Aspect Ratio Converter

Changing a width and a height independently does not convert a file to a new ratio; it changes the shape of every object inside it. A face becomes wider, a circle becomes an oval, and type that was set on a grid no longer sits on one. Every pixel survives, which is why the file size looks reasonable and the result still looks wrong.

The two methods that keep the picture intact are crop and pad. Cropping holds one side of the original and cuts the other away, so the frame fills up but pixels are discarded. Padding keeps every original pixel and extends the canvas until it reaches the new shape, so nothing is lost but the picture no longer fills its frame.

There is a third option that beats both, and it is worth saying plainly because tools tend to hide it: go back to the source and export again at the target ratio. Cropping and padding are recovery methods for files you cannot re-export. If you still have the layered original or the camera file, re-exporting costs a minute and loses nothing.

The arithmetic is two divisions. Working it out explicitly is worth the minute, because it tells you the cost before you commit the edit.

1. Write the target ratio as a single number by dividing the first figure by the second. For 4:5 that is 4 ÷ 5 = 0.8; for 9:16 it is 9 ÷ 16 = 0.5625.
2. Divide the source width by the source height to get its own ratio.
3. If the source ratio is larger than the target, the source is wider than the frame: cropping keeps the height and narrows the width, so the cropped width is the source height multiplied by the target number.
4. If the source ratio is smaller than the target, the source is taller than the frame: cropping keeps the width and shortens the height, so the cropped height is the source width divided by the target number.
5. For padding, do the opposite side: the frame has to be big enough to hold both the width and the height the target shape demands, so take whichever is larger and derive the other from it.

Worked example

Convert 1080 × 1920 to 16:9. The target number is 16 ÷ 9 = 1.778. The source ratio is 1080 ÷ 1920 = 0.5625, which is smaller, so the source is taller than the frame: cropping keeps the width of 1080 and shortens the height to 1080 ÷ 1.778 = 607.5, which rounds to 608. The cropped result is 1080 × 608, and it keeps 31.7% of the original pixels — the rest was above and below the frame. Padding goes the other way: a 16:9 frame holding the full 1920 of height must be 1920 × 1.778 = 3413 wide, so the padded canvas is 3413 × 1920 and every original pixel survives at full size, with 2333 pixels of background beside it.

Crop when the edges of the frame are expendable. A photograph with a centred subject, a screenshot of a single window, or a graphic whose margins are empty all survive cropping well, because the part that gets cut carries no information.

Pad when they are not. Text that runs to the edges of the frame, a logo locked to a corner, a wide group shot — anything whose meaning reaches the boundary — cannot afford to lose the outer band, and background is the cheaper currency.

There is also a third case that decides itself. If the target frame is only slightly different from the source, both methods are cheap and the choice barely matters. It is the conversions between a landscape and a portrait shape that hurt, because that is where one of the two sides has to give way almost entirely.

• Compare the two percentages first. A conversion that costs 5% either way can be decided on taste; one that costs 68% cannot.
• Check whether the crop lands on a face or a headline. The tool reports which edges go, not what was underneath them.
• Prefer padding when the piece will be seen at more than one ratio. Background bars are uniform and predictable; a crop that works in one placement can decapitate the subject in another.
• Do not pad twice. Extending the canvas and then scaling the result to fill the frame is just cropping with extra steps, and it softens the image on the way.

Converting changes where the edges of a picture are. It cannot add anything that was never in the frame. A 4:5 portrait crop of a wide landscape does not create the missing height; it discards the sides and shows the middle, which is why the result often looks like a photograph of a corridor.

The same limit applies in reverse. Padding a portrait into a wide frame adds background, not information, so the picture sits in a band across the middle of the frame and the rest is empty. Deliberate use of that space — a caption, a colour field, a repeated logo — is what separates a padded image that looks designed from one that looks like a mistake.

Everything above is about a file that already exists. When you still get to choose the framing, the arithmetic turns friendly. Compose at the widest ratio the asset has to serve, keep the subject inside the narrowest one, and every later conversion becomes a plain crop with nothing important near the edges.

Concretely: if one shoot has to yield both a 16:9 upload and a 9:16 vertical cut, frame for 16:9 but hold the subject and any text inside a centred 9:16 column. That column covers only about a third of the wide frame's area, so it is a genuine constraint — but anything placed inside it survives both versions, while the wide framing still reads as a normal composition.

This is the only advice on this page that costs nothing to follow and cannot fail later. Converting a file you can no longer re-shoot means accepting a real loss — two thirds of the pixels in the worst cases. Planning for both ratios up front means that loss was never going to happen.

• Settle every ratio the asset must serve before you frame it, not after the first platform crops something off.
• Treat the widest ratio as the master and the narrowest as the safe zone. Whatever sits inside the safe zone survives every cut.
• Keep type and logos well inside that zone. Lettering is the first thing a crop removes and the least forgiving when it goes.

If you are converting to a specific placement rather than to a bare ratio, the platform pages list the size each one expects and the area its own interface covers up. That is usually the number that decides the crop.

• Instagram: Story size
• YouTube: Thumbnail size
• X: Header size
• Pinterest: Pin size
• LinkedIn: Profile banner size
• TikTok: Video size
• 16:9 aspect ratio calculator
• 9:16 aspect ratio calculator
• 4:3 aspect ratio calculator
• 1:1 aspect ratio calculator
• Social media aspect ratio guide
• Screen Ratio Calculator

The aspect ratio calculator names the shape a size already is, and the platform sizes index gives the exact upload size for each placement.

Add a free, live aspect ratio calculator to any page in one line of code - no account, no API key, no ads. Every embed credits the tool with a Powered by AspectCalc link, so your readers always know where it came from. See the full embedding guide for height and styling options.

<div class="aspectcalc-embed" style="max-width:100%">
  <iframe id="aspectcalc-frame" src="https://aspectcalc.com/embed?h=640"
          width="100%" height="700" style="border:0;border-radius:12px"
          loading="lazy" title="AspectCalc aspect ratio calculator"></iframe>
  <p style="margin:8px 0 0;font:13px/1.4 system-ui,sans-serif;color:#64748b;text-align:center">Powered by <a href="https://aspectcalc.com/" style="color:#2563eb;text-decoration:none">AspectCalc</a> - free aspect ratio calculator</p>
</div>
<script>
  // Optional: grow the frame to the widget's real height.
  // Delete this whole block to keep a fixed-height frame instead.
  addEventListener('message', function (e) {
    if (e.origin !== 'https://aspectcalc.com') return;
    if (!e.data || e.data.type !== 'aspectcalc-height') return;
    var f = document.getElementById('aspectcalc-frame');
    if (f) f.style.height = e.data.height + 'px';
  });
</script>

The frame starts at a fixed height, so your layout never jumps. Resize it with ?h= - for example https://aspectcalc.com/embed?h=520 - or keep the small script in the snippet and the frame will follow the widget's real height instead.