2254: JPEG2000

Explain xkcd: It's 'cause you're dumb.
Revision as of 03:58, 8 February 2024 by (talk) (Explanation)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
I was actually a little relieved when I learned that JPEG2000 was used in the DCI digital cinema standard. I was feeling so bad for it!
Title text: I was actually a little relieved when I learned that JPEG2000 was used in the DCI digital cinema standard. I was feeling so bad for it!


JPEG2000 is a standard for digital image storage created by the Joint Photographic Experts Group from 1997 to 2000 to improve on the original JPEG standard, published in 1992. The original JPEG standard is the most widely used image format in the world for both digital cameras and the World Wide Web, while the newer and improved JPEG2000 standard is relatively rare. As of 2020, it is supported by Photoshop, the Safari browser, and GIMP, but it remains unsupported or poorly supported by other popular software, including Google Chrome and Mozilla Firefox browsers. Meanwhile, competing format WebP which appeared 10 years later is supported in all major browsers and has much wider support in other applications as well. As a result, the conventional file name extensions for files using the JPEG2000 standard, .jp2 and .jpx, remain unfamiliar to many users while the .jpg extension, denoting the original standard, is well known. The JPEG2000 standard was seen as an improvement by its creators, supporting many features not included in the original standard, such as multiple resolutions, progressive transmission, a lossless compression option, and alpha channel transparency. The complexity of fully implementing the standard, as well as patent concerns, may have slowed adoption.

Cueball and Hairbun seem to have some desire for or stake in JPEG2000 adoption. Cueball begins to worry after more than 20 years without much progress but Hairbun is confident that it will eventually prevail, and she cares more about its eventual use than rapid adoption.

The core concept of this comic is that engineers often expect that a superior technology or standard will catch on, though often other factors keep an "inferior" standard dominant. (See various comics referencing Dvorak keyboards, as well as the term "betamaxed.") The "we are in this for the long haul" statement might refer to the engineers believing that superior technology will eventually win despite the evidence to the contrary. Its humor comes from the fact that as of the comic publication in 2020, JPEG2000 shows no sign of becoming a widely-used standard.

The title text suggests that Randall feels bad that the standard hasn't been adopted, perhaps because he empathizes with the engineers who worked hard to develop it or anthropomorphizes the standard itself, which has been ignored by most of the computer-using world. Also he may actually believe it is the better standard that should have been more widely used. DCI, short for Digital Cinema Initiatives, is a collaboration of several major film studios to establish standards for the security and proper display of digital films. Version 1.0 of the DCI’s “Digital Cinema System Specification” was released in 2005.


[Cueball and Hairbun are sitting on office chairs opposite each other on a shared desktop with a small division wall between them. They are both working on their own respective computers.]
[Cueball leans back and stops typing. Hairbun continues to type.]
[Cueball makes a statement as he looks over at Hairbun who looks back at him when she replies. She has moved one hand off the keyboard down to her lap. Cueball's keyboard has disappeared!]
Cueball: I'm starting to worry that JPEG 2000 isn't catching on as fast as we expected.
Hairbun: Don't worry! We're in this for the long haul.


  • Cueball's keyboard seems to have inexplicably disappeared in the last panel.
  • A JPEG2000 version of the image file is available here: https://imgs.xkcd.com/comics/jpeg2000.jp2 . It is only 20% smaller than the PNG version, and has visible compression artifacts.
  • JPEG2000 is also supported as an image compression method in PDF.

comment.png add a comment! ⋅ comment.png add a topic (use sparingly)! ⋅ Icons-mini-action refresh blue.gif refresh comments!


Cubeball's keyboard has disappeared in the third panel. 22:24, 13 January 2020 (UTC)

Pretty sure the woman in this comic should be called Hairbun. Updating transcript... Ianrbibtitlht (talk) 01:31, 14 January 2020 (UTC)

Jpeg2000 is widely used on archive.org (scans are stored as .jp2 there). For example, the image of this page [1] is internally from a jp2.zip file: https://ia601604.us.archive.org/BookReader/BookReaderImages.php?zip=/16/items/in.ernet.dli.2015.215730/2015.215730.The-Theory_jp2.zip&file=2015.215730.The-Theory_jp2/2015.215730.The-Theory_0099.jp2 where BookReaderImages.php seems to be able to read .jp2 in zip and send it to you as a legacy format your browser can handle. Yosei (talk) 01:48, 14 January 2020 (UTC)

I wonder if as a result of this comic, xkcd fans will cause rapid adoption. Mikemk (talk) 05:11, 14 January 2020 (UTC)

Looks like it just isn't worth it.

