Gradient Dithering

I realise Illustrator is a vector drawing program, but a lot of us use it to build dynamic user interfaces to output bitmap graphics. There are tools to export bitmap images (save for web and export as PNG) to this should also be optimised for it. One area that illustrator is really lacking in outputting good quality bitmap graphics is the lack of gradient dithering that can be found in photoshop.
Marc Edwards has a great writeup on this.
http://bjango.com/articles/illustratorandappdesign/
I'm sure it could be implemented as an Appearance Effect since dithering occurs at a pixel level.

They don't list Layer style gradient dither in Photoshop's feature list, but it is in CS6. I don't have acces to Illustrator CS6 and would like to know if these features/issues have been addressed. Have searched high and low on the net to see if anyone has mentioned any of these things and have found nothing, hence me asking on the forum.
I wouldn't expect Adobe to put 'Now with better circles!' on the feature list.

Similar Messages

Quality of images / video on new MBP

Have recently bought a new MBP, and have been noticing that with some images and video, the quality is a bit off.
I just tried taking a screen grab from the old powerbook, and one from the new MBP to show the difference, but they both look wrong on the MBP, and both look OK on the Powerbook :
http://www.handprintdvd.co.uk/mbpdisplay.htm
So not that much help!
But there are basically sort of hotspot areas on the skin tones mostly.
If anyone can shed any light on this, as to whether it may be an issue with the display, or if there is anything I can change in any settings somewhere to fix it, that would be much appreciated.

Having rummaged further, it looks like it's the issue with gradients / dithering.
Seen a few posts on the subject, but no solution.
A lot of stuff about how they display thousands instead of millions of colours, and that's just how it is - but there's no way this is within acceptable limits of a £1,599 laptop.

3 Things I'm hoping to see in Illustrator CS6

1. Better Circle Algorithm
See Marc Edwards' article at http://bjango.com/articles/illustratorandappdesign/
2. Gradient Dithering
Photoshop is finally able to dither gradient layer styles so it can't be that far fetched. Again, Marc outlines this in his article.
3. Inner Shadows
Although I have fine tuned my workarounds to the Nth degree, it's still not as simple as photoshop. 1a would be to have the drop shadow setting produce the same result as Photoshop when using pixel measurements.
These 3 features plus the new pattern editor would (in my opinion) make Illustrator the best UI design tool currently available.

They don't list Layer style gradient dither in Photoshop's feature list, but it is in CS6. I don't have acces to Illustrator CS6 and would like to know if these features/issues have been addressed. Have searched high and low on the net to see if anyone has mentioned any of these things and have found nothing, hence me asking on the forum.
I wouldn't expect Adobe to put 'Now with better circles!' on the feature list.

Dithering (gradient banding) in 32 bit Win 7 on iMac

Heyo, all!
Love my Mac but some things, mostly games, aren't available except on the PC.
When I use boot camp and boot up using Windows 7, I am seeing dithering (gradient banding) on the desktop and in some applications. Generally anything that displays unbroken color gradients and shading. I've tried updating the video drivers but nothing has worked, to date. I'm using a ATI Radeon HD 4850.
Odd thing: In VMWare fusion, no problem, shows fine. When I was using WinXP, no problem, shows fine.
Am I missing setting, somewhere?

sorry if I ask a few silly question :o) :
I assume you are running Eclipse version is 3.5.2 (help/about eclipse platform) ?
And you do "file-new-Other" (don't forget the "other") and you don't see a tab "Oracle Service Bus"?
And if you do "servers, new, server" can you see "Oracle WebLogic Server 11gR1 PatchSet 2 " ?
pierre

Animated gif and gradient color problem

Hi there,
I have been trying to create animated gif with gradient background. My animation works alright but the problem is that when I preview either in browser or in the preview mode, the gradient doesnt look smooth. I want to know if there is a way to work around this? I have seen some on other websites and it looks perfect. I'm wondering how they do it.
I'll appreciate anybody's help.

