The History of Raster Image Formats

Here’s a weird one: the word has nothing to do with computers originally. It’s from the Latin rastrum, a rake. Scrape, drag, rake. That’s it.

Turns out old CRT monitors worked almost exactly like that. An electron beam dragged itself across the screen one line at a time, lighting up pixels as it went, the same motion as a rake pulled through dirt. Someone noticed the resemblance and the name stuck.

A raster image, in plain terms, is just a grid of pixels. Nothing fancier than that. The whole idea traces back to Bell Labs in the late 1960s, where a researcher named A. Michael Noll built some of the earliest scanned raster displays. Every format below owes something to that work, even the ones invented decades later.

JPEG 2000: the upgrade nobody asked loudly enough for

“JPEG” stands for Joint Photographic Experts Group. It is the name of the committee which initiated the JPEG standard and also other still picture coding standards. The “Joint” stands for ISO TC97 WG8 and CCITT SGVIII. In 1987 ISO TC 97 became ISO/IEC JTC1 and in 1992 CCITT became ITU-T. Currently on the JTC1 side JPEG is one of two sub-groups of ISO/IEC Joint Technical Committee 1, Subcommittee 29, Working Group 1 (ISO/IEC JTC 1/SC 29/WG 1) – titled as Coding of still pictures. On the ITU-T side ITU-T SG16 is the respective body. The original JPEG Group was organized in 1986,[9] issuing the first JPEG standard in 1992, which was approved in September 1992 as ITU-T Recommendation T.81 and in 1994 as ISO/IEC 10918-1.

JPEG stands for Joint Photographic Experts Group, which, honestly, is a pretty dull name for something this influential. It’s just the committee that built the format, formed in 1986 by two older standards bodies, ISO and CCITT (the latter now called ITU-T). They shipped the first JPEG standard in 1992.

Here’s the thing about JPEG though. It was never the most technically impressive format out there. It won because it hit a sweet spot: good enough quality, small enough files, right at the moment the internet needed exactly that. Timing mattered more than perfection.

More than thirty years later, despite plenty of better alternatives now existing, JPEG is still probably the most common photo format on earth. Old habits die hard, especially when half the web’s infrastructure is still built around them.

On paper, JPEG 2000 should have replaced JPEG. It compresses a little better, and its real trick is flexibility. You can cut the compressed data at different points and pull out a lower resolution or lower quality version on the fly. Hospitals use this. So does digital cinema, where JPEG 2000 quietly survives to this day.

But flexibility costs something. Encoding and decoding takes more horsepower than plain JPEG, and the visual mistakes look different too. Regular JPEG gets blocky because of how it slices images into little 8×8 squares. JPEG 2000 doesn’t do that, but it leaves a soft blur or ringing around sharp edges instead.

It became an official ISO standard in 2000. And then, not much happened. Web browsers never bothered adding support, and as of 2026, that’s still true. Not one mainstream browser handles it natively. Outside specialist corners like medical imaging, most people have simply never heard of it.

GIF: built for an internet that moved painfully slow

CompuServe launched GIF on June 15, 1987. The goal was simple: give people a decent color image format for downloading files, replacing an old black-and-white format that just wasn’t cutting it anymore.

What made GIF actually take off was LZW compression. It squeezed images far more efficiently than the run-length methods older formats like PCX and MacPaint relied on. When you’re downloading over a modem slower than your current Wi-Fi router, that difference is the whole ballgame.

The original 1987 version gets called 87a. Two years later came 89a, adding animation delays, transparent backgrounds, and room for extra metadata. You can actually tell them apart just by peeking at the first six bytes of a file. They literally spell out GIF87a or GIF89a.

GIF also had interlacing, a neat trick where the image loads blurry first and sharpens as more data comes in. On a slow connection, that let you bail out early if the picture clearly wasn’t what you wanted. Combined with small file sizes, this made GIF one of the first two image formats early websites actually used, the other being a far less memorable black-and-white format called XBM.

