I still use dithering daily with 24 bit color images. In fact, my test for whether a modern image editor is serious and good is whether it supports dithering when converting from floating point color channels down to 8 bits per channel. Photoshop does this.
The most important reason for me is not display on the monitor, but printing the image. I do a lot of large format printing, and printers smash parts of your color space and make some gradient/banding problems stick out like a sore thumb. Dithering is critical when printing!
Gradients with banding can easily show up when you resize a large image down to a smaller size. This is a good reason that resizing probably anything less than 16 bits per channel images should be done in a higher precision format than the image.
I haven't tried to use an error diffusion dither when converting 16 bits per channel down to 8... I'm not sure how much that matters. It might make a difference, and it sounds fun to code either way, but in my experience, a good random number generator suffices to make color bands vanish.
> It also has uses when reducing 48 or 64bpp RAW-format digital photos to 24bpp RGB for editing.
The author touches upon it here, but I think it's worth generalizing further: If you have high or maybe infinite precision in your color values, dithering will look much nicer than simply rounding to the nearest value. A concrete example is a color gradient. If done naively with rounding, color bands will be clearly visible. With dithering, they will be almost impossible to see.
The article doesn't mention the void and cluster[0][1] ordered dithering (and modernized variants) which have the advantage of ordered dithering that they are highly parallelization, e.g. via GPU shaders but do not leave the easily spottable patterns of ordered dithering.
For example the madVR[2] video renderer uses it for realtime video dithering to avoid banding shallow color gradients from 10bit sources or debanded internal 16bit representation on <= 8bit displays.
I find it amusing that someone would have a dedicated GPU, yet cannot fully render color images. I mean, I'm sure there's actually plenty of use-cases, but it's still funny.
"Fully render" is a strange way of putting it - most consumer hardware has only 8-bit colour depth, and lots of video compression algorithms will turn smooth gradients into blocky messes that would benefit from dithering.
If you're processing the decode at 10 or 16 bits and need to render at 8 bits, it's much better to dither than truncate.
> and lots of video compression algorithms will turn smooth gradients into blocky messes that would benefit from dithering.
More accurately: Either you have a 10bit-per-channel video (or better) which has smooth gradients which will need dithering when rendering to an 8bit-per-channel display.
Or you have 8bit video (more common) which will have to be passed through a debanding filter first. But then how do you get the debanded result onto the screen without reintroducing the banding? By dithering it.
One technique I've experimented with is iteratively diffusing errors not just towards un-visited neighbors below and to the right, but also into the future. We are all interested in the future, for that is where you and I are going to spend the rest of our lives. [1]
It works nicely with video, and it makes still images (and images you're slowly panning and zooming with the Ken Burns effect [2]) look alive and detailed, as if they were live video.
Another variation of the serpentine scanning (aka boustrophedon transform [3]) you can apply when iteratively dithering an image is to rotate the scan direction 90 degrees each frame, so even frames scan horizontally, odd frames scan vertically, and the vertical and horizontal scan directions rotate round every four frames. That results in a completely uniform diffusion, even when each scan frame only diffuses errors to the next cell.
It spreads the error out over time, as well as space, which has a pleasing effect on the eyes, I think. Any one frame has artifacts, but they tend to cancel each other out over time, break up the log jams, and dance around local minima and without getting stuck at fixed points.
I've implemented some 8 bit anisotropic heat diffusion cellular automata, that exhibited a subtle drift because of the scan order. But rotating the scan order 90 degrees each frame completely eliminated the subtle drifting effect I was getting when using a fixed scan order. Here's a discussion about it I had with Rudy Rucker [4] who inspired some of the rules and pointed out the problem, and a demo [5], and source [6].
