Media Glossary

Plain-English definitions for resolution, aspect ratio, bitrate, file formats, and the other technical terms you'll meet while compressing media — each with a real-world example.

Resolution & Dimensions

The number of pixels that make up an image or video frame, usually written as width × height (e.g. 1920 × 1080). Higher resolution means more detail but also a larger file size.

Example

A 1080p HD video has a resolution of 1920 × 1080 pixels.

The smallest single point of color in a digital image. Images and videos are made up of millions of pixels arranged in a grid.

Example

A 4K photo contains roughly 8.3 million pixels.

One million pixels. A 12 MP camera produces images roughly 4000 × 3000 pixels. Megapixels measure total pixel count, not necessarily image quality.

Example

A 12 MP phone camera outputs about 4000 × 3000 pixels.

Dots Per Inch / Pixels Per Inch — how many pixels are packed into each physical inch when an image is printed or displayed. Higher DPI means sharper prints.

Example

A print at 300 DPI looks crisp; the same image printed at 72 DPI looks soft and blocky on paper.

The width and height of an image or video, measured in pixels. Sometimes called the "size" of the media, not to be confused with file size.

Example

An Instagram square post uses dimensions of 1080 × 1080 pixels.

Reducing an image or video's resolution by throwing away pixels. Used to shrink files for the web or a specific platform.

Example

Shrinking a 4000 px-wide DSLR photo down to 1080 px for Instagram.

Enlarging media by adding pixels that weren't in the original. It can't add real detail, so results often look soft or blurry.

Example

Scaling a 500 px logo up to 2000 px makes its edges blurry.

A small, reduced-size preview of an image or video used to represent the full media before it's opened or played.

Example

YouTube shows a 1280 × 720 thumbnail for every video on its homepage.

The physical dimensions an image will occupy when printed, determined by its pixel dimensions divided by the DPI.

Example

At 300 DPI, a 3000 × 3000 px image prints at a sharp 10 × 10 inches.

Changing the pixel count of an image — either up or down — by recalculating pixel values. Downsampling and upsampling are both forms of resampling.

Example

Resampling an 8 MP photo to 2 MP keeps the same visual content but with fewer pixels.

Aspect Ratio

The proportional relationship between an image or video's width and height, written as W:H (e.g. 16:9). It describes the shape of the frame, not its size.

Example

A 16:9 video is 16 units wide for every 9 units tall.

A square aspect ratio where width equals height. The classic Instagram feed format.

Example

An Instagram square post at 1080 × 1080 pixels is 1:1.

A slightly squarish rectangle — wider than it is tall. Traditional TV and many presentation slides use 4:3.

Example

A 1024 × 768 PowerPoint slide is in 4:3.

The standard widescreen ratio for YouTube, HDTV, and most landscape video.

Example

A Full HD video at 1920 × 1080 is 16:9.

A tall, portrait orientation — the vertical flip of 16:9. Used for vertical video platforms.

Example

A TikTok shot at 1080 × 1920 pixels is 9:16.

A ratio slightly wider than 4:3, common in DSLR photography. Many still cameras shoot natively in 3:2.

Example

A 6000 × 4000 photo from a DSLR is in 3:2.

A very wide, cinematic aspect ratio used for films and ultrawide monitors.

Example

A movie theater frame is often 21:9, wider than a standard 16:9 TV.

A portrait ratio taller than square but not full vertical — popular for Instagram portrait posts because it takes up more screen space than a square.

Example

An Instagram portrait post at 1080 × 1350 is 4:5.

A landscape ratio standardized for social link previews — wide and short.

Example

LinkedIn and Facebook link preview images use 1200 × 627 (1.91:1).

A tall portrait ratio commonly used for Pinterest pins and portrait photography.

Example

A Pinterest pin at 1000 × 1500 is 2:3.

A near-square ratio used in large-format and medium-format photography.

Example

A 2500 × 2000 print is in 5:4.

Black bars added to the top and bottom of a video when a wider aspect ratio is shown on a narrower screen.

Example

A 21:9 movie on a 16:9 TV shows letterbox bars above and below the picture.

