New 20" iMac Screens Show 98% Fewer Colors

Trintech points us to an AppleInsider article about another class-action lawsuit directed against Apple Inc. This one claims that the displays on new 20" iMacs are only capable of 6-bit-per-pixel color, 98% fewer colors than Apple advertises. Rather than the 8-bit, in-plane switching (IPS) screens used in 24" iMacs and earlier 20" models, "[t]he new 20-inch iMac features a 6-bit twisted nematic film (TN) LCD screen," according to the article, "which the [law] firm claims is the 'least expensive of its type,' sporting a narrower viewing angle than the display of the 24-inch model, less color depth, less color accuracy, and greater susceptibility to washout." Apple recently settled a very similar class-action suit about the displays on MacBook and MacBook Pro models.

No April Fools articles this year.

[suso]

Good job slashdot, I think you successfully managed to show that reality is stranger than fiction by holding back on the fake articles this year. And you've thoroughly confused everyone.

Re:No April Fools articles this year.

[shaitand]

OMG PONIES!!

Re:No April Fools articles this year.

[pwnies]

You called?

Re:No April Fools articles this year.

[mapsjanhere]

It's 6 bit per color in a rgb scheme, making it 18 bit or 262,144 total.

Re:No April Fools articles this year.

[electrictroy]

Thanks for the clarification. I was sitting here and thinking to myself, "That can't bee right. 6-bits of color is how much my RGB Amiga 500 used in 1987 (64 colors)."

So it's 6 bits per color (red, green, or blue) to achieve 18 bits total (thousands of colors). Versus a "real" monitor that can do 24 bits total, aka millions of colors.

Yeah. Definitely false advertising.
Lousy Apple.
Starting to act like Microscrew.

Re:No April Fools articles this year.

[pipatron]

Technically, you could only define 32 colours of those 64 (from a total palette of 4096!), the other 32 were actually the same colour but at half the brightness, hence the name of the display mode: EHB - Extra Half-Brite. This was very useful since you could use that extra bit-plane as a shadow-plane, and most palettes had dark and bright versions of the colours anyway.

Of course, this doesn't make it any less superior, just saying...

Re: Amiga video modes.

[Thought1]

Actually, there were two 6-bit modes on the Amiga - EHB, as described, and HAM (Hold And Modify), which caused the pixels defined as colors 32-63 to be defined as "the color of the pixel to the left, but with its (R|G|B) value replaced with ...", thus allowing for all 4096 colors on-screen at once, but usually with a slight fudge-factor, depending on your image and how you arranged your 32-color palette. And that's not getting into the later chipsets, which mostly just added bits... (:

Re: Amiga video modes.

[fbjon]

Ah those were the days. Trying to paint in HAM mode, but changing a pixel changed some number of pixels to the right of it, so you had to start painting from the left.

Also, regarding the article, why the heck is Apple of all manufacturers using TN panels, everyone knows they suck! A supply issue perhaps? I know there was a panel factory that went up in flames a while ago, which caused the Lenovo L220X to be severely short in supply.

Re:No April Fools articles this year.

[badasscat]

Thanks for the clarification. I was sitting here and thinking to myself, "That can't bee right. 6-bits of color is how much my RGB Amiga 500 used in 1987 (64 colors)."

No offense intended, but I can't believe in this day and age that people who are otherwise generally well-versed in computers and computer peripherals are still not even aware of this specification for LCD screens - which is probably the most important one.

Everybody gets so fixated on response times and viewing angles, but none of that amounts to a hill of beans without color rendition and accuracy. The most important specs to look at on any LCD screen are bits per pixel and gamut. Contrast is also useful to know if you also know black level threshold. Without that, though, contrast ratio is useless because it's much easier to make an LCD screen brighter than it is to make one darker, and LCD screens these days are by and large capable of much more brightness than would ever be usable. A contrast ratio of 10,000:1 is meaningless without knowing the starting point for that range.

Unfortunately, most manufacturers make the specs that are actually important almost impossible to find. Even a lot of manufacturers who could brag about these things - because their screens do all the right things in color rendition and accuracy - choose not to. Dell, for example, is probably the largest manufacturer of 8bpp Super-IPS screens with wide color gamuts. Their higher-end screens, which are still pretty cheap relative to most screens marketed towards professionals, are among the more capable out there. But I have never seen Dell actually try to make this argument - I have never seen them argue that colors on their monitors are more vibrant and true-to-life (to use the marketing-speak that they'd probably go with), even though they could.

The reason is that people don't seem to know or care. And they should. You're looking at a screen in some cases almost every waking hour you have (if you're like me and work on computers, then go home and switch on your laptop), and many people are using them for things like photo editing or home video production. People should be demanding good color rendition.

It's almost shocking that Apple, of all companies, does not provide 8bpp panels across their entire line. At the very least, given their reputation as a manufacturer of computers for creative professionals, they should be making it clear which screens are 8bpp panels and which ones aren't. And they should be publishing their screens' gamut as well.

Re:No April Fools articles this year.

[randyest]

Great post overall, until about here:

Dell, for example, is probably the largest manufacturer of 8bpp Super-IPS screens with wide color gamuts.

Dell doesn't manufacture LCD panels. They're an assembler and OEM. They buy LCD panels from Philips, LG, Samsung, etc but they don't make any of their own.

Re:

[rsilvergun]

Everybody gets so fixated on response times and viewing angles, but none of that amounts to a hill of beans without color rendition and accuracy.

