How high resolution can a picture be before the human eye stops caring?

Question

I found an article discussing a collective that sells ludicrously high resolution panoramas, at least one of which is comprised of over 1000 images with a resolution of 7 gigapixels.

Other than having a marketing claim and the ability to zoom in super far onto an image on a computer and still see detail, would there be any reason to stitch an image like this together out of hundreds of pictures as opposed to a number far less than that?

At a normal viewing distance (to be generous, let's say one where the wider sides of the picture just barely fill the field of vision), would an observer be able to tell the difference in detail between a print of a 7 gigapixel image and a print of a 24 megapixel image?

There's surely a limit to the number of pixels humans can process.

Answer 1

I love this question.

There are several concepts of resolution and several aspects that affect each one. I will be playing with different names, sorry for that, ppi, lpi, pixel density, etc, but you simply focus on the number.

These variables are

1. Distance to the viewer/physical size.

2. Density of the display media.

3. Interaction with the image.

3. Interaction. On a computer you can interact with the image. If you use google maps with the satelite view, you are zooming because you do want to see details on a specific zone. The original image would be... realy, realy, realy, realy hughe, but what you want to see is a portion of it.

2. Leaving this interaction (zooming) aside, The resolution is caped to the resolution the media is using phisically.

On a monitor could have a resolution of 1920x1080px and if you put a bigger image this will not matter. On a print medium this maximum resolution is a bit harder to find.

But notice that i mentioned a generic monitor, regardless of the physical size of it.

When we introduce a physical size we use a new unit, unit per inch.

On an offset print (magazine) this resolution is normally 150 lpi. On a 23 diagonal inch monitor this unit density is arround 100 pxpi and on a 6 diagonal inch cellphone is about 350pxpi

1. But the most important thing here is what will the physical distance to the viewer this device will be from the user.

A 23 inch monitor at 20" will have more or less the same size to the viewer than a 6" cellphone at 4". (this are rough numbers... Im not using math to calculate it)

But if I want for example a really big print, and I want to see a lot of detail, for example a city map, where I want to see the streets and buildings I need more information, a bigger file.

But if the total image is going to be viewed at once this simplyfies things.

My rule of thumb is to consider the maximum dimension of the image (for example the base of a monitor, 20") as the distance from the viewer (20") to simplyfy things.

This means that an image used for a magazine, of 30cm at 300ppi can be used as a street artboard of 3 mt viewed at 3 mt with 30ppi.

30 cm at 300ppi are 3,543px on this long side.

This image is not that big, yes, a 24Mpx is fine to be used in most scenarios. a 24 Mpx image has 6,000px wide.

---

The original Retina Claim was to use a 300ppi device at 10 inches would match the maximum resolution of the average human vision. Some claim that it is more like 450ppi.

Another important issue is, do the viewer really care of this small resolutions?

A 300ppi vs a 600ppi device at 10 inches sounds like a great improvement... in reality, no, most people will not see a real practical difference between the two.

One has a micro tiny dot, the other has a half micro tiny dot.

This includes what important is the image to the viewer, does it contains small text or important details or not? etc.

I will edit this answer in a few months because I am preparing some practical investigations and surveys about this.

---

Commercial decisions.

Notice some patterns. Normally 4k resolution is used for big screens. It has sense, you want to stick your nose in the monitor to have an "environmental" experience.

And small phoens are pushing the resolution because you, with for example, googles where you insert the phone (and stick your nose, literally), you are using magnifying glasses for the same reason.

This images will be bigg because, again the interaction with it. Turning your head to see more of it, panoramic, or zooming in.

Answer 2

The apparent sharpness of images is based on the acuity of the human eye and the viewing conditions as to distance and lighting level as well as the contrast of the image.

For photographic purposes, it goes like this:

A friend is holding a coin and you see it as a disk. You friend backs away and you see the disk of the coin getting smaller. As the distance between you and the coin increases, at some point you will no longer perceive the coin as a disk, it becomes a point with no desirable dimension. If you have 20/20 vision and the day sunlit, this happens when the disk is 3000 diameters distant. As an example: A coin 10mm in diameter is viewed as a point from a distance of 30 meters (0.4 inches viewed from 100 feet. While the 1:3000 benchmark is bases for the unaided human eye in bright sun, it is far too stringent for photographic purposes. This is because photography yields much lower image contrast and our images are viewed under lower light levels. The photo industry generally accepts that a point is seen when the angle of view is about 3.4 minutes of arc. That’s a circle viewed from only 1000 times its diameter. This reduction in acuity fits photo prints and viewing screens nicely. This works out a circle 1/100 of an inch in diameter viewed from 10 inches or 1/50 of an inch viewed from 20 inches (0.5mm viewed from 500mm). The key here is, if the image is viewed from a typical viewing distance, it makes no difference if the circle is smaller than 0.5mm.

Now the pixel is the smallest fraction of a digital image that carries intelligence. When it comes to the camera, the size of the circle must be much smaller. This is because todays camera are miniature of subminiature thus the image they produce is likely useless unless enlarged. To make an 8X10 inch print or display from a full frame (FX) you must blowup this image a minim of 8.5X. Thus the size of the circle must be smaller i.e. 0.5 ÷ 8.5 = 0.058mm or smaller. To make the same 8X10 inch image from a DX you must magnify sensor image 1.5 X more (crop or magnification factor) thus 8.5 X 1.5 = 12.75. In other words, with a compact digital you must blowup the image 12.75X. Now the circle size at the imagining sensor is 0.5 ÷ 12.75 = 0.392. Let me add that the industry often uses a circle size of 1/1000 of the focal length. This methodology generally takes into account the viewing distance and the degree of magnification needed to have the final display perceived as being sharp.

If the final display requires an enormous blowup or the viewing distance will be nearby , then the circle size must be adjusted to accommodate more stringent conditions. As to the megapixel count, this will be a variable that intertwines with the sensor size, the degree of magnification, and the viewing distance. Also it should be noted that the contrast of the image also plays into the equations as does the brightness of the viewing light and the acuity of person doing the viewing. This is not an easy topic, it features with I call a dose of gobbledygook.