White pixels take more power?

I just had a debate with my friend over whether or not white pixels take more power than darker ones. He says that when he set his desktop background to all white, it took more processing power than a black background, he tested it multiple times. I counter-argued that all pixels are allocated a specific hexadecimal value in the RAM, and thus the color should have no effect on the power needed. THIS IS NOT MONITOR ENERGY.

who's right?

All I know is the blue ones are the tastiest.

Black isn't necessarily black, still has to draw those pixels. Back when Google supported the Earth Day (yeah, wiki it) by "Turning out the lights", the bloggers & news portals were quick to note that making Google's homepage all black wasn't exactly helping.

Some LCD's use more energy when showing darker pixels rather then bright ones (so it costs less but hurts our eyes more ;|).

I wouldn't have thought it effects processing power at all, however i believe how it works is this.
-A B&W LCD needs a small amount of power (as in power from the monitor not PC) to turn a pixel black. So, the blacker the display the more power is used.
-On a color i believe this is reversed (although this may be true of OLED/Plasma and older LCD im unsure about TFT). It takes power to make the color pixels turn transparent at varying degrees to allow the light to pass through. So with this in mind the black pixels may have a marginal powersave quality but i would have thought it was negliegable. However, the amount of light energy from the lamp is the same, it is just being masked (athough i wonder if all that trapped light energy may convert to a little heat? but this shouldnt mean the lamp uses more power). With CRT though then white is defo more power hungry.

CRT - White consumes more power, because all 3 beams (Red, Green, Blue) are used at the same time on full power.

TFT - White consumes little less power, because 2 lamps are on all the time using full power to create white and the LCD part in front of the light will cover up everything to make the screen go black, but lights will still be on.

White pixels make you sterile!

In terms of processing power - i.e:processing cycles/speed- black and white will be the same. Copying memory around costs the same whether it contains zeros or 1s.

In terms of electrical power used by the system- 1s use power, zeros don't- so all white should use more- but the effect will be miniscule.

The processing power difference could be possible if some kind of alpha blending algorithm is at work. (Partially transparent windows, smooth fonts etc.) Calculations might be optimized when drawing on all black pixels.

Seems unlikely, but that could be possible.

I think John Galt is right. If you're talking about processing power, the difference lies in the number of pixels, and the format the instructions are given. (jpeg, png, gif, etc.) (No, etc. is not a file type. :P )

But as far as electrical power, I think Digital Anime is right, but I'm not sure on this one, because I'm not sure how monitors work. I use what's displayed, not what displays.

The particle generator's overrun will be far less that the difference generated by the fluedcapacitor.
Therefore the nomen will exceed the claviature. (If deltatime is smaller)

Of course, if the fluxcapacitor breaches the heisenberg compensators past the limits of the isopalavial interface, which controls the main firomactal drive unit, you'll need to reconfigure the activation circuitry of the bioneural gelpack.

I like blue smarties.

I don't like this thread though. Bit stupid.

Was this a later edit or did people not read it?

It was there from the start. People didn't read it.

It was there from the start. People didn't read it.[/quote]

People are stupid, a lot of people never read the posts before theirs. It really ticks me off. Sure I put both in my response, one was what chwaga wanted, and the other was responding to others replies.

ok, Actually it varies frmo screen to screen-

it depends on the polarising panels on the front and back of the display.

The two states operate like this:

1) The display defaults at transparent and dark pixels are created by applying voltage.
2) The display defaults at opaque and light pixels are created by applying voltage, or dark pixels by removing voltage.

1) uses more power on dark pixels.
2) uses more power on light pixels.

It makes more sense for a monitor to default to 1, as computers tend to display more white. Also, you can see this on older monitors as the screen slowly goes a dark grey when they turn off.

I know this frmo playing with those old LCD games as a kid :-)

Cygnus, as the three posts before yours were talking about, he wants CPU power, not electricity used by the monitor.

So?

actually the debate involved both, but I think we're both right in a way...

Uhmm..

Green pixels make you horny...

As far as I can see, if your pixel data is held in the AARRGGBB format, either white, black, pink or horned green, they are all a byte and as such, would essentially take the same processing time, behind the scenes, they have all the same amount of 1's or 0's collectively!

Dabz

aye for the actuall image, its all an uncompressed bitmaps worth of vram (more for alpha), jpges might compress on disc but not in vram thus why should it take more or less processing power to display pure white instead of pure black or any shade of grey inbetween (or colour for that matter)

so i dont get that side of the argument

as for monitor power consumption ive no idea and tbh dont care ... pc's arnt very eco friendly at the best of times so trying to have a 'black' day to save the world ???

There's no way to know how some Windows desktop rendering code may be written. It might first blank the memory to all zeros (for whatever reason) and then draw the background color, in which case it has to do nothing on all black. Also, using an all black image may take more power than no image and just black color.

Ps. If the desktop in question was eg. Ubuntu Linux, you can just go through all the code to prove your friend wrong. Might take some time, though. ;)

Yipee- Grumpy sod- I only responded to the other guys, not the OP. It's pretty obvious to me that the CPU has nothing to do with the colour of pixels.

I know, and I'm good at it. :P But, seriously, monitor energy has nothing to do with what he wants, so there's no point in anyone talking about it...although I do understand where you're coming from.



Good point...you may be right, although it's unlikely.

And I'd like to see someone go through all the code for a whole operating system, and find how it displays pixels on a monitor. :P That would be fun...