Cobby Fred wrote:
I have to thank you all for your assistance so far. I must say I only know how to put some few things together and I do not understand some of the technical words you used such as dither, geometric dithering, blending mode... But in all I understand that if I blend colors that are close I'll get something close to smooth.
Here's Wikipedia's page on dither: http://en.wikipedia.org/wiki/Dither Scroll down till you see the images of the cats on the right-hand side. It's a little over half-way. In the original, the cat's pretty much a gradient. If that image is saved as a .gif, see how you get wide areas of color? It almost looks like a paint-by-number picture. This similar to what happens to your radial gradient when you save as .gif. I used the word "geometric," because the radial gradient will produce arcs of grey colors. The third image of the cat has dithering. This means that, instead of setting the colors in the original image to the closest one in the restricted palette, the software tries to figure out how to speckle the image to produce an average that would look better at a distance. Turning on dithering when you go to output your gradient will help make your image smoother, just as the third cat image is smoother than the second. Here's where you turn on dithering if you go through File > Image Preview:
Here I have my unflattened .png file
No, it's flattened. It's a single bitmap.
Here's a noisy image I created. It has a radial gradient at the bottom then a texture layer that has it's visibility off, because I copied it to make a greyscale version to use. I've set that to have 25% opacity and a blending mode of Multiply. Both the opacity and blending mode controls are at the top of the layers panel. This is a FW .png (Edit, no, it isn't! The forum has stopped supporting that! ):
If I export the gradient alone, without the texture layer, I can get this (settings are "64-color" greyscale palette and no dithering):
or this (turning on dithering):
While the dithered image is better, it still shows the arcs. If I set the texture layer back visible, and export with dithering on, I get this:
The noise in the texture breaks up the geometric arcs, which makes the image look smoother. Any small texture will do that. Choosing the 64-color greyscale palette gives me either 22 or 26 colors.The 32-color palette gave me 12. You should experiment with the .gif palette settings to see what you like best. Remember that the more colors you assign to your background the fewer colors you have available for your foreground objects.
As a final point, the red text looks blurry. When I zoom in on it (400%), some of the lines are partially "broken," specifically the verticals on the L, T, and N. You might want to try a different font or adjusting the size so that the aliasing (mapping) to the pixels is cleaner. Italic text in small sizes is difficult to use.

17" MacBook Pro (late 2008) display banding in smooth gradients

While running the Screen Savers: RSS Visualizer or Spectrum, I noticed the graphics background looked different in sort of a high contrast, banded, sort of way. I compared it to the same Screen Saver on my Mac Pro w/30" Cinema Display and that looked fine and smooth, with no banding what-so-ever. Has anyone noticed a difference? Is this what I am to expect with this latest 2008 17" MacBook Pro LCD glossy display??? As a pro-photographer I find it less than perfect for editing purposes and will probably take it back, unless it is malfunctioning and it can be fixed. I've even seen 24" iMacs look practically indistinguishable to the 30" Cinema Display. Are they not able to make the 17" display as well?

I made an appointment with the Genius Bar and they ran an external FireWire Video Graphics Card Tester on the 17" MacBook Pro and found no problems with the NVIDIA GeForce 8600M GT card itself. Although, after consulting with other repair technicians and observing the Screen Saver "RSS Visualizer" and more notably "Spectrum", the Genius's agreed that they was an obvious undesirable increase in dithering contrast compared to LCD iMacs and LCD Cinema Display panels. They concluded it may be a problem with the LED display and wrote up a ticket and took it in for a display replacement repair.
I went nearby to a Best Buy store to see if other 17" and 15"LED MacBooks exhibited the same appearance and found that they also had an undesirable increased contrast in dithering of gradient tonality compared to Apple LCDs. While I was there, I received a call from the Apple Genius requesting I pick up the MacBook Pro since he reneged on his decision to replace the display after further checking and found no evidence of a faulty display. He said "That's the way it apparently is supposed to be." So, currently, my conclusion is that the new LED displays regardless of model are inferior to an LCD in representing smooth tonal gradations.
I hope others more knowledgeable will chime in here to explain what the heck is really going on and offer their experience with these new LED displays. For me, I will continue to test this 17" MacBook Pro LED display with high res images from a Nikon D300 and PhaseOne P45+ comparing them on my MacPro Quad Xeon w/NVIDIA GeForce 7300 GT card and 30"LCD Cinema Display. It may be that the resolution in the Screen Saver is not up to par for these new LED displays.
Visually Yours,
Bruce Christianson