Black bars added to the left and right when a narrower (more square) video is shown on a wider screen.

Example

A 4:3 old-TV clip on a 16:9 screen shows pillarbox bars on the sides.

Compression & Bitrate

The amount of data processed per second of video or audio, measured in kilobits (kbps) or megabits (Mbps). Higher bitrate generally means better quality and a larger file.

Example

A 1080p video at 5 Mbps looks smooth; the same video at 500 kbps turns blocky.

A compression method that discards some data to shrink the file. The loss is often invisible, but repeated re-saving degrades quality ("generation loss").

Example

JPEG photos and MP3 audio are lossy — each re-save slightly reduces quality.

A compression method that reduces file size without removing any data, so the original can be perfectly restored.

Example

PNG and FLAC files compress without losing a single pixel or sample.

Visible glitches caused by aggressive lossy compression — blocky patches, banding in gradients, or blurry edges.

Example

A JPEG saved at quality 10 shows blocky squares and smeared edges.

An encoding mode that keeps the bitrate the same throughout the entire file. Predictable file size, but less efficient for scenes with varying complexity.

Example

A 2 Mbps CBR video uses 2 Mbps even during a static, unchanging scene.

An encoding mode that raises or lowers the bitrate depending on how complex each scene is. Generally better quality at a smaller size than CBR.

Example

VBR lowers bitrate during a still talking-head scene and raises it during an explosion.

A quality-based setting used by encoders like x264 — you pick a target quality and the encoder adjusts bitrate to maintain it. Lower numbers mean higher quality.

Example

CRF 23 is a balanced default for x264; CRF 18 looks near-lossless but makes bigger files.

A sequence of video frames starting with a keyframe. The GOP length controls how often a full frame is saved — longer GOPs compress better but are harder to edit or seek.

Example

A GOP of 60 means one full keyframe is stored every 60 frames.

A block of pixels (often 16 × 16) that video codecs process as a unit. When bitrate is too low, these blocks become visible as squares.

Example

A heavily compressed video shows its macroblocks as a grid of blocky squares.

The step where a codec rounds off fine detail to save space. Higher quantization means more compression and more visible loss.

Example

High quantization flattens subtle skin tones into bands of flat color.

A method where the encoder scans the video twice: first to analyze complexity, then to encode with optimally distributed bitrate. Slower but more efficient than single-pass.

Example

A two-pass encode at 4 Mbps looks better than a single-pass encode at the same bitrate.

A slider or value (usually 1–100) that controls how much an image is compressed. Higher = better quality and larger file; lower = smaller file and more artifacts.

Example

A JPEG at quality 80 looks nearly identical to quality 100 at about half the file size.

The math used to guess new pixel values when resizing. Different methods (nearest, bilinear, bicubic) trade speed for smoothness.

Example

Bicubic interpolation produces smoother edges than nearest-neighbor when upscaling.

File Formats — Image

The most common lossy image format for photographs. Supports millions of colors and achieves very small file sizes. Does not support transparency.

Example

A phone photo exported as a 2 MB .jpg file for uploading to social media.

A lossless image format that supports transparency. Ideal for graphics, logos, and screenshots, but produces larger files than JPEG for photos.

Example

A logo with a transparent background saved as a .png for use on any backdrop.

A modern Google image format offering both lossy and lossless compression plus transparency, often producing much smaller files than JPEG or PNG.

Example

A WebP version of a photo can be about 30% smaller than the same JPEG.

A next-gen image format based on the AV1 video codec. Compresses even better than WebP and supports HDR, but encoding is slower and support is still growing.

Example

Netflix serves AVIF images to compatible browsers to cut bandwidth.

An older image format limited to 256 colors that supports simple animation. Small and universal, but poor for photos. Modern video (MP4) is usually a better choice for motion.

Example

A short looping reaction animation shared on social media.

A high-efficiency image format used by default on modern iPhones. Offers JPEG-quality images at roughly half the file size, but compatibility with some platforms is limited.

Example

An iPhone photo saved as .heic at ~1 MB instead of ~2 MB as JPEG.