I'm a color blind programmer.

Re:

[Bigman]

One reason that computer retailers are vague about specs is that it means that they can change the spec at any time without customers having any comeback. So if the start off using 8bpp screens and then want to change suppliers for cheaper 6bpp, if they have specified in the specs they are 8bpp they are stuffed. If all the spec says is the resolution and that the viewing angle is "better than x degrees" they can go shopping with little to worry about.

I had 2 "identical" HP flat panel displays in my job-befo

Re:

[mr_matticus]

Then you have to explain that apple is using LCD specifications that were outdated in the late 90s. (Sadly, the last 6bit pixel LCD I have even seen was a 1999 Pentium II Laptop.) But hey, Macs rule, right?

So you haven't seen a laptop in eight or nine years? Where have you been? Haven't seen a typical consumer LCD from a no-name brand, or the budget version from a bigger label? You seem to have missed quite a bit in the near-decade you've been away from civilization!

There are no 8-bit notebook panels (with the sole exception of some 17" models), and most TN panels (which make up the lion's share of the market, because they're the cheapest) are 6bpp, too.

But hey, don't let the facts get in the way.

Re:No April Fools articles this year.

[Anonymous Coward]

False advertising? Where exactly does apple say the iMac's display outputs 'millions of colours'?

The 'millions of colours' option means (and has meant ever since the Mac IIfX) that a 24bit pallete is used rather than a 16bit pallete. How many colours the display natively supports is a completely different matter.

The culprit here isn't Apple, it's every consumer that ever bought a cheap display and priced true 24bit displays out of the low and midrange market. On any platform if you use a '32bit' or '24bit' setting, you're more than likely not going to be getting that resolution (or even close to it) on your display, whoever the manufacturer.

Show me any audio interface which claims to be 24bit or 16bit resolution and I'll show you it's actual SNR. 98% is a also misleading, like audio color perception is logarithmic. A 16 bit audio CD has 99.996% less audio thingamyjigs than a 24bit professional audio card, but strangely they sound about the same to the untrained ear. 'Creative' declared and actually measured noise floors have always been miles off (when you look at the number 96dB, it's a misprint, they actually mean 69dB)

Who would have thought that the 'low end' mac would have a 'low end' display in it. Nobody expects any other low and mid-range machine to bundle with a professional display, so why should apple users just because they edit some fotos from time to time.

Apples to apples, not apples to lemur testicles.

Professional graphic artists, photographers, compositors and video editors should rightly demand a full pallete, but then that's what those pricey cinema displays are for.

Re:

[NulDevice]

24-bit (and higher) sound is, however, incredibly useful for recording. While the listener can't tell on the raw signal, when you start doing effects processing, you want the most bits available to reduce interpolation errors. Those you CAN hear.

(of course, your 24-bit signal has to actually be clean in the first place for this to even matter, which is another issue entirely)

Re:No April Fools articles this year.

[Capsaicin]

The reason why apple got caught with their pants down in their lawsuit is because for decades professional graphic artists and photographers have used and relied on apple.

Despite being marked down as Troll, this is actually quite insightful. It also shows the depth of the mistake Apple has made. The real cost to Apple is not in settling the litigation, but in the trust that will be lost in the professional graphics market. Up till now you could buy a Mac and be confident that you were getting a machine that was suited for graphics work. While most professionals wouldn't settle for the 20", nor an iMac for that matter, this is the most negative publicity, in one of it's core markets, that Apple could have (not) hoped for.

Re:

[fbjon]

I use the database at flatpanels.dk [flatpanels.dk]. Just input the main part of the model number of the monitor, and you'll get what panel(s) it comes with. There should be other databases around as well. Avoid TN, S-IPS/S-PVA/etc. is probably good enough. But do check reviews, as the panel isn't the whole story, the backlight is important too, as well as a few other factors.

Re:

[electrictroy]

Isn't the proper term "sub-pixel"? (wherein 3 R, G, or B subpixels make a whole pixel)?

(just curious)

Re:

[randyest]

"98% fewer colors" is exactly what it says -- 90% fewer discrete colors. It says nothing about gamut or color range, so it's not misleading at all.

Re:No April Fools articles this year.

[kextyn]

I admit I have no idea how to figure out how many colors there are with 6 bits per pixel...but I did find this website which talks about 8bit and 6bit LCDs: http://compreviews.about.com/od/multimedia/a/LCDColor.htm [about.com]

Re:

[Anonymous Coward]

I admit I have no idea how to figure out how many colors there are with 6 bits per pixel...

2 states for each bit (on or off); 6 bits per pixel; 3 subpixels per pixel (red, green, blue)

(2^6)(2^6)(2^6) = 262,144

Re:

[psydeshow]

I think they mean 6 bits per color channel per pixel. That's just weird.

Almost like going back to web-safe colors again.

Re:No April Fools articles this year.

[Firehed]

It may be weird, but it's also remarkably common. About half the LCDs on Newegg are reported as showing 16m or 16.2m colors, rather than 16.7m (2^24). A far cry from the 280k-odd colors of a 6-bit-per-channel display, but the number they're reporting is based off of the results of a 6-bit panel using dithering. Many cheaper screens from all manufacturers follow this trend, especially those advertised towards gamers. They sacrifice color reproduction in order to get the pixels to twist faster - all of the reported 2ms panels are 6-bit dithered displays, which gives awful color reproduction (not critical for games most of the time, but a big problem for photo/video work). Of course, anything faster than 16ms is absolutely pointless since you're dealing with a 60Hz signal, but that's aside the point. More notably, the 6-bit panels are quite a bit cheaper, as one would expect.

