TOC

1. Four terms that get mixed up

2. Web

3. Social media

4. Video

5. Print: pixels = inches × PPI

6. Checking a file you already have

Throughout, each row is labeled:

• [Official] — published by the platform or standards body itself, checked for this article.

• [Convention] — widely used numbers that no standard publishes. Treat as safe defaults, not requirements.

• [Example] — an illustration of the arithmetic, not a spec.

Four terms that get mixed up

PPI and DPI are not interchangeable. Adobe puts it plainly: DPI measures physical ink dots in print, PPI measures pixels in the digital original — "DPI as the print output, and PPI as the digital input." The U.S. federal digitization guidelines (FADGI) went further in their 2023 edition and replaced "dpi" with "ppi" throughout, to keep the two measures apart. Vendors still label settings "dpi" out of habit: Epson's scan guide gives resolution in dpi while describing the pixels it captures.

One trap follows: a file's pixel count is fixed; the PPI value stored alongside it is only a label for intended print scale. Calling an 800-pixel image "300 dpi" does not add pixels. The real density at any print size is pixels ÷ inches, which is the calculation at the end of this article.

Web

Why 1200×630: it clears every consumer's minimum — Meta's own floor for link-share images is 600×315 — and the few pixels of difference between consumers (LinkedIn's 1200×627, Meta's 1.91:1) never show at preview scale. Ship 1200×630 and keep the text in the center band.

Social media

Checked against each platform's own help pages on 2026-10-09. Anything the platform does not publish is marked as convention.

Instagram

Facebook

X (formerly Twitter)

LinkedIn

Pinterest

YouTube

Video

Pixel counts for orientation: 1920×1080 = 2,073,600 (2.07 MP); 3840×2160 = 8,294,400 (8.29 MP); 1080×1920 = 2,073,600 (same pixels as 1080p, turned 90°).

Print: pixels = inches × PPI

Print has no fixed pixel size. It has paper sizes and a density you choose. The whole calculation is:

pixels on the long side = inches on that side × PPI

Worked examples at 300 PPI (the standard density for quality photo printing, per Adobe's PPI guide):

A4 is defined in millimetres, so convert first (25.4 mm = 1 in), then multiply:

210 mm ÷ 25.4 = 8.268 in      297 mm ÷ 25.4 = 11.693 in
8.268 × 300 = 2480.3 → 2480    11.693 × 300 = 3507.9 → 3508

Run the arithmetic backwards and PPI becomes a check: a 2400×3000 file placed at 8×10 in is exactly 300 PPI (2400 ÷ 8 = 300). The same file at 16×20 in would be 150 PPI — fewer pixels per inch, visible softness at close range.

How much density is enough? The FADGI technical guidelines for cultural heritage scanning set performance levels at 300 ppi for basic capture and 400 ppi for the top tier, and note that an image captured at 400 ppi renders at true size when displayed or printed at 400 ppi. Epson's scan guidance puts everyday numbers at 96–200 dpi for screen use and 200–300 dpi for print and OCR. For commercial photo printing, 300 PPI is the standard target — a benchmark, not a universal requirement; what actually decides quality is pixels ÷ inches at the size you print. For archival document work, the FADGI top tier runs 400.

Checking a file you already have

1. Read its pixel dimensions. File properties on Windows/macOS, or open it in any image viewer: the width × height in pixels is the raw truth.

2. Get the ratio. Divide width by height and reduce (1600÷900 = 1.778 → 16:9). The ratio tells you which crop family the file belongs to before any platform touches it.

3. Compute real PPI for a target print size. Pixels ÷ inches — but only when the shapes match. A 1200×1800 photo (2:3) on a 4×6 in sheet (also 2:3) is a clean 300 PPI both ways. On an 8×10 in sheet (4:5) the two shapes disagree, so you must choose: fit the whole photo inside the frame (6.67×10 in, uniform 180 PPI, side margins); fill the frame (scale to 8×12 in, crop 2 in off the height, uniform 150 PPI); or distort it to exactly 8×10 — the only case where the two directions disagree (150 across, 180 down), which is what stretching looks like. Scaled correctly, an image has the same density across and down; two different numbers only appear when the shapes are forced to disagree.