Here's one of my favorite spooky Heizenbugs:
>The original version of this code written in C running on a Sun did have an interesting bug: I was not initializing the "error" accumulator that carried the leftover of the average from cell to cell, so when different kinds of background activities were happening on the Sun, the error accumulator got initialized from the stack frame with a random undefined value! I noticed it when every time I typed to a terminal window, the dithering shivered! It was really spooky until I figured out what was going on!
For an interesting case of modern artistic use of dithering, see the discussion related to development of indie game "Return of the Obra Dinn" [1] (by Lucas Pope, author of "Papers, please"), where a participant contributed a dithering scheme he invented [2] for a specific purpose of making faces of in-game characters better looking and easier to recognize.
(Just in case and for easier browsing, I took liberty of uploading a copy to github: https://github.com/akavel/WernessDithering, although I'm not clear on what's the license of the code, unfortunately; that said I hope the numbers in the matrix are not patented.)
(edit: lol, didn't notice there's another thread on Obra Dinn already on HN, now I'm surprised! :)
I don't know if it's some kind of nostalgia factor (since a lot of old 16-color EGA software used it, I think), but I find ordered dithering output strangely appealing. I have also used it sometimes where size mattered more than style on the assumption that GIF/PNG compression rates would be better on ordered dithers. ImageMagick can do a lot of dithering styles and it's fun to play with.
Another fun fact I always think of when dithering comes up is how Tim Sweeney's software rendered or the original Unreal game used dithering instead of bilinear interpolation for texture mapping. This was very impressive at the time.
This article is really nice. I used it as my starting point when I once won a code golf competition on image quality in dithering. Using Fortran.
The algorithm is based on Sierra Lite, but I added a random element to the direction in which the error is propagated. This removes essentially all dithering artifacts.
Has anyone tried using the equivalent of animated GIFs to help with dithering? If you have a limited palette of colours, perhaps you could produce two dithered versions, with pixels sometimes having different colours in alternate frames. If the image is refreshed fast enough, two colours could blend into a third.
The most extreme example could be 'dithering' a grey square into 1) a black square and 2) a white square, and when they are rapidly switching between the two it might appear as a grey instead.
I guess that in practice, the refresh rates of monitors are too low to make it seem anything other than a terrible flickering image, but on old CRTs the effect might work a little better. It's also memory and CPU intensive, but I'd still be curious to see if it could be used successfully, and if it improved the quality compared to 'just' a single dithered image.
You don't even need the animation speed for this. In the GIF format you can have a new color palette for every frame so paint the first 256 colors in the first frame. Add 256 more colors in the right spots in the second frame etc.
This was especially common on 8-bit and 16-bit computers to fake 'high colour' displays on machines that had only limited colour capability. They often took advantage of the fact that people were using their machines with TVs with relatively slow-changing phoshor screens, so the flicker was less evident than on a modern monitor.
The most extreme version I've seen of this was on the ZX Spectrum, which had not only a very limited 15 colour palette, but also limited to 2 colours within each 8x8 block of the screen. Some bright spark came up with the idea of flipping rapidly between R, G, and B frames to give (limited) per-pixel RGB. Unfortunately it did flicker quite badly because of the extreme changes in colour levels (only two levels of each channel), and the fact that it required 3 whole frames to make a single colour virtual frame.
Wow, that example parrot image is pretty impressive, given that it is just three colours, and I'm viewing it on an LCD!
I never knew anyone had tried that before on a spectrum. At first I thought you were just talking about the other trick, getting more than 2 colours per 8x8 by changing the palette as the raster scanned down the screen. The multi-colour parrot is way more adventurous!
I'd love to see the parrot image on an old CRT to get a feeling for what the effect might look like with the phosphor afterglow. Leaving the spectrum behind and using a bigger palette range, like on the ST, the effect seems much less epileptic fit inducing, because you can pick closer colours to switch between.
> The most extreme example could be 'dithering' a grey square into 1) a black square and 2) a white square, and when they are rapidly switching between the two it might appear as a grey instead.
