Aspect Ratio Calculator
Convert width and height into ratios like 1:1, 4:5, and 9:16 used by social platforms.
- Aspect ratio
- 16:9
- Decimal
- 1.78
- Width
- 1920
- Height
- 1080
Runs in your browser. Nothing you enter is uploaded.
Overview
Two images at 1920×1080 and 1280×720 are different sizes and the same shape. Ratio is the property that survives scaling, which makes it the thing to reason about when you are deciding whether an asset will fit a slot, and pixel dimensions the thing to reason about when you are exporting it.
This calculator converts between the two. Enter a width and a height and it reduces them to their simplest whole-number ratio and shows the same relationship as a decimal, with a set of buttons for the common social and video dimensions so you can load a known shape and compare.
The reduction is exact arithmetic, which means it is also occasionally unhelpful - some dimension pairs do not reduce to anything recognisable, and knowing why saves you assuming the tool is wrong.
Reduced ratios and the ones that do not reduce cleanly
The ratio shown is the fully reduced form, obtained by dividing both numbers by their greatest common divisor. 1920×1080 reduces to 16:9 because both share a factor of 120. 1080×1350 reduces to 4:5 because both share 270. These are the satisfying cases, and they are satisfying because the dimensions were chosen to be clean multiples in the first place.
Arbitrary dimensions do not cooperate. A photograph cropped by hand to 1187×763 has a greatest common divisor of 1, so the "reduced" ratio is 1187:763 - technically correct and completely useless. This is the moment the decimal figure becomes the more helpful of the two outputs: 1.56 tells you immediately that the crop is close to 14:9 and nowhere near 16:9, which the whole-number form obscures entirely.
So read the decimal when the ratio looks strange, and read the ratio when it looks clean. A practical habit is to compare decimals rather than ratios when checking whether an image will fit: 16:9 is 1.78, 4:5 is 0.80, 1:1 is 1.00, and 9:16 is 0.56. Anything within about 0.02 of a target will letterbox imperceptibly; anything further will visibly crop.
- 16:9 = 1.78 · 4:5 = 0.80 · 1:1 = 1.00 · 9:16 = 0.56 · 1.91:1 = 1.91
- A ratio with large numbers on both sides means the dimensions share no common factor.
- Compare decimals, not ratios, when checking fit against a target.
- Within about 0.02 of a target ratio, cropping is imperceptible.
Matching a ratio to the slot it has to fill
Every destination has a shape it will display and a behaviour for anything that does not match. Feed posts on Instagram accept a band of ratios and crop outside it. Stories and Reels are 9:16 and will letterbox or fill anything else. A YouTube player is 16:9 and pillarboxes a vertical upload. An Open Graph card is nominally 1.91:1 and crops from the centre.
The consequence is that the ratio decision should happen when you frame or crop, not when you export. Deciding to publish a 16:9 landscape shot as a 4:5 portrait feed post means throwing away nearly half the width, and if the subject sits near an edge there is no crop that keeps it. Shooting or framing with the destination ratio in mind costs nothing and removes the problem.
The presets here cover the shapes that come up most often - square and portrait feed posts, the 9:16 vertical of Stories and Reels, a 16:9 video thumbnail, and the wider cards used for link previews. Loading one and then adjusting a single dimension is a quick way to see how far your own numbers are from a standard.
Euclid’s algorithm on width and height
The greatest common divisor of your two numbers is found by the Euclidean method: repeatedly replace the larger number with the remainder of dividing it by the smaller until the remainder is zero, at which point the surviving value is the divisor. Both inputs are then divided by it to give the reduced ratio, which is displayed as two whole numbers separated by a colon.
The decimal is simply width divided by height, rounded to two places. It is computed independently of the reduction, so it stays informative on dimension pairs where the ratio does not simplify. Values above 1 are landscape, below 1 are portrait, and exactly 1 is square.
Both inputs are integers with a floor of 1 - an empty or invalid entry is coerced to 1 rather than producing an error, which is why clearing the field shows a ratio against 1 rather than nothing. There is no rounding or tolerance in the reduction, so the tool will not tell you that 1187:763 is "approximately 14:9"; that inference is yours to make from the decimal.
1920×1080 and 1080×1350 side by side
Input
First: width 1920, height 1080 Second: width 1080, height 1350
Output
First → Aspect ratio 16:9 · Decimal 1.78 Second → Aspect ratio 4:5 · Decimal 0.80
These are the two shapes most content has to move between: the landscape 16:9 that video is shot in and the portrait 4:5 that gets the most vertical space in a feed. The decimals show how far apart they are - 1.78 against 0.80, a factor of more than two. Converting the first into the second means keeping roughly 45% of the original width, which is why a 16:9 frame with the subject off-centre often has no usable 4:5 crop at all and needs reshooting rather than recropping.
Limitations
- The reduction is exact with no tolerance, so dimensions that share no common factor produce a large unhelpful ratio rather than a nearest standard.
- Inputs are whole pixels only. Non-integer or fractional dimensions cannot be entered.
- Nothing is validated against platform requirements - the tool does not know which ratios a given destination accepts.
- The presets are a fixed list and reflect conventions at the time of writing rather than live platform specifications.
- No image is involved, so the calculator cannot tell you whether a specific crop keeps your subject in frame.
- It converts between ratio and dimensions only; it does not compute the dimensions needed to hit a target ratio at a given width.
When not to use this tool
- If you need to actually produce the image at a target ratio, the image resizer applies the dimensions and handles the fit.
- If the question is which crop keeps the subject, work visually in the image cropper rather than numerically here.
How to use it
- 1
Enter the pixel dimensions you actually have, or load the preset for the destination you are targeting.
- 2
Read the reduced ratio if it looks clean and the decimal if it does not - the decimal is the reliable figure for comparison.
- 3
Compare that decimal against your target’s: within 0.02 is safe, beyond 0.05 means visible cropping or bars.
- 4
If the gap is large, decide the ratio at the framing or cropping stage rather than trying to fix it at export.
Questions
Why does my ratio show as something like 1187:763?
Because those two numbers have no common factor, so the fully reduced form is the numbers themselves. It means the dimensions came from a manual crop rather than a standard size. Read the decimal instead - it tells you which standard ratio the crop is closest to.
Which ratio gets the most space in a feed?
Portrait, generally - 4:5 on Instagram’s feed and 9:16 for Reels and Stories occupy more vertical screen than square or landscape. That is a display-area argument rather than a performance one, but more screen means more chance of stopping a scroll.
Is 16:9 still the right ratio for video?
For anything watched on a player - YouTube, an embedded video, a television - yes, because that is the shape of the player. For anything watched in a feed on a phone, vertical is the shape of the screen. Many people now frame 16:9 with a vertical-safe centre so one shoot yields both.
How close does my ratio need to be to a target?
Within about 0.02 in decimal terms is imperceptible in practice - the crop or letterbox amounts to a few pixels. Beyond roughly 0.05 the difference becomes visible as trimmed edges or bars, and that is the point at which it is worth producing a dedicated export rather than letting the platform decide.