I'm almost positive that my Macbook Pro does this as well; honestly, quite unacceptable for a "pro" machine. It's especially noticeable at the brighter edge of a gradient (ex. the Photoshop color palette).

Most people aren't going to really notice. Dithering is reasonably effective, and it still manages to give the illusion of most of the spectrum (certainly far more than 6-bit/64 levels per channel, rather than 8-bit/256). But at the end of the day it's still an illusion, and the difference IS there.

Re:No April Fools articles this year.

[tftp]

Most people aren't going to really notice.

Yes, I am sure that's what they said at Apple. "Those suckers will never figure out what we sold them!" This is just another proof of "corporate honesty" being an oxymoron.

Re:

[bloobloo]

It may be weird, but it's also remarkably common. About half the LCDs on Newegg are reported as showing 16m or 16.2m colors, rather than 16.7m (2^24).

Eh? How does that work?
If you lose just one bit of colour information, you go from 16.7 to 8.4 million colours. I think they must just be rounding or writing it down poorly.

Re:No April Fools articles this year.

[gnasher719]

Eh? How does that work? If you lose just one bit of colour information, you go from 16.7 to 8.4 million colours. I think they must just be rounding or writing it down poorly.

I think you were not here last year or so, when exactly the same claim was made against Apple monitors (and again, only against Apple monitors, strangely enough not against any other identical monitors):

Each subpixel can display one of 64 values, lets say from 0 to 63. However, each subpixel also can change its value over time. During four consecutive clocks, the sub pixel can have two different values. For example, to produce the values 31 1/4, 31 1/2 and 31 3/4, change the value in a pattern 32-31-31-31, 32-32-31-31, or 32-32-32-31. That way, you achieve 253 different values from 0 to 63 in quarter steps. 16.2 million = 253 * 253 * 253.

Re:

[glittalogik]

Unfortunately, the onlydetermining factor in response time is the LCD panel itself. DVI/VGA cables, screen resolution, and signal refresh rates won't make a lick of difference, any more than standing on one leg when you turn on the microwave will cook your food faster.

Asus, Viewsonic, Samsung and many others have 2ms LCD monitors out nowadays that should set you in good stead for gaming, and prices are as always dropping rapidly.

Re:

[mr_matticus]

No 14" Dell notebook has ever had an 8-bit panel, ever. Give me the model number of the LCD and I'll point you in the right direction.

Re:No April Fools articles this year.

[_xeno_]

No, the article just wasn't clear. It actually means 6-bits per color channel per pixel. In other words, 18-bits per pixel instead of full 24-bits per pixel. And the reduction from 2^24 to 2^18 does indeed reduce the number of colors from about 16 million to 262,144 - a reduction of about 98% of the entire color space.

And as someone who owns a 18-bits per pixel monitor, trust me, you can tell when working with static imagery. Maybe not when playing games or playing movies, but you can tell. The little gradients on Slashdot look terrible on that monitor. It helps that it doesn't do any form of dithering, but even on my cheap Acer laptop that also only does 18bpp, you can clearly see the dithering.

Since Apples are frequently used for photo work and print work, using only 6 bits per color channel is simply unacceptable. Coders probably won't care, but graphic artists most certainly will.

Re:

[gEvil (beta)]

I have a 6-bit LG monitor that I picked up on clearance a few years ago and I have to say that it does an excellent job of dithering. The viewing angles aren't that great, but as far as color goes, it's really not too bad. It's hard to tell even with some of the gradient tests that are out there. OTOH, I've put those same tests up on some of the LCD monitors at work and they look horrible. Point being that there are some 6-bit panels out there that manage to display colors pretty well.

It's truly April Fool's....

[afxgrin]

There's no army of Apple fan boys coming to their defense in this particular case. Come on, you can start posting defensive comments now, it's past noon.

Re:No April Fools articles this year.

[poliopteragriseoapte]

Yes, this is a disaster for graphic artists. One of the BIG advantages of having an Apple machine has always been that the notion of color profiles are built-in, and uniformly applied; even Safari knows how to interpret them in photos. This is ridiculous. Thank goodness one can still get a Mac Mini plus an external LCD... but if the Mac Mini goes, so does Apple's superiority in graphics, and this is a big deal -- and ought to be a big deal to Apple.

Re:

[fracai]

6 bits per channel, so 18 total bits (6*3) vs 24 total bits (8*3)

I was confused at first as well until I worked out the common factor in the discrepancy.

Re:

[Rakishi]

It should be 6 bits per sub-pixel although pixel is probably an accurate way of saying it as well. If you know the context it does make somewhat more sense than saying each pixel can only show 18 bits instead of 24. LCDs have three pixels/sub-pixels next to each other for red, blue and green. 6 bits for each one means that each trio can produce 2^18 colors. At 8 bits they can produce 2^24 colors.

(2^18)/(2^24) = 0.016 2%

Re:No April Fools articles this year.

[gnick]

(2^18)/(2^24) = 0.016 2%

Umm... No. (2^18)/(2^24) = 1.5625%. You were somehow off by a factor of about 96.