A vector image format made of mathematical shapes rather than pixels. Scales to any size without losing sharpness — perfect for logos and icons.

Example

A company logo that stays crisp whether shown at 32 px or 3200 px.

An older, uncompressed Windows bitmap format. Files are large and it offers no compression advantage; rarely used on the web today.

Example

An old Windows desktop wallpaper file that's 6 MB for a small image.

A flexible, high-quality image format popular in professional photography and print publishing. Supports lossless compression and layers, but files are large.

Example

A print-ready magazine photo delivered to a publisher as a 50 MB .tiff.

Unprocessed sensor data straight from a camera, with no in-camera compression or color tweaks. Offers maximum editing flexibility but huge file sizes.

Example

A wedding photographer shoots .cr2 or .nef RAW files for later editing.

Adobe Photoshop's native layered design file. Preserves layers, masks, and editable text — the working source before exporting a flat image.

Example

A poster design kept as a layered .psd so each text layer stays editable.

A legacy vector format used by print shops for logos and illustrations. Largely replaced by PDF and SVG, but still common in commercial printing.

Example

A brand logo sent to a printer as an .eps file for signage.

A Windows icon format that bundles multiple sizes in one file. Best known as the favicon format shown in browser tabs.

Example

A website's favicon.ico appears as the tiny icon in the browser tab.

Animated PNG — supports full-color animation with transparency, unlike GIF's 256-color limit. Larger but far higher quality.

Example

A smooth, full-color animated sticker shared in a chat app.

A texture format used heavily in 3D games and real-time rendering. Supports GPU-friendly compression for fast loading.

Example

A character's skin texture stored as .dds so a game can load it instantly.

File Formats — Video

The most widely supported video format — a container that typically holds H.264 video and AAC audio. Great balance of quality and small file size for web and social media.

Example

A YouTube video downloaded as a .mp4 plays on nearly every device.

An open, Google-developed video format designed for the web. Often produces smaller files than MP4 and is well supported in modern browsers.

Example

A website serves a .webm clip that loads faster than the equivalent MP4.

Apple's QuickTime video container format. Common on macOS and iOS; high quality but larger files, so it's usually converted to MP4 for the web.

Example

A video exported from iMovie as a .mov before being converted for YouTube.

An older Microsoft video container format. Largely replaced by MP4 and WebM; files tend to be large with limited compression.

Example

An old home video saved as a 1 GB .avi file with weak compression.

An open-source video container that can hold many audio, video, and subtitle tracks in one file. Popular for high-definition archival, but not universally supported on all devices.

Example

A 4K movie packaged as .mkv with five language tracks and subtitles.

Apple's professional editing codec — high quality and fast to edit, but produces very large files. Used in post-production, not for final web delivery.

Example

A video editor keeps ProRes .mov files while editing, then exports MP4 for upload.

Adobe Flash video, the dominant web video format in the 2000s. Now obsolete after Flash was discontinued, but you may still encounter old clips.

Example

An early-YouTube clip still sitting on an old server as a .flv file.

Windows Media Video — Microsoft's older compressed video format. Decent compression but limited cross-platform support today.

Example

An old corporate training video delivered as a .wmv file.

A low-quality, low-bandwidth video format designed for early 3G mobile phones. Files are tiny, but quality is poor by modern standards.

Example

A grainy video recorded on a 2007-era feature phone saved as .3gp.

An Apple video container similar to MP4, sometimes with DRM copy protection. Used by iTunes Store purchases.

Example

A movie bought on iTunes downloads as a .m4v file with Apple's DRM.

File Formats — Audio

The most common lossy audio format. Small files, universal compatibility, and good-enough quality for most casual listening.

Example

A song saved as a 5 MB .mp3 that plays on any device.

An uncompressed audio format — CD-quality but large. Standard for professional recording and editing.

Example

A 3-minute song as a WAV is about 30 MB.

A lossless audio format that compresses to roughly half the size of WAV while preserving every original sample. Favored by audiophiles.

Example

A FLAC album is ~150 MB vs. ~300 MB as WAV, with identical audio quality.