GIMP seems to be able to load JPEG2000 images. To export as JPEG2000, you need an external plugin. Fabben (talk) 12:02, 14 January 2020 (UTC)

That’s correct, I changed the text. --17:06, 14 January 2020 (UTC)

I wonder if Randall is deliberately referencing Valve's Artifact's long haul. Even has a loose connection with image artifacts. 12:20, 14 January 2020 (UTC)

Would a brief description of the .png format (more typically used for comic images) be appropriate? -- (talk) (please sign your comments with ~~~~)

I was pretty sure that patents were the main problem with adoption, at least in time when .gif patents were a problem. However, it seems the patent status is getting better and it isn't helping ... meanwhile, WEBP, which is using similar technology, is gaining traction. ... which would also answer the question of the previous commenter: while brief mention of PNG might be worth it, mention of WEBP and similar alternatives would be more important -- Hkmaly (talk) 23:46, 14 January 2020 (UTC)

In the US, patents are issued for 20 years, counting from when the application was submitted, which means it should be coming off patent any day now. JamesCurran (talk) 21:44, 30 January 2020 (UTC)

It's also used for textures in Second Life. In fact, that page also states that decompressing JPEG2000 is much more processor-intensive than other image compression methods, so I guess that might be another reason for the lack of general adoption? EddyM (talk) 00:50, 15 January 2020 (UTC)

JPEG2000 is not at all unknown in the geospatial community. Both USGS and NASA offer various aerial and satellite imagery products in JPEG2000 format only. I assume it is one of the most versatile non-proprietary photographic imaging formats out there. 06:30, 15 January 2020 (UTC)

When I clicked on th3 .jp2 as ljnked in the Trivia, my tablet wanted to open it only in my (pre-installed bog-standard) ebook reader or GPS Essentials (perhaps confirming's comment, just above). But mention of JPEG2000 takes me back (25 years or so!) to a time a similar scare to the GIF patent issue had motivated alternatives to the 'public' common picture standard. And reminds me also of the "masking" technique used on (regular?) JPEGs, based upon keyword-hash shuffling/deshuffling of selected 8x8-pixel DCT units of a JPEG image (and of the hues apllied to the curves) to reversibly censor images, IIRC driven largely by Japanese censorship rules. Somewhere on an old hard disk I must still have the reverse-engineered 'solver' I wrote for that, written in Delphi... ;) 18:14, 15 January 2020 (UTC)

Odd that Randall would use a lossy JPEG2000 image for a cartoon rather than a lossless one. A friendly reminder that JPEG is best for photography and is not intended for line drawings. Thisisnotatest (talk) 08:04, 16 January 2020 (UTC)

That's his loss, then! 17:15, 16 January 2020 (UTC)

I was browsing a series of 70+ page PDFs that was a very high quality image scan, and the PDF browser would regularly grind to a halt for a second or two when trying to move forward a few pages. I eventually discovered that the images were embedded in JPEG2000. They were definitely small file sizes and definitely high quality, but it was just too much. I decoded the entire 500+ pages and re-encoded them as jpeg. Bigger file size, lower resolution, but scrolling was smooth as butter again.

Randall is correct to not really care about the standard's failure, per se, except insofar as he feels sorry for it. The difference between the technical impressiveness of these improvements and their unimportance to reality reminds me of the VHS vs Beta issue. Yes, Beta had the ability to reproduce sound and video of a higher fidelity, but only in a trivial sense indistinguishable to most people under normal conditions, whereas VHS was better at things that were indeed important, like being able to record a full two hour movie when Beta could handle less than one hour. The same thing happened with OS/2 vs Windows...OS/2 was purely object-oriented, a technical distinction that was completely irrelevant to real life, but required four times as much RAM as the typical brand-new computer came with, so it failed. Being able to save 32 bit color profiles and choose whether the compression is lossless is important to me as a graphic artist, but doesn't matter one whit to the typical user, who wouldn't even notice the difference. —Kazvorpal (talk) 17:55, 25 January 2020 (UTC)

The link to the term "betamaxed" doesn't seem to be working... (talk) 03:32, 8 February 2024 (please sign your comments with ~~~~)

Well, it worked, but was redirecting to a generic page lacking the definition. ;)
Anyway, replaced it with another link that (currently) defines the term. 04:01, 8 February 2024 (UTC)