Gradient problems ...funny texture in screen

I just received my new macbook the other day.
The screen seems to have a texture pattern in the display.
I am assuming this is a inherent thing in the glossy display.
The main issue i am having is the color saturation and gradient problems.
I have been using an imac for the last couple of years, and have only started to use leapord for the first time on the macbook. I though apple intentionally desaturated...but in fact, it seems like the screen has a dull saturation.
I am always seeing un smooth gradients and a lot of dithering. In fact, it's almost in everything. I have read things on the internet about 6bit displays and such but never got any real facts.
I have heard of people getting the display replaced because of this issue.
I am not sure if it is common in all macbooks or just some.
Any info would be awesome.
TIA,
ADAM

Thanks for your reply,
I have a few different problems with my screen.
The back of the laptop screen is being illuminated by the lcd,especially around the edges.... It is not as bright as the illuminated apple logo, but is pretty strong.
I've talked to many friends that have mac books and have never seen this issue. I am sending my macbook back and they are sending me another one for that reason.
I also looked at many macbooks just the other day. I looked at numerous friends macbooks and even went to best buy to compare their screens with mine. The screens on those macbooks are far better. Two issues, first, the gradients are really bad. They are not near as noticeable on other macbook screens. Sometimes you have to tilt the screen to notice it, but on my macbook, not very far. It is almost impossible to do graphics work on it. The best I can describe this....it reminds me of a PC that is missing a graphics driver (lots of dithering on subtle color changes.)
Second, the screen has a very light texture being displayed on the screen, kind of a micro check pattern. Again, this is not noticeable on other macbook screens (I originally though it might be on all "glossy" screen).
All in all, Apple is doing the right thing and letting me get it replaced with another. I hope that this macbook I am sending back just had a defective screen...and it's not apple changing the screens on them. If they are in fact changing the screens, then i have to say that they are inferior to the old ones....( and your paying the same price). If the macbook I get in the mail a few days has the same problem, I will, unfortunately, have to return it for a FULL refund. The laptop will not be suitable for the work I do.

Display Dithering (poor quality?)

A couple display observations:
- On light gray screens, such as the first bootup screen, noise/grain is detectable.
- Once in Leopard, display dithering is apparent. For example, hide the dock and look at the darker gradients towards the bottom of the default screen background. Also, look at gradients, such as those in iPhoto's fullscreen viewing mode's toolbar
Have you noticed display dithering on the Macbook Air?

Hiya!
Gee...this is beginning to remind me of the screen problems that people were initially reporting with the iPod touch when they managed to get them . :S Perhaps you might want to try a monitor test of some kind to see how the colors on your screen are supposed to look? Something like this:
http://www.lagom.nl/lcd-test/
Maybe you'll want to see how your screen does if you load up that page first and look at the graphics on it. If something's wrong and you have to go to the Apple Store to have a Genius look at your MBA, you can try loading up this page to show them what's going on.

Gradient with dither creates artifacts beyond gradation

Photoshop CC / Mac (also occurs in CS6, I just never took the time to document it)
Create a gradient mask from 0 - 255, the mask contains a dot pattern in the area that extends beyond the gradation and should be a solid 0,0,0. The dot/pattern has a value of 1,1,1. The gradient was created with in the blue guides (see screen grab) with dither on. The dither is expected within that range but not beyond the gradation. I have tried with and without the "use Graphics Processor" checked.
To test this it's easiest to see against a transparent background, and enhancing with levels.
Is this a known feature