Re:

[goofy183]

6 bits per color, I'm assuming (I didn't RTFA) that there are 3 colors per pixel so you have 18 bit color instead of 24 bit color.

Re:

[davidwr]

Maybe next year, they'll unveil "wiki.slashdot.org" due to overwhelming demand by careless users who want to be able to edit their posts and fix their spelling (and everyone else's posts too!)

Maybe next year, they'll unveil "wiki.slashdot.org" due to overwhelming demand by careless users who want to be able to edit their posts and fix their spelling (and everyone else's posts too!).

There, fixed that for you.

Re:

[Jardine]

Maybe next year, they'll unveil "wiki.slashdot.org" due to overwhelming demand by careless users who want to be able to edit their posts and fix their spelling (and everyone else's posts too!).

There, fixed that for you. Citation needed.

Uh oh

[sltd]

Mac Fanboys converging in 3... 2... 1...

Re:

[geekoid]

only after the through on their black wrinkled Tee Shirts, and un-comb(de comb?) their hair.

Re:

[sm62704]

Can I mod the ironic moderation as "funny"?

If anything...

[MrNemesis]

...the new OSX interface has shown us that we don't need so many colours. Colours in a computer eat up the memory bits and distract us from our reverence. Personally, I'm going to take Steve's advice and go get my eyes chromed.

Re:If anything...

[sm62704]

You're absolutely right. I hate colours myself. I much prefer the American colors, thay're much brighter and prettier than the British colours. Damn that Jobs and his British colours! And he calls himself an American!

Re:

[ThePromenader]

Technically, American 'color' contains less bytes than British 'colour'. So although American colors (especially in advertising) tend to be more towards the pure (rgb + complimentary), can we consider them to be the more 'simblastic' of the two?

If only...

[v(*_*)vvvv]

the Windows Guy could retaliate in one of those commercials.

But cutting costs is part of innovation, so Apple is still the best, OBVIOUSLY.

Re:If only...

[Haeleth]

the Windows Guy could retaliate in one of those commercials.

Unfortunately, the vast majority of Windows PCs (including pretty much every laptop ever made) also use these "inferior" screens, and nobody's tried to sue Dell yet.

The fact is that most people can't tell the difference, and aren't interested in paying four times as much to get a product that isn't noticably better unless you make your living working with colour.

This is a storm in a teacup.

Re:If only...

[kesuki]

Dell lets you pay extra to configure your laptop with a real screen. you pay through the nose, but still they let the person decide at checkout.

Re:If only...

[dal20402]

There is no better laptop screen, because no one makes one. You can increase the resolution by paying extra, but you're still getting the same cheap TN crap that everyone uses and that Apple is getting sued for advertising as capable of displaying "millions of colors."

Crap TN panels are slowly but surely taking over the desktop space too. It's hard to find a non-TN panel under 23" these days, and even many 24" and all 27" panels use the sucky technology.

Unfortunately, Americans still largely drive tech trends, and we rarely care about anything but "big and cheap." (We say we do, but then we actually still buy "big and cheap.")

Re:If only...

[tolan-b]

This isn't about the MacBook suit, this is about 20" iMac desktops.

Incidentally a guy (Mac user) on our forums ran some tests on his Thinkpad and found that it does indeed have an IPS display. So although TN screens may be common on laptops they're not ubiquitous.

Re:If only...

[dal20402]

This isn't about the MacBook suit, this is about 20" iMac desktops.

I realize that. I was responding specifically to the inaccuracy in the parent post.

Incidentally a guy (Mac user) on our forums ran some tests on his Thinkpad and found that it does indeed have an IPS display. So although TN screens may be common on laptops they're not ubiquitous.

IBM made several ThinkPads with IPS panels 2-4 years ago, although none were produced in large numbers. The 14" and 15" IPS screens are no longer being made. The only one I know of still being sold is the X-series tablet, which has a 1440x900 12" IPS screen that I believe is also now out of production.

TN was just too big and cheap for IPS to survive. There was no money for the panel makers in producing a tiny quantity of $100 more expensive laptop screens for the few buyers with enough basic perceptivity to tell the difference.

Not so much

[Sycraft-fu]

For one, there are laptop screens that use other panel types. For example LG Display makes the LP201WE1 which is a full 8-bit laptop LCD panel.

Also it is easy to get non-TN panels for desktop displays, you just have to be willing to pay more. For example the LG L1910S is a 19" S-IPS monitor. However, it's going to run you like $350, not the $150 you may be accustomed to for monitors that size. Same deal with larger panels. Yep, you can get 24" TN panels, and you can get them for an extremely good deal. Just

Re:

[thegnu]

Unfortunately, the vast majority of Windows PCs (including pretty much every laptop ever made) also use these "inferior" screens, and nobody's tried to sue Dell yet.

The majority of Windows PCs are non-specific about the superiority or inferiority of their screens. Dell doesn't lie about it. No fraud, no suit.

The fact is that most people can't tell the difference, and aren't interested in paying four times as much to get a product that isn't noticably better unless you make your living working with colour.

Re:

[MojoStan]

Unfortunately, the vast majority of Windows PCs (including pretty much every laptop ever made) also use these "inferior" screens, and nobody's tried to sue Dell yet.

The majority of Windows PCs are non-specific about the superiority or inferiority of their screens. Dell doesn't lie about it. No fraud, no suit.

