Does an LCD or LED monitor consume lot more power if the pixels are changing their colors or does the backlight have way more impact on power consumption, which makes changing pixel colors' power consumption negligible?
Yes and no.
Some modern LCD/LED monitors does. It can dynamically change it's back-light illumination level dynamically.And conventional LCD monitors, controllers won't do it.
[example google black edition was designed to save energy]
On the other hand OLED technology, yes it does. In OLED it don't have a back-light. Just pixels are tiny LED cells. They are specially made polymer organic LEDs. [Plastic electronics is a hot topic in here these days.]
Added: Its the backlight that consumes power;
The LCD part is like static RAM but refreshed at a slower rate and the storage is a voltage that controls brightness of a pixel. Some dynamic power from mux'ing the row/columns, but basically ITS THE diffused white BACKLIGHT that consumes power... as indicated in more detail below.
Monitors are the biggest power sinks you can manage on laptops. It also is hardest on the eyes. So remember to minimize the brightness to minimum needed. LED's now have a luminous efficacy of 60 ~ 120 lumens / watt and CFL's are maybe 50 lumens/ watt on 3mm tubes on edge lit backlights. but for residential T8's 70~100 lumens/watt is possible. in 5000'K. So when displays are spec'd at 500Lumens say per sq m. the efficacy determines the power dissipated. But with Zone controlled LED backlights as @sandan indicated, you not only get lower power consumption from zone dimming based on black content, but also improved contrast ratio.
I actually measured the power consumption of my LCD monitors with a power meter at one point.
Basically, barring confounding factors (like dynamic contrast, as others have pointed out):
Darker colours require more energy to display.
The change is quite small in comparison to overall display power consumption, but it held true across multiple display panel types (TN, IPS), and backlight sources (CCFL, LED).
This was an answer for the question Does a webpage with a black background save energy? on Skeptics.SE. As such, it refers to Blackle.com, which is a (bogus) website that supposedly "saves energy" by displaying google with a black background, rather then a white background.
It got deleted because apparently it was "Anecdotal" (and no one ever actually clarified what it would require to make it not "anecdotal". I took measurements, reported them, and the methodology used to take said measurements. What more do they want?)
Acer S211HL (LED Backlight):
- 18.1-18.5W - Google Homepage (e.g. mostly white screen), browser in full-screen mode.
- 20.4-20.7W - Blackle Homepage (e.g. mostly black screen), browser in full-screen mode.
Samsung SyncMaster 245BW (CCFL Backlight):
- 63.5-63.9W - Google Homepage (e.g. mostly white screen), browser in full-screen mode.
- 64.5-64.9W - Blackle Homepage (e.g. mostly black screen), browser in full-screen mode.
MacBook Pro 17" Sandy-Bridge (LED Backlight, possible confounding aspects (it's a laptop))
- 20.5-21.2W - Google Homepage (e.g. mostly white screen), browser in full-screen mode.
- 21.7-22.3W - Blackle Homepage (e.g. mostly black screen), browser in full-screen mode.
So, with LCDs, darkening the screen consumes more energy.
It seems the power variance is somewhat dependent on the LCD in question. However, I tested three different LCDs, which use different panel types (TN, IPS), Backlight sources (LED, CCFL), and the increase in energy to display a black subject was common across all of them.
All these monitors were driven via DVI (well, the MacBook is whatever internal bus it uses), at their native resolution. The Acer is connected to a Mac Mini, and the Samsung is a Windows Box. If I wanted to be really thorough, I would use the same video source for all three monitors, but I don't think it's needed here.
The power draw is specified in ranges because the readout measurement on the Kill-A-Watt is somewhat noisy. I took measurements by letting the monitor sit at the specified display image for a minute or two, then monitored the maximum and minimum value displayed on the Kill-A-Watt. Those are the numbers reported for the range in each monitor's power draw.
Note that Kill-A-Watt meters are rated to have a measurement accuracy of 0.2%, calibrated from the factory. Just because they're inexpensive does not make them necessarily inaccurate.
I used Chrome in Full-screen mode on all platforms.
The Only situation where something like Blackle would be of use, or indeed beneficial is with OLED Screens. These screens effectively have a backlight-per-pixel, and they do indeed save power when displaying dark scenes.
However, with any modern single-backlight LCD, the only effect on power draw is the actual panel's power draw, which increases as the transmitted light decreases, the inverse of what Blackle claims.
I also measured the power consumption of my LCD monitors with a power meter at zero brightness. Dynamic contrast is off and to be honest, it's totally pesky so it is allways off anyway. After some hours I checked my measurements with the meter.
My monitors consumes more power when displaying full white!
Dell P2414H monitor - AH-IPS panel with LED backlight measured at zero brightness
total white 10.0 Watt
total black 8.5 Watt
Fujitsu Siemens P20-2 monitor - S-PVA panel with CCFL backlight measured at zero brightness
- total white 30.0 Watt
- total black 28.0 Watt
You may think that I mixed up measurement results, but it is not ! I dare everybody to measure his/her monitor and post results !
The claim that every LCD/TFT monitor consumes less with total white is untrue.
I cannot measure monitor with TN panel as yet because I haven't got. I will measure a TN monitor at my workplace and will post results.
Fujitsu Siemens P20-2 monitor measured again with strobe effect youtube video:
When playing Youtube video "Black and White Strobe Light" full screen the consumption is 28.5 Watt average at zero brightness. The backlight have way more impact on power consumption. This monitor consumes 45 Watts (when display black) and 48 Watts (when display white) at 100% brightness. Full screen strobe 46.5 Watts.