(sorry for the delay, I just found this report on someone else's bug list)
This is the dither option for the gradient tool. You just did an extreme adjustment so the 1/255 dither pattern (1 bit) becomes visible. The same would happen if you did a color conversion with dithering enabled - there would be 1 bit differences even in flat colored areas because a high precision color value was dithered down to an 8 bit color value. You can see a similar issue in 16 bit/channel - but it requires an even more extreme adjustment to see the lower bit.
That is normal, and has been there since the gradient tool was written (or at least as far back as our source database goes, but photoshop 1.0 didn't have dithering on the gradient tool).
If you do not wish to see the 1 bit dithering pattern, then you can disable dithering for the gradient tool.
Now, the fact that it is a pattern -- that is something we might work on. It's using a fixed noise table with offsets, and the offsets sometimes form a pattern for certain image sizes (that are close to multiples of the prime factors used in building and iterating the table). I have ways to make it more random, but it'll take a good chunk of time to code up and test.

When to use dithering?

Hey everybody,
Right now I am exporting roughly 100 titling effecs, one at a time, in 16 bit depth at 1920x1080 resolution. Do I need to be using dithering? I remember reading somewhere (the Motion help manual, I think) that dithering should only be used for 8 bit depth projects. Is this the case? Or is it necessary to to use dithering on 16 bit and 32 bit projects? BTW, my project is mostly monochromatic, white text with an indent filter and drop shadow. The only color is the various shades of white, red and blue on some notepaper images I'm using. The final output will at one point be up on a 50 foot screen, so it needs to be pretty high res. Thanks.
-G
1.6 GHz G5 1.5GB RAM Mac OS X (10.4.3) Final Cut Studio

8 bit projects with shallow monochromatic gradients to lessen/avoid banding.
Patrick

Trouble with brush, gradient and filling

Hello
Few days ago i've found a problem with using brush, gradient and filling layer. Each time i started a new document, filled it with a solid color, it looked ok. But when i added a new layer, added soft brush on canvas and changed the mode to any mode, the result looked very pixelated. I tried with all brush i have and found it just happened with all brush which have outer glow style ( like soft basic brush, lighting...etc..) . It was the same when i used gradient, filled a selection and changed the mode of this layer. You can see it on screenshots below:
Add brush with normal mode:
Overlay mode
Apply gradient with normal mode
Soft light mode
Add more brush with Soft light mode
I've used photoshop for years and never seen such a case before. Any idea is appreciated. Thanks in advance.

"Pixilated" may be the wrong term. If you're talking about the appearance of bands in the data, perhaps you're referring to Posterization.
What bit depth is your document (e.g., 8 bits/channel? 16 bits/channel)? What color space (profile) is it in?
If indeed you're talking about the posterization, look into your Dithering settings. Dithering, when enabled, will mix pixels of adjacent colors to very effectively hide the appearance of posterization.
The following are two gradients, one rendered without [ ] Dither enabled, and one with, then contrast-enhanced with a Curves layer to show the difference in the appearance of posterization more clearly.
-Noel

Gradient Banding

Over the years, working with Illustrator, I've come to expect that Illustrator just has issues with creating gradients, and the only way to get a good smoothly transitioned gradient is to use Photoshop. I'm really suprised that Adobe hasn't solved this by now. I've been reading posts about gradient banding only in dark shadows, that the banding goes away when printing, or people just trying to say that a particular gradient just can't be recreated due to the limitations of a mathmatically defined gradient. I understand completely that the gradients in Illustrator are mathmatically defined, but essentially, so are the gradients in Photoshop, even if they end up being rendered as a bitmap, correct? I also understand that images can have stepped gradients as a result of either color profile gamuts, or bit depth, however the banding I've run into is not a result of monitor issues, limited color gamut, bit depth, or excessively dark gradients and the banding is in Illustrator, shows up in the exported file (tiff, I've also tried rasterizing the AI file in Photoshop with 16bit bit depth and still got banding) as well as prints, because it's IN the file. Below is a Tiff file that shows a gradient created in Illustrator CS6 on the left, Photoshop in the middle and the original art from Illustrator on the right. You won't be able to see the banding in the file unless you download it and view it at 100%.
So, the question is, is there a workaround in Illustrator, or a alternate way of creating gradients directly in Illustrator that will prevent this banding that I've somehow been missing in the 10 years I've been working with it? Also, Adobe, is anyone there trying to solve this problem?
-Matt