I just had to check the specs of Dell's (forgotten?) iMac competitor: the XPS One [dell.com]. From the specs:

• Display
  Large Size ( 20" )
  Widescreen
  High Definition: WSXGA (1680x1050) resolution at 16.7 million colors

Hmm... looks like an 8-bit panel.

• Viewing Angle (up to 80 degrees)
  Fast pixel-response rate (5ms typical for fast motion)

Fuck... that looks like a 6-bit TN panel. I'm assuming a viewing angle of "80 degrees" translates to "160 degrees," which is typical of TN panels. Also, I don't think current 8

How can you judge colour quality?

[Ed Avis]

I don't have a Mac, but I do sometimes buy computer monitors. I can understand specifications like the physical size, resolution, viewing angle and (just about) contrast ratio. But do manufacturers publish specs on what colour depth is supported? Is there some quantitative measure of how well a display shows different colours and how wide the gamut is? How can I avoid getting caught out like these hapless iMac buyers?

Re:How can you judge colour quality?

[randyest]

But do manufacturers publish specs on what colour depth is supported? Is there some quantitative measure of how well a display shows different colours and how wide the gamut is? How can I avoid getting caught out like these hapless iMac buyers?

Yes, of course. The LCD manufacturers will spec 6-, 8-. or 10-bit color for their panels. Then Apple will buy the 6-bit and claim it's an 8-bit. Then you sue Apple and get your money back and lunch with Steve, or something like that.

But seriously, yes, LCD (and any decent LCD mfgr) will spec the color bit depth of a panel. A really good mfgr (NEC, LG, Samsung) will have gamut charts available to OEMs and possibly end users. But if Apple chooses not to share, or worse just lies about it, there's not much you can do other than try to do some independent research to figure out what panels Apple uses, then contact the panel mfgr to (try to) get some specs.

Re:

[Mogenpwr]

Most people dont realize (hell, nobody does) that 6 bit panels (as opposed to 8 bit panels) do not lose the 2 MSB; they lose the 2 LSB. As a result,instead of the LSB being a digial 1, it's valued at a digital 4. They are fully capable of displaying the full 24 bit color palette, but they have problems when the image is very dark. To illustrate the point, when the image is Full White (R,G,B=255,255,255) with an 8 bit panel you may see the signal swing from 254-255. On a 6bit panel the signal will swing

Re:

[Toonol]

I believe this has already been tested in court (I think against Palm?), and that the 'dithering defense' was duly defeated. So the perceptual definition may be irrelevant to the cast at hand.

Re:

[Overzeetop]

My guess is to do the research to find out who makes the panels for the model you might want and go from there.

I got a cheap 19" LCD which clearly has a smaller color space based on watching video simultaneously on the LCD and on a HT projector. The LCD clearly has marked color transitions where the (also LCD) projector does not. At least, I presume that's the issue. I don't really care, as the screen is mainly for setup of the HT computer, not watching, but I can see how somebody might be pretty disappoin

Re:How can you judge colour quality?

[xlsior]

Pretty much any monitor advertised as 16.2 million colors is using a 6 bit panel with hardware dithering. Those advertised as 16.7 million colors tend to be 8 bit.

Re:MOD PARENT UP

[vux984]

Mod parent up. This is absolutely true. I'd estimate that the vast majority of LCD panels on the market are 6-bit screens. Whether you are buying Benq, LG, Dell, Viewsonic, it doesn't matter. Most of them are 6 bit.

They are cheaper, and they have faster response times.

8-bit LCD panels are almost a niche specialty 'pro product' in today's market, and unless you went out of your way to buy an 8 bit screen odds are you took home a 6-bit TN panel, advertised as showing "16.2 million colours" without even knowing it.

Its not just Apple. Although they seem to have gone beyond marketing deceptiveness to outright lies and deserve to be taken to task about it.

But don't for a minute think all those free Dell monitors bundled with low end PCs are anything better. Hell, even the ones you can pay to upgrade to aren't often anything better than 6-bit.

Re:

[(H)elix1]

Whether you are buying Benq, LG, Dell, Viewsonic, it doesn't matter. Most of them are 6 bit. ... But don't for a minute think all those free Dell monitors bundled with low end PCs are anything better. Hell, even the ones you can pay to upgrade to aren't often anything better than 6-bit.

For those interested in looking up the monitors, here [wikipedia.org] is a handy guide that gives you the inside scoop on most of the Dell flat panels. Also why the the 200x, 240x, and 300x series monitors get the loving they do and were wo

Re:

[nxtw]

Typically, 24" screens and greater are not TN. This article [behardware.com] claims that the first 24" TN panel came out in mid 2007.

I can't imagine that there are many larger LCD TVs with TN panels, even among the cheap ones; the viewing angles would be unacceptable.

I wish it was that easy

[Sycraft-fu]

Unfortunately, that isn't a good indicator anymore. This is in part because companies are deceptive, but mostly because retailers don't know what they are talking about. At any rate, just do a little searching around and you'll find 6-bit TN panels that are listed on a site as "16.7 million colours". The reason is that the site isn't even checking, they just put that for ALL monitors.

It also goes the other way too. I am thinking about getting an NEC 2690WUXi which is a pro monitor. It is, of course, an 8-bi

Re:

[Phroggy]

I just checked. The tech specs page for Apple Cinema Displays says "Display colors (maximum): 16.7 million". The tech specs pages for the MacBook [apple.com], MacBook Pro [apple.com], MacBook Air [apple.com] and indeed the iMac [apple.com] all mention "millions of colors" (which is what Apple has traditionally called 24-bit color, as opposed to "thousands of colors" which is 16-bit mode and "256 colors" which is obviously 8-bit mode).