I remember seeing old STN grayscale LCDs that seemed to work like this (and I even implemented it myself on the TI-83 as a test)
There is another interesting application for dithering that I've read about in the recent Uncharted 4 Brain Dump (Ctrl-F for "Dithering") here https://redd.it/4itbxq: Use dithering instead of alpha blending to fade out close objects. Alpha blending can be quite expensive while dithering just omits pixels. The result looks like this: http://allenchou.net/wp-content/uploads/2016/05/dithering-1-... (best visible at the top left corner).
> For simplicity of computation, all standard dithering formulas push the error forward, never backward. If you loop through an image one pixel at a time, starting at the top-left and moving right, you never want to push errors backward (e.g. left and/or up).
Would the image look a lot different if you dithered it backwards from the bottom right pixel?
Are there dithering algorithms that consider the error in all directions instead of pushing the errors forward only?
The answer seems to be that changing the image parsing direction gets rid of some artifacts but introduces others while not vastly improving on faster and simpler approaches.
Also if you don't know what your target palette is you can do k-means clustering over colorspace, and the palette is the cluster centers (each point you're giving to k-means is an <r,g,b> or <h,s,v> vector).
Back when I was at university, we had a computer graphics course. Each week we got new assignment that we needed to program, write to a floppy disk (it was before Internet became widely available) and give it to the prof next week.
Dithering was one of the assignments. We were required to implement black-white quantization and then Atkinson and Floyd-Steinberg. We were given the freedom to choose our own images.
During development at the dorm my favourite picture to debug on was pretty racy (think along the lines of full version of "Lena"). I totally did not intend to put it to the floppy disk...
Not only I got the 10 - the highest number of points for this assignment, I got +2 on top of that with the comment from prof: "for the choice of test images in the best tradition of the field".
I am curious if you knew about the Lena photo and history. It seems likely this remark by the professor was inappropriate, but given that context, provided by joshvm https://news.ycombinator.com/item?id=11888153, I think that muddies the water.
I understand your sentiment. I did not mean to glorify objectification of women with this story. Just something mildly funny that happened almost two decades ago, in a country without a strong tradition of feminism, to a 17-year old me who did not know any better then.
How can an virtual object (a photo of a woman seen on a monitor) be objectified? The model meant that photo to be used as an object (looked at). If she wanted people to understand her personality more, she would have just written a book.
* In geek contexts, they are usually a way for heterosexual men to bond over their common attraction to women. This is othering for anyone who is not a heterosexual man, including, obviously, women, and contributes to their invisibility in the field. This sensation of exclusion is very visceral when in a small minority, as women can be in geek settings.
* There is a long tradition of sexual images, suggestions and approaches being used to shame, scare, harrass or brutalise women. This is common enough that most women will have had personal experience of it. Therefore many women are unable to sensibly assume good faith on the part of unknown men seeking to make a situation sexual and feel mentally uncomfortable at best and physically intimidated often.
* While in many areas this restriction is loosening, women are stigmatised as well as celebrated for being too sexual. This traps women into a double bind when responding to sexualized environments, because even by getting the joke they may reveal themselves as too sexual.
Keep in mind the commenter received extra points for this.
This is othering for anyone who is not a heterosexual man
I disagree. Women look at other women as well. It's very natural to feel excluded when groups of people over something you don't get excited about, whatever the subject. If, for example, people can contain their excitement about Magic cards when I'm around, I'm fine.
There is a long tradition of sexual images, suggestions and approaches being used to shame, scare, harrass or brutalise women.
Yeah, so a picture of a professional model is something completely different. You are dragging this point into the discussion but it doesn't add anything.
...they may reveal themselves as too sexual.
Just bring it back to basics. Men have dicks they want to use, women want to be admired. If people could just respect eachother while taking these basic needs into account, everyone would be fine.
--
I am happy that people here in the Netherlands are so much more tolerant when it comes down to things like sexuality, public intimicy and nudity. It seems to me that in the US people get offended very easily and want to limit another persons freedom of expression just so that they can be offended a little bit less.