Over the years...I've come to expect that Illustrator just has issues with creating gradients...
And what other vector drawing programs have you used which have no occurrence of gradient banding?
...and the only way to get a good smoothly transitioned gradient is to use Photoshop.
Contrary to popular misconception, raster imaging is not immune to gradient banding. Consider: Everything you view on your monitor is a raster image.
I'm really suprised that Adobe hasn't solved this by now....I understand completely that the gradients in Illustrator are mathmatically defined, but essentially, so are the gradients in Photoshop, even if they end up being rendered as a bitmap, correct?... Also, Adobe, is anyone there trying to solve this problem?
What do you expect Adobe to do about math?
I've been reading posts about gradient banding only in dark shadows, that the banding goes away when printing, or people just trying to say that a particular gradient just can't be recreated due to the limitations of a mathmatically defined gradient.
This has been discussed at length many times in this forum. What part of what you've read do you not understand, or to what specifically do you take exception?
I also understand that images can have stepped gradients as a result of either color profile gamuts, or bit depth, however the banding I've run into is not a result of monitor issues, limited color gamut, bit depth, or excessively dark gradients...
How do you know the banding you're seeing is not a result of any of those issues?
...and the banding is in Illustrator, shows up in the exported file (tiff, I've also tried rasterizing the AI file in Photoshop with 16bit bit depth and still got banding)...
Where are you viewing the Illustrator, TIFF, and Photoshop grads? (On your monitor.)
...as well as prints, because it's IN the file.
It's in what you're viewing (the print). To see if the same banding pattern is in the image file, open the image file, count the bands, and actually measure and compare the pixel color values between the bands.
Below is a Tiff file that shows a gradient created in Illustrator CS6 on the left, Photoshop in the middle and the original art from Illustrator on the right.
No, below is a JPEG assembled from screenshots from your monitor.
You won't be able to see the banding in the file unless you download it and view it at 100%.
You can see the banding at multiple zooms. The bands change because of the resampling that your video system performs when it resamples on the fly when zooming or scaling.
...is there a workaround in Illustrator, or a alternate way of creating gradients directly in Illustrator that will prevent this banding that I've somehow been missing in the 10 years I've been working with it?
There is no workaround that will prevent all banding in all grads in all situations in Illustrator or any other program. So you'll have to be much more methodical and specific in your question if you really want to understand what is going on.
All images have banding, just as all raster images have "jaggies" (square-shaped pixels). The issue in both matters is to make them fine enough to be negligible. Visible "jaggies" is a mathematical conseqence involving the factors of pixel size and halftone dot size. Similarly, visible banding is a mathematical consequence involving the factors of number of tones (for each color channel or separation ink) that an imaging device (monitor or press) can render, and the distances spanned by each possible value step.
When you send a linear grad command to a PostScript device, you are sending a command to vary from color A to color B across distance D in however many steps the device can handle. The device does what it physically can within its hardware limitations. Same is true for your monitor. But the limitations differ between the devices.
So the banding you see on your monitor differs from the banding you will see on your desktop printer and from the banding you will see on film seps and from the banding you will see on press. The specifics depend on the specifics of the artwork and the imaging systems.
Your monitor can display 32 bit color. But how many of those values are available to paint a gradient across a distance depends on how many different colors--and how different they are--to be displayed across how many monitor pixels.
Your desktop printer has a fixed number of printer spots. The printer spots are used to build up halftone dots of varying size. How many different sizes of halftone dots (how many levels of gray) the printer can build is a mathematical function between the number of printer spots and the number of halftone dots available within a given distance. If the differences in adjacent tones (for each color separation involved) is finer than the differences between possible gray levels of the halftone screen, posterization occurs and you see it as bands in a linear grad.
An imagesetter has a greater number of printer spots and can therefore build more different sizes of halftone dots for any given halftone ruling than can your desktop printer. So it is less prone to banding because it is able to render more levels of gray. But it is not immune. You can still exceed its possibilities by demanding finer differences in tone steps than it can reproduce.
Banding is quite often ameleorated in print because (depending on the number of component inks involved in the subject color), the bands occur at different places on each color separation. So the resulting bands in the composite are narrower than on the individual seps, and therefore less evident. The same principle occurs on your desktop printer, but at a courser scale, because it is still a lower resolution device.
Banding is just as likely to occur in mathematically applied linear grads in raster imaging. The rasterization is not deferred to print time, it goes ahead and occurs. But the same principle applies. You are still telling the program to change the values of a fixed number of pixels in 8-bit channels. The result is being displayed on your monitor's array of pixels. When printed, the result is being rendered on the printer's (or film's) array of halftone dots. So the banding quite likely occurs at different locations from where they occur on your monitor. In other words, I don't care what you're doing, your monitor is a poor simulation of the banding that will occur in final print. For one thing, everything you see on your monitor involves three "separations" (RGB), not four.
The myth that raster imaging is somehow magically immune to banding stems from the fact that raster images are very frequently not really linear grads, but scattered dithering patterns (otherwise known as noise). If you use a linear grad tool in a raster image program to paint a linear grad across a broad distance of pixels, you can create banding just as you would in a vector-based program. If you do, then the bands are "nailed down" (rasterization is not deferred to print time) and if the imaging device has sufficient resolution, it will faithfully reproduce that same banding.
Back when raster file size was a much greater issue than it is nowadays, there were 8-bit raster imaging programs (Color It! being one example) which could make beautiful images using only 256 different colors. How did they do it? By effective use of dithering. The differences in adjacent colors were greater than those of a 24-bit image. But they were mixed and scattered so as to simulate additional colors, just as the droplets from an airbrush spraying a single color can result in the perception of multiple tones, just as a halftone screen does the same thing. Well, the nature of raster imaging is like that. Zoom way into what looks like a smoothly graduated sky and you'll see that it is in fact a scattered mixture of different values (noise).
Vector-based grads don't involve noise. They're just mathematical commands sent to the printer's interpreter. If you must, though, you can assign noise as a raster effect to your vector objects, and it can reduce visible banding because what you would really be sending to the printer ultimately is a dithered raster image.
So you just have to use some educated (i.e.; experienced) discernment when creating grads and when interpreting what you see on your monitor. And bear the general rules-of-thumb in mind:
The greater the distance spanned, the fewer component inks involved, the smaller the color change, the lower resolution the imaging device...the greater the likelihood of banding.
Set up some simple tests using simple grads in methodically arranged different values. You can build such a test array on a single page, just as you would build an array of color chips. Print it. Study the results.
JET