Netscape Navigator 2.0 added looping for animated GIFs in 1995, which is basically where “GIF culture” as we know it began. Funny enough, it took ages to spread everywhere. Facebook didn’t even support GIFs until May 2015, years after sites like Tumblr had already made them a staple.

PNG: the format a lawsuit basically created

PNG exists because of a legal mess. In early 1995, people realized the LZW compression inside GIF was patented by Unisys, meaning any software that created GIFs technically owed royalties. On top of that, GIF was capped at 256 colors right as computers capable of showing millions of colors were becoming normal. Something had to give.

The idea kicked off on a Usenet thread titled “Thoughts on a GIF-replacement file format” in January 1995. One contributor, Oliver Fromme, who had written a popular DOS JPEG viewer, half-jokingly proposed the name PING, short for “PING is not GIF.” That eventually got trimmed down to PNG.

The team behind it made one deliberate choice early: no animation. PNG would be single images only. That gap got patched later by other people. In 2001 came Multiple-image Network Graphics, or MNG, which never really caught on. Then in 2008, some Mozilla developers released Animated PNG, or APNG, and that one stuck. By 2026, every major browser supports APNG, including Edge, which picked it up once Microsoft switched the browser over to Chromium.

EXIF: the data riding along inside every photo you take

You’ve probably never opened an EXIF file on purpose in your life. Doesn’t matter. Nearly every photo you’ve ever taken carries one anyway. The Japan Electronic Industries Development Association (JEIDA) built the original spec, with version 2.1 landing in June 1998. The organization later renamed itself JEITA and kept the updates coming: version 2.2 in 2002, nicknamed “Exif Print,” version 2.21 in 2003 with Adobe RGB support, and version 2.3 in 2010, revised again in 2013.

This is the data working quietly behind the scenes. Camera model, shutter speed, GPS coordinates, timestamp. Nearly every camera and phone manufacturer supports it, which is why that information just appears automatically when you check a photo’s details on pretty much any device.

Need any of your own images converted to a newer format like WebP or AVIF?  Raster Conversion service handles that in a few clicks.

WebP: Google’s first real shot at dethroning JPEG

Google rolled out WebP on September 30, 2010, built on technology it picked up through buying On2 Technologies. It handles both lossy and lossless compression, and it’s technically a sibling of the WebM video format since both lean on the VP8 video codec underneath.

A year later, in October 2011, Google bolted on animation support, ICC color profiles, XMP metadata, and tiling for huge images. The pitch was simple: smaller files than JPEG, roughly the same visual quality.

It took a while to spread, as these things always do, but by 2020 WebP had full support across every major browser. These days it mostly plays backup. AVIF takes the lead spot on most sites now, with WebP stepping in only for the rare browser that hasn’t caught up to AVIF yet.

AVIF: the one that actually won, quietly

This one wasn’t part of the original story, but leaving it out in 2026 would be a mistake. AVIF is built on the AV1 video codec, an open, royalty-free standard backed by a group that includes Google, Netflix, and Mozilla. It compresses noticeably better than both JPEG and WebP at matching quality.

Chrome picked it up in 2020, Firefox in 2021, Safari in 2023. Slow rollout, sure, but it’s basically over now. Global support sits around 93 to 95 percent, high enough that most major sites just default to AVIF and only fall back to WebP or JPEG for the small sliver of older browsers still hanging around.

The one real downside is encoding speed. AVIF files take noticeably longer to produce than WebP or JPEG XL at similar quality, which matters if you’re churning through huge batches of images. For a finished photo already sitting on a web page though, that cost is completely invisible to whoever’s looking at it.

BPG: one developer, one ambitious attempt

Better Portable Graphics, or BPG, came from a single person: Fabrice Bellard, the same programmer behind FFmpeg and QEMU. He built BPG in 2014 on top of the intra-frame encoding used in HEVC video compression, and tests that year showed it beating JPEG, JPEG XR, and WebP on file size at equal quality.