Forget abstractions for a second; there are people whom this makes uncomfortable (we know because they said so), and the cost of that is much higher than the very small benefit of using one image over another as a test image. Simple cost/benefit analysis says to use something else.
Obviously in certain circumstances only - few of us here would censor Charlie Hebdo.
We only care about discomfort that we think is reasonable, or more accurately, not a lie designed to control the actions of others.
In this case, most people agree that sexualizing the workplace is a strong negative and are in agreement with you. But as a general rule, you'd have much less support.
I totally did not intend to put it to the floppy disk...
So he didn't mean to turn it in. But what he does in his own time, isn't that his business?
Your analysis also tells us that you should not look at porn in your own free time, as the potential cost is much higher because it makes 2 parties uncomfortable, if caught.
Bear in mind that computer vision has used a cropped image of Lena Söderberg (i.e. the infamous Lena image) for decades now. If you believe the anecdote, the image was not chosen as an in-joke or as a way of sexualising women. The researchers needed a high quality image of a human face (hence they cropped out her body) and they were tired of stock images. Someone in the lab had a playboy magazine and history was made. Most people have no idea that the full version of the image even exists.
The image is so famous that Lena herself was invited as a guest to ICIP 2015 and chaired the best paper award. I think she also gave a talk. I dare say she doesn't have a problem with it.
> This is othering for anyone who is not a heterosexual man
Except, you know, for bisexual men and women, lesbians, heterosexual trans-men, and so on. A lot of people could theoretically bond with one-another over a shared attraction to a given gender. (Not making any point to the spirit of your argument; just poking at the letter of it in hope that you refine it.)
The point you were making wasn't about what was or wasn't happening here, though; it was about what such statements "usually [serve to do] ... in geek contexts." Geek contexts involve a lot of the types of people I mentioned. (For example, according to a recent survey of LessWrong members, there are actually more transwomen than ciswomen in that community.)
> In geek contexts, they are usually a way for heterosexual men to bond over their common attraction to women. This is othering for anyone who is not a heterosexual man, including, obviously, women, and contributes to their invisibility in the field.
There's nothing here suggesting that that shared sexuality can't be used as a bonding mechanism among other groups. Just that in "geek contexts" it's usually intended as a bonding mechanism between straight men, and that this is othering to people who aren't straight men.
-3DFX Voodoo (all models in 16bit color depth) let you enable hardware dithering block (2x2/4x4 ordered dither, zero performance penalty). They did it to save framebuffer space (24bit textures, 16bit framebuffer). It made 3dfx graphics look significantly better than nvidia/ati in 16bit depth. Earlier cards used 4x1 filter on the output, Banshee and later models gained 2x2 filter providing "22 bit like quality" as 3dfx called it.
-All crappy (TN) LCD panels (staple of garbage bin supermarket 1366x768 laptop, and older 'gaming' fullhd ones) use FRC which is a form of temporal dithering https://en.wikipedia.org/wiki/Frame_rate_control
One legitimate use of dithering was in the days of CGA/EGA/other fixed palette hardware. You can relive it here 'Joel Yliluoma's arbitrary-palette positional dithering algorithm': http://bisqwit.iki.fi/story/howto/dither/jy/
The most important reason for me is not display on the monitor, but printing the image. I do a lot of large format printing, and printers smash parts of your color space and make some gradient/banding problems stick out like a sore thumb. Dithering is critical when printing!
Gradients with banding can easily show up when you resize a large image down to a smaller size. This is a good reason that resizing probably anything less than 16 bits per channel images should be done in a higher precision format than the image.
I haven't tried to use an error diffusion dither when converting 16 bits per channel down to 8... I'm not sure how much that matters. It might make a difference, and it sounds fun to code either way, but in my experience, a good random number generator suffices to make color bands vanish.