rmalarz Loc: Tempe, Arizona

https://www.digitalcameraworld.com/news/we-should-stop-measuring-resolution-with-megapixels?fbclid=IwAR0qxclb9UPDKfez3x1BX5QMoas-SNvdhqt7ZPwJwA99aIFHrEjLOGgvZ9s
--Bob

camerapapi Loc: Miami, Fl.

Bob, in the past we were all familiar with the "megapixels race." Camera manufacturers were using more and more megapixels to entice buyers into buying their cameras. Indeed, a double of megapixels is needed to see significant changes.
You remember the Nikon D70, a camera with only 6.1 effective megapixels. A 13x18 inch enlargement was a piece of cake and going larger was too. We do not take into consideration the distance at which an enlargement should be looked at. A mural does not need a lot of resolution.

When Nikon introduced the D800 with 36.3 million pixels I was using if memory does not fail a D300 with only 12 megapixels. I took a workshop in West Virginia to photograph in the Canaan Valley wilderness. The instructor had a D800 and when his images and mine were in the computer I could not see the difference when the images were enlarged to 50% and that already is a big enlargement. His prints were not very different from mine. I began to question why so many megapixels.

I am not familiar with measuring the horizontal resolution of cameras, indeed I have never made any claims that I know the subject I only base my observations on what I see. The D70 was a very capable camera when it came to prints and so was the D300. Look at the small Olympus cameras, for years Olympus kept their 17 megapixel sensors and just a few years back they began to introduce 20 megapixels in their cameras. Is anybody going to tell me that there is a big difference in resolution between 17 and 20 megapixels? I do not think so.

Megapixels and resolution is a very interesting topic that we do not discuss here often, perhaps the majority of us do not understand it very well. I thank you for such an interesting article.

Longshadow Loc: Audubon, PA, United States

I'm primarily interested in the fact that with more pixels, I can crop more in the editor and expand the crop to the "canvas" size with less pixelation.

Agreed, the viewing distance changes with the size of the print, but too many people view the "details" of, lets say an 8x12, from less than a foot when they are holding a print.
I'd love to get a foot from a billboard and see what the "pixels" look like. The billboard looks great from the interstate.

Many printers cannot "print" a bazillion pixel image, the print algorithms simply combine/drop what it can't handle.

But I suppose it depends on what one wants to be concerned.

srt101fan

For some reason the idea of favoring higher resolution cameras because it allows greater flexibility in cropping gets roundly pooh-poohed here. It’s apparently some mysterious aversion to “throwing away pixels”.

Longshadow Loc: Audubon, PA, United States

You mean some people don't like my opinion(s) or needs?....
Damn.
I would have never thunk that.

(If I don't want some pixels, I will throw them away.)

zug55 Loc: Naivasha, Kenya, and Austin, Texas

How one counts megapixels is just a matter of convention. You could make the same argument about how f-stops are calculated. How does it make numeric sense that a lens at f/2.8 transmits twice as as much light than at f/4? The ratio works if you square the numbers--2.8 squared is almost 8 (7.84), while 4 squared is 16. Which is the method by with megapixels are counted. (To make this even more confusing, this is also how that ISO numbering system works.)

Megapixel counts do matter, no matter how you count them. But there is some truth to the basic argument presented in the article--which the author does not address. A high MP count does not guarantee a high-resolution image. If that were the case cell phones, often with over 100 MP, would offer the best resolution, which they clearly don't.

What matters is the size of photosites--photosites provide the pixel's visual data. Their size is both a function of megapixel count and sensor size. Cell phone photisites are extremely small which means that they capture much less light, which translates into lower image quality (resolution, dynamic range, noise, color saturation, etc. etc.) and particularly inferior low-light capabilities.

A MP count comparison thus only is meaningful if it factors in the sensor size. Like with other numeric conventions, photographers need to learn what these numbers actually mean.