A lossy audio codec that generally sounds better than MP3 at the same bitrate. The standard audio track inside most MP4 videos and Apple Music.

Example

A 128 kbps AAC stream sounds as good as a 192 kbps MP3.

An open-source lossy audio format. Efficient compression and royalty-free, used by Spotify and Wikipedia.

Example

Wikipedia's audio samples are stored as .ogg files.

A modern, low-latency audio codec that handles both speech and music well at low bitrates. Used by Discord, WhatsApp calls, and YouTube live.

Example

A Discord voice call at 32 kbps Opus sounds clear with minimal delay.

Codecs & Containers

Software that compresses (encodes) and decompresses (decodes) video or audio. The codec determines how the media is compressed inside a file.

Example

H.264 is the codec that compresses the video inside an MP4 file.

The file "wrapper" that bundles video, audio, subtitles, and metadata together. The codec inside is what actually compresses the media.

Example

An .mp4 container can hold H.264 video plus AAC audio in one file.

The most widely used video codec today. Excellent compatibility across devices and platforms, with a solid balance of quality and file size.

Example

Nearly every phone recording and YouTube video uses H.264.

A newer codec that compresses video about 50% more efficiently than H.264 at the same quality. Smaller files, but licensing and device support can be limited.

Example

A 4K movie in H.265 is roughly half the size of the same movie in H.264.

An open, royalty-free video codec that outperforms H.265 in compression. Increasingly used by YouTube and streaming platforms, though encoding is slower.

Example

Netflix uses AV1 to cut data usage on mobile devices.

Google's open, royalty-free video codec — the predecessor to AV1. Used for much of YouTube's 4K and HDR video.

Example

YouTube serves VP9 for 4K video to compatible browsers.

The older video codec used on DVDs and early digital television. Reliable but inefficient by modern standards.

Example

A Hollywood movie on a standard DVD is encoded in MPEG-2.

Metadata, Color & Motion

Hidden information embedded in photos by cameras — date, camera model, lens, exposure settings, and sometimes GPS location. Often stripped before sharing for privacy.

Example

A photo's EXIF reveals it was taken with an iPhone on July 4 at specific GPS coordinates.

Text metadata attached to images — captions, keywords, photographer credit, copyright. Standard in news and stock photography.

Example

A news wire photo carries its headline and byline in IPTC fields.

Adobe's extensible metadata format, often saved as a sidecar file alongside RAW images. Stores editing instructions and keywords.

Example

Lightroom writes your edits to an .xmp sidecar so they can travel with a RAW file.

A defined range of colors a device can display. sRGB is the web standard; Adobe RGB and Display P3 are wider gamuts used in photography.

Example

A photo in sRGB looks right in every browser; the same photo in Adobe RGB may look washed out on the web.

The total range of colors a display or file can represent. A wider gamut shows more vivid reds, greens, and blues.

Example

A Display P3 screen shows richer greens than a basic sRGB monitor.

A format that preserves brighter highlights and deeper shadows than standard video. Requires a compatible capture, file, and display to see the benefit.

Example

An HDR movie on an HDR TV shows a sun that glows without washing out to white.

Regular video without HDR — a narrower range of brightness. Most existing video and screens are SDR.

Example

A normal YouTube clip on a regular monitor is SDR.

The number of bits used to store each color channel. 8-bit supports ~16.7 million colors; higher bit depths (10-bit, 16-bit) allow smoother gradients.

Example

An 8-bit sky may show visible banding; a 10-bit sky blends smoothly.

A compression trick that reduces color detail more than brightness detail, since the eye is less sensitive to color. Written as ratios like 4:4:4 (none) or 4:2:0 (common in video).

Example

YouTube video uses 4:2:0, halving color detail to save space with little visible loss.

An extra channel alongside color that records how transparent each pixel is. Without it, images can't have see-through areas.

Example

A PNG logo's transparent background is stored in its alpha channel.

See-through areas in an image, letting whatever is behind it show through. Supported by PNG, WebP, GIF, and SVG — but not by JPEG.

Example

A transparent sticker PNG can be dropped onto any colored background.