4. Ignore the PPI label when judging sharpness. It records intended scale, not content. Only the pixel count and the print size together determine what the reader sees.

Sources

Here is the catalogue you are probably looking for, followed by where each ratio actually shows up.

TOC

1. The common ratios at a glance

2. 16:9 — video and modern screens

3. 4:3 — the older screen shape

4. 3:2 — still photography

5. 1:1 — the square

6. 4:5 and 3:4 — portrait formats

7. 9:16 — full-screen phone video

8. 21:9 and other ultrawide numbers

9. 5:4 and 7:5 — the quiet standards

10. A-series paper — the √2 ratio

11. How to read any ratio yourself

The common ratios at a glance

Decimals are rounded to three places; each was checked by dividing the example dimensions (1920 ÷ 1080 = 1.7777…, 1080 ÷ 1350 = 0.8, 2560 ÷ 1080 = 2.3704…, and so on).

16:9 — video and modern screens

16:9 (1.778) is the frame shape of high-definition television. The ITU's Rec. 709, the standard behind HDTV, fixes the common image format at 1920×1080 pixels — 2,073,600 of them — at exactly 16:9. Its successor for ultra-HD, Rec. 2020, keeps the same shape at 3840×2160 and 7680×4320. Everything downstream inherits it: 720p (1280×720), 1080p, 4K, most TV panels, and a great many laptop screens.

For content, YouTube asks video thumbnails at 16:9 and recommends uploading them at 3840×2160 with a minimum width of 640 pixels. X's video specifications list 16:9 (and 1:1) as the recommended aspect ratios for uploads. When a platform "supports many ratios" it still tends to display 16:9 in its classic embeds and players.

4:3 — the older screen shape

4:3 (1.333) was the ratio of CRT televisions and the first generation of flat panels, and it survives in a few places with strong opinions behind them. Computer monitors of the era — VGA, SVGA, XGA — were all 4:3, so the first digital cameras matched them: most point-and-shoot cameras shoot 4:3, the Four Thirds and Micro Four Thirds systems are built around it, and Apple's iPad line kept the shape until the 11-inch iPad Pro moved to 1.43:1 in 2018. Silent-era 35 mm film was shot at 4:3 too, and Apple's iPhone camera guide still lists 4:3 alongside 16:9 and 1:1. A 4032×3024 photo checks out the same way: 4032 ÷ 3024 = 1.3333.

If an old photograph scans at 1024×768, that is 4:3. Print sizes are their own system: a 4×6-inch print is 3:2 (6 ÷ 4 = 1.5), so 35 mm and DSLR photos fill it exactly, while a 4:3 photo is cropped slightly top and bottom.

3:2 — still photography

3:2 (1.5) comes from 35 mm film: a 36×24 mm frame, which is 36 ÷ 24 = 1.5 exactly. Full-frame and APS-C digital sensors kept the family proportion, so a typical DSLR file of 3000×2000 pixels is 3:2. A handful of 2-in-1 laptops (Microsoft's Surface line among them) chose 3:2 for document work: taller screens fit more of a page.

3:2 and 4:3 are close enough that crops between them look mild, which is why photos "just get a little cut" in frames and feeds rather than failing to load.

1:1 — the square

1:1 is the neutral shape. Medium-format photographers used 6×6 cameras for it, platforms adopted it because it tiles well in grids, and profile pictures almost universally render square (they are then shown in a circle, which is a mask, not a different ratio). A 1080×1080 file has an aspect ratio of exactly 1:1 — width equals height, so 1080 ÷ 1080 = 1.