Photoshop CC gradient banding

I've been a using Photoshop for 20 years (as a design professional) and only since I switched to CC a few months ago, I get awful obvious banding when creating gradients. Even in rgb files, which sometimes used to occur in cmyk. Nothing seems to get rid of it, and I'm NOT looking for a noise or blur solution, please, so don't suggest that.
I did the same actions in Adobe 4 which I still have on another computer and it doesn't occur. I hate this. Any suggestions? I have a new latest iMAC so it doesn't seem like that would be the issue.

Greetings from Boise. I'm in a big rush so didn't go through all the answers, but have some suggestions (late to the party, I know).
In RGB banding is more likely if you have a dominant hue, such as Blue (which accounts for far less than 10% of any Pixel's Luminance value). Additionally, it is more likely if you are working in 8 bits per Channel: this is due to "Quantization," or the Luminance "steps" between various Channel values. In 8 bits, there is a noticeable difference between, say, 125 and 126. Again, this is most noticeable with a dominant Hue. Human vision is very sensitive to Luminance Deltas (differences).
Work in 16 bits, aRGB (not PPRGB, nor sRGB). I know there are arguments for sRGB, but: http://www.forensicallyfit.net/2012/04/16/argb-srgb-gamut-accuracy-conversion-delta-e/
When you build the Gradient, in the Gradient Dialogue Box ensure that you select adjust the "Smoothness" control up. This will produce a dithering in the Gradient to combat Posterization. Additionally, when you have the Gradient Tool Selected, check your Options ToolBar to ensure that the Dither checkbox is selected.
There are some very clever things you can do like overlaying and averaging Gradient Layers, but I mention that because it is only necessary in extreme cases.
Here's a good exercise: generate a monochromatic (0-255) Linear Gradient without Smoothing and without Dither. Note the shape of the Histogram: it is "platykurtic," meaning that it is very flat and regular.
Now-without changing the Gradient Stops or Colors, apply Smoothing and/or Dither, or both:
Now the Histogram is neither evenly distributed nor smooth. This tells you-primarily-that what amounts to Noise (synthetic value Pixels) are being generated by PS in the background to blend the various Quantization Levels. You will probably not find them objectionable in even large prints. This is a superb way to mitigate "banding."
Finally, you should know about a "gotcha." You'll notice (using your Info Window) that the Pixel Values of a 16 bit Document range from 0-32768, which represents 15 bits, not 16 (which would range from 0-65535). So, you are actually working in 15 bits, but the difference is unimportant to the human eye...unless, like me, you hallucinate a lot. Ahem. Anyway, you are in 15 bits (this is done to accomodate Blending Mode math).