A graph showing how many pixels exist at each brightness level, from black (left) to white (right). Used to judge exposure.

Example

A histogram piled against the right edge means the photo is overexposed (blown highlights).

The setting that makes white look truly white under different light sources, removing color casts from warm or cool lighting.

Example

A photo under indoor bulbs looks orange; correcting white balance makes it neutral.

How much light reaches the sensor. Underexposed images are too dark; overexposed images are too bright with lost highlight detail.

Example

An underexposed night photo is muddy and dark; bumping exposure brightens it.

The span between the darkest and brightest parts a camera can capture at once. A wide dynamic range preserves both shadows and highlights.

Example

A sunset photo needs high dynamic range to keep both the bright sky and the dark foreground.

Random speckles in an image, usually from shooting in low light at a high ISO. Looks like film grain or static.

Example

A dark indoor phone photo is full of colored noise from a high ISO setting.

An effect that boosts edge contrast to make an image look crisper. Overdone, it creates glowing halos around edges.

Example

Light sharpening makes a slightly soft photo look crisp; heavy sharpening adds visible halos.

A trick that scatters pixels of available colors to simulate shades a format can't actually represent, smoothing gradients.

Example

A GIF with only 256 colors uses dithering to fake a smooth sky gradient.

The curve that maps brightness values. sRGB uses a gamma of roughly 2.2; mismatched gamma makes images look too dark or too bright.

Example

A raw render in linear gamma looks washed out until converted to sRGB's 2.2 gamma.

How many still frames are shown per second of video. 24 fps looks cinematic, 30 fps is standard for web, and 60 fps looks smooth.

Example

A cinematic film uses 24 fps; a sports broadcast uses 60 fps for smooth motion.

A complete, self-contained video frame that other frames reference. More keyframes mean larger files but easier seeking and editing.

Example

Jumping to a point in a video snaps the playhead to the nearest keyframe.

Interlaced video draws every other line, then fills the gaps (old TV method). Progressive draws each frame in full — the modern standard.

Example

1080p is progressive; 1080i is interlaced and can show comb-like artifacts on motion.

Constant Frame Rate keeps FPS steady throughout. Variable Frame Rate changes FPS on the fly — common in screen recordings that idle.

Example

A screen recording at VFR drops to a low FPS when nothing moves, then spikes during active use.

Streaming & Conversion

Converting media from one codec or format to another by re-encoding. Changes quality and file size, and takes time.

Example

Transcoding a .mov into an .mp4 so it plays on the web.

Moving media from one container to another without re-encoding — fast and lossless, but only works if the target supports the same codecs.

Example

Remuxing H.264/AAC from an .mkv into an .mp4 in seconds, with zero quality loss.

Delivering video or audio over the internet so playback starts before the whole file is downloaded.

Example

Watching Netflix without waiting for the entire movie to download first.

HTTP Live Streaming — Apple's protocol that splits video into small chunks for adaptive delivery. The standard for live and on-demand streaming.

Example

A live sports stream on an iPhone uses HLS to switch quality mid-playback.

Dynamic Adaptive Streaming over HTTP — an open standard, similar to HLS, that lets players switch quality as bandwidth changes.

Example

YouTube uses DASH to drop you to 480p on a weak connection, then back up to 1080p.

A streaming technique that keeps several quality levels of the same video and switches between them live as the viewer's connection changes.

Example

On a slow train, a video drops to 360p; on Wi-Fi, it bumps back to 1080p automatically.

Pre-loading a few seconds of video before and during playback so it doesn't stutter. When playback pauses to load, that's buffering.

Example

A video that freezes with a spinning loader is buffering — waiting for more data.

The delay between when something is recorded and when the viewer sees it. Live streaming adds a few seconds of latency for processing.

Example

A live streamer's words reach viewers about 5–10 seconds after they're spoken.

Splitting a video into many small segments (chunks) that are fetched one at a time, enabling adaptive streaming and live delivery.

Example

HLS breaks a video into 6-second .ts chunks that the player fetches as it plays.

Ready to compress?

Put these terms to work — shrink and format your media in seconds, right in your browser.

Open the compressor