It had everything going for it on paper. Small files, strong quality, light enough to run on handheld devices and early IoT gadgets. But no major browser ever gave it native support. Bellard released a JavaScript decoder so sites could display BPG images anyway, but a JavaScript workaround was never going to beat formats baked directly into the browser. BPG ended up more of a curiosity than an adopted standard.

HEIF: the format already on your iPhone

High Efficiency Image File Format, or HEIF, came out of the Moving Picture Experts Group and is defined under MPEG-H Part 12. The group behind it claimed it could pack roughly twice the image information into the same file size as JPEG, at noticeably better quality. Requirements were drawn up in 2013, and the spec itself was finalized around 2015.

HEIF’s real break came when Apple made it the default photo format on iPhones starting with iOS 11 in 2017, using the .heic extension. One decision, hundreds of millions of devices overnight, and most people never even noticed the format behind their photos had changed. It’s still the default today. Most apps just quietly convert HEIC to JPEG whenever you share a photo somewhere that doesn’t know what to do with it.

TIFF: the format scanners needed

TIFF solved a very boring, very real problem. Back in the mid-1980s, every desktop scanner maker had its own proprietary image format, and none of them talked to each other. Aldus Corporation, the company behind PageMaker, published the first real TIFF spec, Revision 3.0, in autumn 1986, after working it out with scanner manufacturers and software developers.

Early TIFF could only handle pure black-and-white images, because that’s all scanners of the time could produce. As scanners got better and hard drives got cheaper, TIFF grew to handle grayscale, then full color. Revision 4.0 shipped in April 1987 with mostly small fixes, and Revision 5.0 in October 1988 added palette color images along with LZW compression.

TIFF never made it onto the web in any meaningful way, but it’s still the go-to choice in professional photography, printing, and archiving today. Lossless compression and high image quality will do that for a format.

PCX: the format that got DOS users through the ’80s

PCX stands for PiCture eXchange, made by ZSoft Corporation out of Marietta, Georgia. It started out as the native format for PC Paintbrush in 1985 and became one of the first image formats DOS users widely recognized by name.

Most PCX files stored palette-indexed images, anywhere from 2 or 4 colors up to the full 256, though the format later stretched to cover true 24-bit color too. It had a good run through the late ’80s and early ’90s before formats like BMP, JPEG, and PNG showed up and mostly pushed it out of the picture.

JPEG XL: the comeback that’s still, technically, in progress

JPEG XL didn’t exist when this history first got written either, and its story is a little chaotic. The same JPEG committee built it, working alongside Google and Cloudinary, and it became an official ISO standard, ISO/IEC 18181, in 2022.

On paper, it beats nearly everything. Files 50 to 60 percent smaller than JPEG. Better compression than AVIF in most cases. True progressive loading, the kind old JPEGs used to have. It can even repackage an existing JPEG into a smaller JPEG XL file without losing a single pixel of data.

Browser support is where things get messy. Google actually ripped JPEG XL support out of Chrome back in 2022, which killed a lot of its early momentum. Then, in a surprising about-face, Chrome 145 brought it back in February 2026, and Firefox 152 followed in June 2026. Both ship it switched off by default though, buried behind an experimental flag most people will never find. Safari’s the only browser that’s supported it natively out of the box, and has since 2023.

So where does that leave it? As of late 2026, JPEG XL support sits around 12 to 15 percent of the web, almost entirely from Safari users. A genuinely excellent format that still can’t be anyone’s primary choice. For now it mostly rides along as a bonus option behind AVIF and WebP, for the browsers that happen to support it.

Closing Thought

That’s the short version: from electron beams raking across a CRT screen to arguing over AVIF versus JPEG XL in 2026. The problem never really changes. People just want smaller files, better quality, and faster loading, all at once, forever. If there’s a format we missed, or something here looks off, drop a comment and we’ll fix it.