Instagram was built on it before adding taller formats; the square remains the one size every platform displays without negotiation.

4:5 and 3:4 — portrait formats

4:5 (0.8) is the standard portrait feed shape. Instagram's help documentation says feed photos keep their original resolution when their aspect ratio is between 1.91:1 and 3:4, and it spells the bound out in pixels: at a width of 1080, the height can be anything from 566 to 1440. So 1080×1350 (4:5) sits comfortably inside that range, and 1080×1440 (3:4, 0.75) is the tallest crop it preserves. Taller files get cropped.

Meta publishes a placement table showing which ratios each Facebook and Instagram ad placement accepts — 1.91:1, 16:9, 1:1, 4:5, 2:3 and 9:16 are the recurring column heads, with 9:16 marked "recommended for both images and videos" on the Stories and Reels rows. Portrait feed images exist because a phone is held upright: a 4:5 image occupies more of the screen than a 1:1 one.

On the film side, 4×5 large-format sheets give the ratio its other name; print it portrait and you have 4:5 (0.8), the same shape as the social feed portrait.

9:16 — full-screen phone video

9:16 (0.5625) is 16:9 stood on its end: 1080×1920, the resolution of full-screen vertical video. Instagram states that Reels accept aspect ratios from 1.91:1 up to 9:16, with a minimum resolution of 720 pixels and 30 fps. YouTube recommends 9:16 for Shorts thumbnails (2160×3840, minimum height 640 pixels). Pinterest recommends 9:16 or 1080×1920 for full-bleed video Pins.

Nothing here is universal — platforms also accept wider uploads and crop them — but 9:16 is the ratio designed for the phone held upright, and every vertical-video surface is built around it.

21:9 and other ultrawide numbers

"21:9" is a marketing name rather than a precise ratio. Written honestly, 21:9 = 7:3 = 2.333. The ultrawide monitors sold as 21:9 are usually something else: 2560×1080 is 64:27 (2.370) and 3440×1440 is 43:18 (2.389). Cinema's "Scope" convention is 2.39:1, which those panels approximate closely; the digital-cinema flat format is 1.85:1.

Use the label for the category (very wide screens, cinematic letterboxing), and the decimal for the math. If a spec sheet says 21:9, check the pixel dimensions before you compute a crop.

5:4 and 7:5 — the quiet standards

5:4 (1.25) shaped the SXGA era: 1280×1024 was the standard high-resolution LCD mode for years, and 5:4 is the shape of the 8×10-inch print (10 ÷ 8 = 1.25). 7:5 (1.4) is what you get from a 5×7 inch print held landscape: 7 inches wide, 5 inches tall (7 ÷ 5 = 1.4). Neither is fashionable, both still appear in frame shops, scanner presets and inherited monitor resolutions.

A-series paper — the √2 ratio

ISO 216 defines the A series: A0 down to A6, each size half the area of the next larger one. The trick is that halving preserves the shape, which requires a width:height ratio of √2 (about 1.414). A4 is 210×297 mm; 297 ÷ 210 = 1.4143, the ideal √2 = 1.4142 rounded to whole millimetres. Fold an A4 page in half and you get A5 with the same proportions — that is the property the ratio buys you.

The B series, also defined in ISO 216, follows the same logic. If you have ever wondered why an A4 page scaled up to A3 "just fits", the ratio is the reason: A3 is A4 with every dimension multiplied by √2 (210 × 1.414 ≈ 297).

How to read any ratio yourself

Two moves cover almost everything:

1. Get the decimal. Divide width by height. 1920 ÷ 1080 = 1.778 → 16:9 territory. 1080 ÷ 1350 = 0.8 → exactly 4:5.

2. Reduce the pair like a fraction. Divide both numbers by their greatest common divisor: 1920 and 1080 share 120, giving 16 and 9.

Sanity check: a decimal above 1 is landscape (width-led), below 1 is portrait (height-led), exactly 1 is square.

Sources