PPI ranks screens backwards
Pixels per inch measures how tightly packed a panel is. It does not measure how sharp it looks, because that depends on how much of your visual field each pixel occupies — and that depends on how far away you are.
The right measure is pixels per degree: how many pixels fall inside one degree of your field of view. And by that measure the ranking inverts.
Common screens by pixel density and by angular resolution at their real viewing distances| Screen | PPI | Typical distance | Pixels per degree |
|---|
| Flagship phone | 461 | 12″ | 97 |
|---|
| 13″ high-DPI laptop | 227 | 20″ | 79 |
|---|
| 27″ 4K monitor | 163 | 24″ | 68 |
|---|
| 27″ 1440p monitor | 109 | 24″ | 46 |
|---|
| 24″ 1080p monitor | 92 | 24″ | 38 |
|---|
| 55″ 4K television | 80 | 100″ | 140 |
|---|
The television has the lowest pixel density on the list and the highest angular resolution — nearly half again the phone’s. Nothing about the panel changed; you simply sit four times further from it, and distance is doing all the work.
Which is why “retina” was never a property of a screen. It was always a claim about a screen at a stated distance, and any screen becomes retina if you stand far enough back.
The monitors below the acuity limit
Normal vision — 20/20, or 6/6 — is defined as resolving detail that subtends one minute of arc. There are 60 arcminutes in a degree, so roughly 60 pixels per degree is where individual pixels stop being distinguishable.
Apply that to the table above and something uncomfortable falls out. The phone, the high-DPI laptop, the 4K monitor and both televisions clear it comfortably. But a 27-inch 1440p monitor at 46 PPD and a 24-inch 1080p monitor at 38 PPD do not. On both of those, at the distance people actually sit, the pixel structure is visible.
The tool reports a “retina distance” for exactly this reason: how far back you would have to sit for a given screen to reach 60 PPD. For a 24-inch 1080p monitor that is 37.5 inches — over three feet, which is considerably further than a desk allows. That screen was never going to look sharp where it is used.
None of which makes such a monitor bad. Text rendering, subpixel antialiasing and simply not looking closely all soften the effect, and plenty of people work happily on 1080p panels. But it does explain the immediate, obvious difference people notice moving to a high-DPI display, and it explains why that difference is much less dramatic on a television.
A diagonal is not an area
Screens are sold by their diagonal, which is a strange thing to quote because it is the one dimension that tells you least. Two screens with the same diagonal and different aspect ratios are different sizes.
A 27-inch screen at four aspect ratios| Ratio | Dimensions | Area |
|---|
| 4:3 | 21.6″ × 16.2″ | 350 sq in |
|---|
| 16:10 | 22.9″ × 14.3″ | 328 sq in |
|---|
| 16:9 | 23.5″ × 13.2″ | 312 sq in |
|---|
| 21:9 | 24.8″ × 10.6″ | 264 sq in |
|---|
A third more glass in the 4:3 panel than the 21:9 one, both labelled 27 inches. And the ultrawide is wider while being much shorter — 10.6 inches of height against 16.2.
This is why replacing a 27-inch 16:9 monitor with a 27-inch ultrawide often disappoints: you gain 1.3 inches of width and lose 2.6 inches of height, and most work is vertical. The honest comparison is on height, or on area, and neither is the number on the box.
A CSS pixel is not a pixel
There is a third quantity that shares the word “pixel” and is not one. A CSS pixel is a reference unit, defined as 1/96 of an inch, so that a 16-pixel font is about the same physical size everywhere regardless of the panel behind it.
devicePixelRatio is the bridge. On a flagship phone it is 3: the browser reports an 852-pixel-wide viewport while the panel has 2556 real pixels, and every CSS pixel is drawn with nine device pixels. On an ordinary monitor it is 1 and the two coincide.
The practical consequence is the one that catches people out with images. An image authored to fill an 852-pixel viewport contains 852 pixels of detail and is stretched across 2556 — so it looks soft on exactly the screens whose density was supposed to make it look good. This is what srcset and the “@2×” and “@3×” asset conventions exist to fix: supply the extra detail, and let the browser pick.
It also means “DPI” means three different things depending on who is speaking. A printer’s DPI is ink dots, a panel’s is really PPI, and a browser’s is a reference density that has no direct relationship to either.
What resolution is actually worth buying
Putting all of it together gives a usable rule: work out the pixels per degree at the distance you will actually be, and stop when it clears 60.
For a desk monitor at around two feet, 4K at 27 inches lands at 68 PPD — comfortably over, with nothing wasted. 1440p at the same size is 46 and visibly short. So the jump from 1440p to 4K at 27 inches is real and the jump from 4K to 5K is largely not, unless you sit unusually close.
For a television the picture is different and less flattering to the marketing. A 55-inch 4K set at eight feet already delivers 140 PPD, more than twice the acuity limit. An 8K set at the same distance delivers 280. The extra pixels are, for a viewer with normal vision at a normal distance, literally invisible — which is why television makers compete on brightness, contrast and colour instead, where the gains are real.
For a phone, 97 PPD at a foot is well past the limit, which is why flagship resolutions stopped climbing years ago and the effort moved to refresh rate and brightness.
The general shape: resolution has hit diminishing returns everywhere except close-range desktop work, and where it has, the remaining differences between screens are not about pixel count at all.
Sources and methodology
The geometry is Pythagoras and a tangent. What the sources settle is the two reference points everything hangs off: that normal visual acuity resolves about one arcminute, which is where the 60-pixels-per-degree figure comes from, and that a CSS pixel is defined as a fraction of an inch rather than as a hardware pixel. Both are conventions rather than facts of nature, and both are routinely misquoted.