Re:

[randyest]

I'm going to have to assume that this post is an April Fool's joke, since it so perfectly is in exact opposition to reality. Nice work!

Class Action?

[randyest]

Strange, the first case that was "settled out of court under undisclosed terms" seems to have been just two guys. Surely there are more than two photographers who bought macs thinking they would get 8-bit color and later realized it was only 6-bit. I wonder why no class-action was initiated? Since it wasn't though, it seems like Apple is still open to potentially thousands or more lawsuits for this false advertising.

That's what it is, right? They say "millions of colors" when it's really 262k colors. Or is there some precedent that lets a company claim dithering = unique color?

Re:

[octover]

What I read was that the capabilities of the screen had to be a major purchase point in order to qualify, and most people would have a hard time arguing that. I'm happy enough with my display, but I am pissed that I was lied to about its capabilities.

No, because quality was obvious

[SuperKendall]

If you are really into photography, you wouldn't buy the non IPS monitors regardless. These guys are idiots for buying a product clearly labeled as having a cheap screen and then being surprised when the screen was, well, cheap!

It's not like dithering is not used in plenty of other applications to produce more colors than the device can physically output - I assume they are going after printer makers next?

Re:No, because quality was obvious

[dreamchaser]

In this instance (not the previous one), the issue is did Apple advertise them as supporting 8 bit per plane or did they not? If they tried selling them as 8 bit and they were really 6, then there is a problem. It's called false advertising.

cheap vs. mislabeled

[davidwr]

If it was advertised as an 18-bit screen we wouldn't even be having this discussion.

18 bits is plenty for many people, but it's not plenty for graphic artists - the very people who buy Macs.

Re:

[hattig]

It's a standard TN screen, as supplied with pretty much any computer under $1000. Indeed gamers prefer this type of display because they don't blur so much as slower response, but better image quality, alternatives.

TN screens are 6-bits per component, alongside a load of other things like poor viewing angles, etc.

They use temporal dithering to achieve 16.2m colours, because they can flick between close together colours very quickly, faster than the human eye can detect. There is another illusion used as wel

Re:Class Action?

[ink]

The CLUT supports 24bpp color, so they advertise millions of colors. If the display dithers down to 262k, it could be argued that the display is still being sent 24bpp info - and thee iMac does have an external video out port, so I think Apple has some wiggle room here...

From this link http://www.apple.com/imac/specs.html [apple.com]

Display

* Built-in 20-inch (viewable) or 24-inch (viewable) glossy widescreen TFT active-matrix liquid crystal display
* Resolution
o 20-inch: 1680 by 1050 pixels
o 24-inch: 1920 by 1200 pixels
* Millions of colors at all resolutions
* Typical viewing angle
o 20-inch models
+ 160° horizontal 20- and 24-inch
+ 160° vertical
o 24-inch model
+ 178° horizontal
+ 178° vertical
* Typical brightness: 290 cd/m2 (20-inch models); 385 cd/m2 (24-inch model)
* Typical contrast ratio: 800:1 (20-inch models); 750:1 (24-inch model)

(apologies for slashdot's mangling of the unicode above)

They make the claim that the "display" supports "millions of colors". And by display, they mean something that has 290 cd/m2 brightness and a 160 degree viewing angle -- which could hardly be referring to the GPU/video card.

Can't say I'm surprised.

[DurendalMac]

I work at an Apple shop, I love Apple products, but I'd be happy to tell you how shitty the 20" Aluminum iMac screens are. They really, really suck, and here's hoping Apple finally gets their head out of their ass and puts a quality screen on what should be a quality product.

Re:Can't say I'm surprised.

[boristdog]

I am not a big Apple fan, but in the past I always knew they at least put out a quality product. I never had problems recommending Apple products to my clients if their needs fit the product.

But in the past few years Apple quality has been slipping. They need to nip this in the bud or they'll be known as just an OS company with crappy hardware.

And for a company that pushes such a visual image - DON'T go cheap on the displays!

I know what they're doing

[jollyreaper]

Apple is just trying to bring back the glory days of black and white screens.

That's OK

[Lxy]

640 colors ought to be enough for anyone.

Macs are for graphics... untrue?

[pwnies]

Hasn't apple prided itself in that mac's are for "fun and artistic purposes" rather than business purposes? It seems to me that apple is shooting itself in the foot here, and then pouring lemon juice on the wound just for good measure.

Quick Ban-Him

[Russell2566]

Quick, ban this guy for posting something that might be construed as anti-apple... We all know they can't do wrong. Someone change Apple -> Microsoft and all will be well...

6-bit colors make gradients look awful.

[Animats]

6-bit colors? In 2008? What were they thinking? The trend is towards 10 bits. At 6 bits, gradients look awful; false edges appear. Go into Photoshop, generate a single color gradient, and then "posterize" to 64 colors to see what this looks like. Yuck.

Dithering won't help; it puts noise into a nice, smooth gradient.

Re:

[randyest]

Do you think 10-bit color provides only 2^10 or 1024 colors? I assume (hope) not, so why would 6-bit color be only 64 colors? The 10/8/6 bits are per channel (Red, Blue, Green) so 6-bit color is 2^18, or ~262k. 8 bit is 2^24 or ~2.7million, 10-bit is 2^30.

