) the images together.

The stacking software is very sophisticated and does a wonderful job recognizing signal and getting rid of noise. For instance, each image taken might have a satellite trail in it, but they are eliminated from the final result because the software recognizes that the trail is not in all the images, so it gets ignored.