T400 LCD Panel Dithering

I have a display issue called dithering.
I run a T400 (2768 GP6) with Windows 7 Ultimate and Lenovo drivers through System Update. It is a switchable graphics model (Intel GMA and ATI HD3400).
No matter which graphic card is set to prior, I get dithering on the LCD Display (1440x900). Most noticeable are color gradients "stepped" and "grainy" colors. It is particularly disturbing on JPG images with compression artifacts on white background, that you usually don't see, the laptop displays dotted areas. Bright areas are like "white burnt", probably because the screen is not able to display light grey.
The dithering doesn't appear on an external display, I thus conclude it is not a software or hardware issue linked to the graphic card.
The dithering appears when running a Linux LiveCD, I thus conclude it is not an issue linked to the Windows installation.
I found in the Lenovo Community an article advising to downgrade the BIOS , when the ATI HD 3400 is set to prior. No matter which BIOS I use (3.13, 3.18 or 3.25), the issue persists.
Do you believe through your experience with these models it is a software issue ? Or a hardware issue where a display replacement might solve the problem ?
Thanks for your attention.

welcome to the forum!
you'll need to reference your defective pixel(s) with the current replacement policy to know if it falls under warranty:
http://www-307.ibm.com/pc/support/site.wss/MIGR-69856.html
ThinkStation C20
ThinkPad X1C · X220 · X60T · s30 · 600

Why my gradients looks like so bad?

When I do a gradient, it looks horrible. Everything is good if I do gradient in white to black etc. But if I do from light black to dark black it looks like this. Different software like Corel Draw gives me a beautiful gradiients. Of course also dark to darker.
I work with CMYK and my graphics card is GeForce 660 GTX
Any tip, advice?

How the gradient appears where people see it should be a bigger concern than just how it appears in Illustrator or Corel.
I did a test in Illustrator CC 2014 and CorelDRAW X7 where I setup the exact same scenario you did. I have a square box with a circular gradient going from 0,0,0,70 in the middle to 0,0,0,100 on the outside. Corel had a significantly better looking gradient than Illustrator did, however the final output of the gradient is what I’m more interested in.
I exported both as 300dpi CMYK TIF files where both images are embedding the U.S. Web Coated (SWOP) v2 ICC profile. Opening both in Photoshop there is almost no visible difference between the two gradients. I placed them both in a single document and masked off one side of the top image. The image below shows how the files look at 300% and you can see that both images have banding. This is because the distance between the two selected colors are too near the other therefore you're going to get more visual stepping. Let me know If you know for certain which side is the CorelDRAW generated gradient. =^ )
If you want to know why Photoshop gradients appear so silky smooth it’s because Photoshop creates raster images which aren’t as mathematically strict as Illustrator and CorelDRAW are. Gradients in Photoshop are generated using dithering to help give the illusion of smoothness. The image below is a 500% close-up of the same 70-100 gradient and if you look close enough you can see that Photoshop is using a random dot pattern to help hide the transition from one transition step to another.

Gradient Dithering

Similar Messages

Maybe you are looking for