How this instrument works
Most modern video assumes square pixels, where one stored pixel is exactly as wide as it is tall, so stored dimensions and displayed dimensions match. A number of older and broadcast video formats, however, store frames using non-square, rectangular pixels — described by a pixel aspect ratio, or PAR — which means the stored pixel grid and the correctly displayed image are different shapes.
PAR is the ratio of a single stored pixel's width to its height. A PAR of 1.0 means square pixels and no correction needed. A PAR different from 1.0 means each stored pixel must be stretched or squeezed horizontally when displayed, so the displayed width equals the stored pixel width multiplied by the PAR — the stored height is left unchanged, since PAR only adjusts pixel width relative to height.
This matters because several well-known formats deliberately use non-square pixels to save storage or bandwidth while still displaying correctly. NTSC DV and DVD store 720x480 pixels but use a PAR around 0.9091 to display correctly in 4:3, and the HDV 1440x1080 format uses a PAR of roughly 1.333 to stretch its stored frame up to full 16:9 HD width on playback.
- Enter the video's stored pixel dimensions into Stored frame width (px) and Stored frame height (px) — the raw pixel grid the file actually contains, before any pixel-shape correction.
- Enter the format's pixel aspect ratio into Pixel aspect ratio (PAR) — check your codec or format documentation; square-pixel formats use 1.0.
- Read Displayed width (px) for the width the video actually shows at once pixel-shape correction is applied, keeping the stored height unchanged.
- If your PAR value isn't already known, look it up from your capture format or codec's documentation rather than guessing — using the wrong PAR will stretch or squeeze the image incorrectly on playback.
Worked example — NTSC DV/DVD at 720x480 with PAR 0.9091
Enter 720 into Stored frame width (px), 480 into Stored frame height (px), and 0.9091 into Pixel aspect ratio (PAR) — the standard NTSC DV and DVD storage format, which uses a 10:11 (about 0.9091) PAR for correct 4:3 display. Displayed width (px) works out to 720 x 0.9091 = 654.552 pixels.
That 654.552 displayed width alongside the unchanged 480 stored height gives a displayed shape close to 4:3, which is exactly why the format uses non-square pixels in the first place — storing the frame at 720x480 saves data compared to storing a full square-pixel 4:3 grid, while the PAR correction restores the correct proportions on playback.
Questions
Why would a video format use non-square pixels at all?
Mainly for storage and bandwidth efficiency within a fixed broadcast standard. NTSC and PAL analog-derived formats were built around fixed line counts and sample rates, so digitising them at 720 pixels wide didn't naturally produce square pixels for a 4:3 or 16:9 picture. Rather than resample to an awkward square-pixel width, the format keeps 720 columns and encodes a pixel aspect ratio so playback software knows how to stretch the image correctly.
What happens if I ignore PAR and just play the stored pixels as square?
The image displays with the wrong proportions — circles look like ovals, and people look subtly too wide or too narrow depending on the direction of the distortion. NTSC DV's 720x480 frame, played back as square pixels without PAR correction, displays at roughly 3:2 instead of the correct 4:3, a visibly squashed picture that any consumer NTSC DV or DVD player automatically corrects for.
Does PAR ever apply to modern HD or 4K video?
Rarely, but not never. Most HD, 4K and web video uses a PAR of 1.0 (square pixels), where stored and displayed dimensions match exactly — this instrument would simply return the stored width unchanged. Some transitional and broadcast-derived formats, like HDV's 1440x1080 frame at a PAR of about 1.333, still use non-square pixels to stretch up to a full 1920-wide 16:9 HD display on playback.
How is pixel aspect ratio different from a video's overall frame aspect ratio?
Pixel aspect ratio describes the shape of one individual pixel; frame aspect ratio describes the shape of the whole picture. A video can have square pixels (PAR 1.0) and still be any overall frame shape, like 16:9 or 4:3, simply by choosing appropriate stored width and height. Non-square PAR only comes into play when the stored pixel grid's raw proportions don't already match the intended displayed frame shape.
Where do I find the correct PAR value for my footage?
Check your camera, capture card or codec's documentation — PAR is a property of the format or standard the footage was recorded or encoded in, not something to guess. Common reference values include 1.0 for square-pixel HD and most web video, roughly 0.9091 (10:11) for NTSC DV/DVD 4:3, and roughly 1.333 for HDV's 1440x1080 anamorphic HD format.