That said, you're right that 6-bit makes gradients (and many more things) look like shit. But, to be fair, not 64-color total shit.

The Amiga had 4096 colors (12-bit total, 4 bit per channel) in the 90's. 1024 total colors, now in 2008, on the be

Re:

[dal20402]

You won't see stripes, but noisy transitions, at least on the iMac. Most of the 6-bit displays (including the iMac one) dither when they are fed intermediate values.

For me, the difference is most dramatic on relatively dark gradients involving green or blue.

In any case, the worst problem with TN isn't the dithering/banding, it's the total lack of color consistency that derives from the very narrow viewing angle.

Dithering does help

[Solandri]

Dithering won't help; it puts noise into a nice, smooth gradient.

You're thinking of spatial dithering. LCD panels can use both spatial dithering and temporal dithering. With temporal dithering on a 6-bit panel, the sub-pixel can only be 64 possible states*, but you flip it rapidly between two states to approximate something in between. This is generally invisible to most people. If you can see older fluorescent lights flicker like I can, you may be able to notice it; but for most people for all intents

All I can say is: It better not have a 9-pins

[Zombie Ryushu]

I'm going to be madder then Hell if I take a monitor home, and look at the connector, and it has a 9-pin RS-232 Connector (Thats not a serial connector, the Serial ones are Female.)

Apple monitors give o1000000 colors

[davidwr]

Apple uses octal.

6 Bit per pixel.

[jellomizer]

Just a reminder this is 6 Bits per pixel not the Bit depth that you set on your OS. Having 64 Colors per Pixel and combination of hardware dithering makes a decent screen for most people. However for true videophobes that would get in the way 8 bit would be prefered. But for most people they wouldn't know the difference betwen 8 bit and 6 bit displays.

Re:6 Bit per pixel.

[randyest]

Wouldn't a videophobe prefer to have no display at all?

Great for the environment

[InvisblePinkUnicorn]

I'd call this 98% color reduction a healthy, green approach, great for the environment... except that green was one of the colors that was removed...

Wish they line had a decent display-less product

[octover]

Things like this make it more painful that Apple doesn't have a good performing display-less model. I don't need the power of the Mac Pro, but need more than a Mac mini. The iMac is about the right specs except I'd prefer something that had room for a couple of hard disks. Also more importantly I upgrade my machine more often than I want/need to upgrade my display. I could easily envision going that route getting a good display and then using it for 2~3 machines and buying a MacBook for my portable needs.

The same as 90%+ of LCDs sold.

[guidryp]

90%+ LCD monitors are TN screens like the low end iMacs. They all claim 16+Million colors. The Panel itself is a LG.Philips LM201WE3(teardowns online). The manufacture web says it is 16.7million colors with FRC.

This would only affect the clueless. It was widely complained about that apple switched to TN panel on the 20" as soon as the Aluminum iMacs came out. It is not a hidden fact, you can tell by the viewing angle specs.

Apple will probably fight this one, because there is a chance the laptops did not have FRC dithering (many laptop screens don't) and thus did not have millions of colors, OTOH the FRC dithering panels are classed as having millions of colors industry wide, and the viewing angles were quoted to industry standards in the spec that would make it clear to anyone who knew or cared about display or even asked anyone for advice that these were TN panels.

In fact you would have to be living under a rock to not know, but that won't stop some people for trying for a small cash grab and lawyers from trying for a big one.

This is not as big of a travesty as it seems

[Cowclops]

I'm the guy that you'd find arguing over how much LCDs suck and how much better CRTs are a couple years ago. But my CRT died last month (Mitsubishi 19" Aperture Grille, it was about the best monitor you could get short of the 22" version of the same), and I picked up a Samsung 226CW. There are only two things it doesn't do as well as the CRT:

Absolute black level.
Off-axis viewing degradation.

The color is actually BETTER, DESPITE the 6 bit panel. The reason why 6 bit is not a big deal is because the panel response is so fast that it can temporally dither two colors into one, and you don't even notice that its doing it. For photography, its actually better color reproduction because its more consistent than CRT. On top of that, the "C" model in particular (as opposed to the 226BW) has a 95 CRI backlight, which means the spectrum the backlight produces is much less peaky and closer to natural sunlight. Altogether, the result is more accurate color than I'd get on a CRT. Plus I get 2ms response time so gaming is fine too.

The 226CW may be TN, but its one of the best panels out there. I thought I was going to be more disappointed than I actually was. In fact, I wasn't disappointed at all because it turned out better in most regards, not just "almost as good." It can produce smooth color because spatial and temporal dithering on fast monitors is surprisingly effective, and its actually more accurate because of the better quality back light.

Not that this was an article about CRT vs LCD, but I'm saying that TN panels have become common not just BECAUSE they're cheap but because the good ones (as cheap as they are) are SURPRISINGLY good. Apple may have used a shitty 6 bit panel instead of, say, Samsung's 6 bit panel, but the number of native colors is surprisingly not that big a deal, even if you're a picture-accuracy freak.

(It doesn't excuse them from not clarifying whether it was TN or IPS though, and in fact it pisses me off that no manufacturers are clear on what overall technology goes into their LCDs)

iPhone?