Wallen Loc: Middle Earth

The main takeaway is that pixel size (and amount in megapixels) should be matched to a sharp lens and vice versa. Add to that the need for a better focus mechanism and less operator error. Sharp lens and high megapixel with smaller pixel diameter sensors are less forgiving to focus & operator errors, as well as physical & mechanical issues.
https://www.uglyhedgehog.com/t-585722-1.html

Longshadow Loc: Audubon, PA, United States

Sharp lens and a High pixel count or sharp lens and a low pixel count, the comparison is between pixel counts.
Why complicate it with additional variables like lens quality, focus ability, .....
Simply compare the difference in pixel count, no matter what the other variables may be.

zug55 Loc: Naivasha, Kenya, and Austin, Texas

I would agree, with the exception that pixel size, that is the size of photosites, matters also, and with it sensor size. 24 MP on a full-frame sensor do not compare to 24 MP on an APS-C sensor--or on a cell phone.

Longshadow Loc: Audubon, PA, United States

Yea, that's part of it also.

Wallen Loc: Middle Earth

Because it is a system. One part always affect the other parts.

Take the pixel diameter as an example. With the same sensor size, higher pixel count will mean smaller pixel diameter. Smaller pixel diameter means more resolution capability, but that is only capability. To enable that, we need a lens that is also capable. And even if the lens is capable, the operator with the camera body, needs to focus that lens correctly.

The system works together. Compromise one aspect and it will not be any better than the lower tier group.

zug55 Loc: Naivasha, Kenya, and Austin, Texas

This is all kind of obvious. If you have a high-performance car you only get the best performance when you use high grade motor oil and gasoline. Also, the car runs better if you paint it red. Okay, just kidding about that one. Better lenses clearly give you better performance, regardless of pixel size and count.

However, one of your statements is not true: "Smaller pixel diameter means more resolution capability." If that were the case, I would ditch my full-frame camera and only use my cell phone. Generally, you get a higher resolution from a higher pixel count. However, smaller pixels (technically photosites) store less data because, well, they are smaller. What that means that you may get better resolution in good light, but the low-light performance falls off. This is why sensor size matters: a larger sensor has more space for pixels, but also allows them to be larger.

I shoot with two full-frame Sony bodies: A7 III (24 MP) and A7R III (42 MP). In good light I use the A7R because of the great resolution, micro-contrasts, better ability to crop, etc. But because of the smaller pixel size you get inferior low-light performance than with the A7.

I used to shoot with a Sony a6000, an APS-C camera with 24 MP. Its performance was worse across the board than that of the A7 III that also has 24 MP, which means that the pixels on the a6000 are smaller. (You could also argue that the A7 III has a better sensor, which is true.) Same number of pixels, but smaller pixels, resulting in lower performance.

Wallen Loc: Middle Earth

You have misread the context.

"With the same sensor size, higher pixel count will mean smaller pixel diameter. Smaller pixel diameter means more resolution capability"

This is about increasing the number of pixel per area. IE:
a full frame with 6mp vs a full frame with 24mp.
A celphone with 1mp vs a celphone with 5mp
Pixel density, or more pixels per sensor size "may*" improve resolution.

The sensors photosites or pixels do not store data. They convert light to electrical signals.
A pixels' response/sensitivity is altogether another topic. Not only size affects a pixels capability but other stuffs as well, like how old the technology, filters, SNR, operating temperature, bit depth etc.

Theoretically, we want the smallest and as many pixel as possible. That will allow us to define the smallest line hence the sharpest & true to life image. Technology, in many aspects, are not there yet.

*Note that i said "capability" in the original post. Because it still depend if the other parts of the system can support the increased density. It will need a lens with a resolving power matched to the sensors.
It won't matter how much MP the sensor has if the lens is soft.
Changing to a sharper lens would not matter if the sensor could not define the lenses resolution.

It's a system. IMHO, the only element that can be changed that will provide the greatest improvement (or not) is the person operating the camera.

Longshadow Loc: Audubon, PA, United States

True for the whole picture, but if one is only reviewing/analyzing a lower tier group.
(Don't forget filter quality, how clean the optic system may be, and if using a tripod or not.......)