[themassiah]

I'm curious - and completely ignorant of how to find this information. What type of screen does the iPhone use? The WikiPedia entry doesn't give that much detail, or I don't know what I'm looking for. Thanks!

Let's add eye strain to the list of grievances.

[ahfoo]

I use one of these exact machines on the weekend and in the last few weeks I've been having serious eye strain. When I come home during the week and use generic 17" LCDs or my 19" CRT the need to rest my eyes constantly goes away by about Wednesday, but it comes back every weekend when I use that 20" iMac. Seems like a pretty direct correlation. It could be something else like the lighting in the room there, but I'm wondering if anybody else who has used one of these had noticed unusual eye troubles after prolonged usage.

Yet Steve Jobs does not care about it

[Orion Blastar]

because he is colorblind. That is why the original Macintosh and Lisa were in black and white with shades of gray. It wasn't that it was cheaper, it was that Steve Jobs is colorblind. 6-bit or 8-bit color, it all looks the same to Steve Jobs.

On the other hand, Windows and PCs are the way they are because Bill Gates has asperger syndrome.

Linux is the way it is because Linus Torvalds worked his way through college as a nude model for art students to paint or draw pictures of the human body. That is why Linux is open, totally naked.

Re:Only 766 colours anyway.

[Actually, I do RTFA]

Even a screen with an 8-bit DAC is only capable of displaying 766 colours - each subpixel can show 255 brightnesses of three distinct wavelengths of light (as each subpixel can show the same black this makes 766, not 768)

Let's start with, it's multiplicative, not additive. That's 255^3, not 255*3. This is because, as you mentition later, the eye combines all three subpixels into a new color.

And if you want to get really picky, you can only display three colours

If you interpert color as a wavelenght of light as opposed to relative excitment of the three colored cones in your eye, then yes. But no one thinks of that definition. Instead, the obvious usage is 'colors preceived'. Even when you talk about color of a pure wavelength, you can only interpert it as combinations of your three cones.

Whether an individual subpixel can display 256 levels is quite irrelevant since dithering is capable of producing a higher colour depth at the expense of colour resolution

So, even if one were to concede all your points, these aren't really 1920x1280x24 displays are they then. Because that 1920x1280 resolution has to get shortchanged for the dithering. So you can say that Apple lied about the resolution instead of the color if you like, but it's awful pedantic.

Yes, it is possible to build CCDs where the R, G and B are cosited, nobody actually uses the Foveon sensor because the difference in the capture picture is not discernable.

I know people who paid a lot more to get a camera with a Foveon sensor, actually. While I might be unable to notice the quality, they (and their clients) can. And you better believe they would be pissed if they ended up with a Bayer filter instead.

If you want to say that the difference is small, and unnoticible to most people, so that is the optimal thing to make, fine. I respect that, and agree with you. But this is flagrant false advertising. A 1920x1280x24 screen was advertised and not delivered. Bitch about Apple's behavior just like any other major company's.

Re:Only 766 colours anyway.

[dal20402]

Whether an individual subpixel can display 256 levels is quite irrelevant since dithering is capable of producing a higher colour depth at the expense of colour resolution. You still get full brightness resolution. And this is ok, because its not really possible to tell the difference.

Try the following exercise:

1. Find a new 20" iMac (or laptop, or other machine with a crap TN panel). Find a good IPS panel such as the one on a 24" iMac. Put them side by side.
2. Open your favorite image editor.
3. Create a diagonal gradient starting with black and ending with 50% pure blue or green
4. The hard part: tell me with a straight face that you can't see the dithering.

At typical viewing distances, subpixels are small enough to dither with reasonable effectiveness. Full pixels aren't, at least where the color transitions are subtle.

Re:Only 766 colours anyway.

[hankwang]

2. Open your favorite image editor. 3. Create a diagonal gradient starting with black and ending with 50% pure blue or green

(Shameless plug) Rather than creating the image yourself, you can also try The Lagom LCD test pages [lagom.nl] (and try lots of other monitor tests as well).

Re:

[Edgewize]

What the hell is the complaint about? Even a screen with an 8-bit DAC is only capable of displaying 766 colours - each subpixel can show 255 brightnesses of three distinct wavelengths of light (as each subpixel can show the same black this makes 766, not 768). ... This whole thing is stupid. It sounds like people nitpicking advertising, without actually being aware of the technical concepts involved the image display process.

Interesting ... you pass judgement with authority, but nowhere in your post do you

Parent is correct

[spitzak]

If you look closely enough you will see THREE pixels, one red, one green, one blue. Each of these (on an actual 8-bit screen) can display 255 different shades of their color, plus black. 255red + 255green + 255blue + 1black = 766 different colors.

This in fact is the only way to count the colors if you want to claim that dithering does not count. (Conversely if you do count dithering you could claim that the screen can display an astronomical number of colors, if viewed from so far away that the entire displ

no, mod it down for "wrong". or perhaps "troll".

[Edgewize]

No. The post in question is a train wreck. Combinatorial math does not work like that. Nor is that the reason that video recording is historically done in YUV. The human eye is very much capable of perceiving millions of distinguishable colors.

Bringing in the mechanics of color perception is irrelevant, not to mention that the post is using misleading and incorrect terminology (it's nothing to do with "dithering") and that it is conveniently overlooking the fact that the three wavelengths that the cones

Re:

[PRMan]

But when the manufacturer sold it to Apple, they probably didn't lie to them about what it was.

If they did, then Apple should turn around and sue them.