Prysm Readies For Launch

Andrew Neale

At ScreenMedia Expo in London at the beginning of May, PSCo were showing the latest generation of the Prysm Laser Phosphor Display (LPD) system on their stand, and announced that they were partnering with Prysm to bring the system to market in the UK.

This was significant – it was the first time the display system has been on public show, as Prysm’s stand at ISE in Amsterdam earlier this year was an enclosed booth and by invitation only.

The units displayed we were told were second generation, but still prototypes, with production units now only weeks away when the system will be officially launched at #InfoComm2010, Las Vegas, in June. The most striking first impression was that the display was brighter overall than we’ve seen up until now, and there were certainly no problems competing with the ambient lighting on the exhibition floor, and other displays on show. The problematic reflections of the glass screen surface on the original prototypes have been largely cured by a new anti-glare surface treatment – you could still see some image banding artefacts on each module, but when viewing from a distance, this wasn’t so noticeable.

The position of the display on the corner of the booth meant that it was easy to appreciate one of the features of the technology, namely the wide viewing angle. Because the light-emitting phosphors are on the back of the glass screen surface itself, the viewing angle is virtually 180 degrees, and you could see this walking down the aisle past the stand.

One main feature of the system that Prysm are keenly promoting is that the power consumption is very low compared to other technologies. At ISE there was a live power meter showing figures in Watts connected to one of the modules, and this was showing measurements in the twenties. It would fluctuate according to the type of pictures being displayed on the screen. Sadly this was not on show at ScreenMedia Expo, so we couldn’t see the consumption this time.

To back up the system’s green credentials, Prysm launched an Eco calculator on their website in April and we’ve taken an in-depth look behind the figures below.

Power consumption of display systems fall into two main categories depending on the primary technology in use:

  1. Projectors and LCD – power consumption is generally constant, and doesn’t vary much according to the images shown on the display. The main power-using component, for example a projector lamp or an LCD backlight, is on constantly.
  2. LED, OLED, Plasma and LPD – the power consumption varies according to the images shown on the display. In these cases, the light output, and therefore power consumption, of these emissive devices is modulated by the image displayed. In the case of a black image, the display is dark and relatively little power is consumed.

This makes it difficult to calculate the on-going power consumption for some types of display, unless you can do some real-world measurements with some typical content being displayed, and ideally measuring over a reasonable period of time.

Prysm have compared the main screen technologies in a table by using a source figure for power consumption (labelled below as Typical Energy Consumption), measured in Watts per square metre of screen area.

They have also sensibly assumed some cooling will be required, and factored in power for cooling to be a third of display power consumption. Given that an electricity unit cost is 15 cents per kW/Hour, the total cost of energy used is shown for each technology type, along with the percentage saved when comparing LPD with the other types.

The standard example in the first table is for a 10sqm screen area display, running 24/7 for 5 years.

Total Energy Consumption LPD LCD LED Plasma LED Rear Projection Tile Large Format Projection
Total Energy Consumption (kW) 2.0 4.6 7.8 8.3 13.3 13.3
Total Energy Cost ($/lifetime) 13,140 30,156 51,115 54,597 87,381 87,381
Additional Cost to Operate vs. LPD 17,016 37,975 41,457 74,241 74,241
% Saved with LPD 56.4 74.3 75.9 85.0 85.0
Display Energy Consumption Amount of energy used while in operation
Typical Energy Consumption (W/m2) 150 345 585 625 1,000 1,000
Display Energy Consumption (kW) 1.5 3.5 5.9 6.3 10.0 10.0
Energy Cost ($/lifetime) 9,855 22,667 38,435 41,063 65,700 65,700
% Saved with LPD 56.5 74.4 76.0 85.0 85.0
Cooling Energy Consumption Amount of energy used while cooling down
Typical Cooling Consumption (W/m2) 50 114 193 206 330 330
Cooling Energy Consumption (kW) 0.5 1.1 1.9 2.1 3.3 3.3
Energy Cost ($/lifetime) 3,285 7,490 12,680 13,534 21,681 21,681
% Saved with LPD 56.1 74.1 75.7 84.8 84.8

At first glance, LPD naturally does very well, and it’s demonstrated that it is the most economic to run by more than twice against LCD, and still more against the others.

However, given the fixed parameters of time and energy cost, we looked more closely at the source power consumption figures and found a case for some adjustments to be made to this model.

Interestingly the figure of 150W/m2 Prysm shows for LPD puts its consumption above the average seen on the module power meter at ISE, and definitely above the “less than 100W/m2” mentioned in their March brochure.

When the display is running digital graphically designed content rather than video pictures it would be fair to say that the power consumption is likely to increase. This sort of content, more typically seen in out-of-home and digital signage applications, tends to have more areas containing higher saturated colours when compared to the video content we have seen on the system to date.

More saturated colour means that the lasers need to be on for longer to excite the phosphor screen brighter. Obviously it won’t all be the worst case of peak white being shown, but we feel that it could be reasonable to increase the average figure to 250W/m2.

A sample of commercial grade LCD panels shows that unless you look at a model which has quite a low brightness, the consumption if anything is being understated. A more realistic average is about 400W/m2.

There is a wider variation for LED screens, with the pitch of LED pixels being a key factor in determining the performance of a display, both in terms of power consumption and picture quality.

Viewing distance from a display is an important consideration when determining what pitch is suitable, and for indoor and higher quality images, 4-6mm pitch is common. Also the power consumption will vary based on the content being played, and the ambient lighting conditions will determine at what brightness the screen will be driven.

The figure of 585W/m2 is probably about correct as an average, but this is the figure most likely to need a proper real-world measurement in the environment in which it is being used to determine a truly accurate one.

The Plasma figure appears to be based on the Orion MPDP panels, but a further look at the latest Panasonic range shows, for example, that their monster 103” panel is much more efficient, at less than 450W/m2. With a few of the other sizes available taken into consideration, a more fair average would be 535W/m2.

LED Rear Projection Tiles can only be a reference to Christie Digital’s MicroTiles product, it being a modular system like Prysm’s LPD, with each display module powered by a LED light source, near throw projector engine. Christie have published conservative figures for maximum power used under full brightness. Prysm appear to have used the typical 110W per tile power consumption figure, and disingenuously taken a 3×3 display (9 tiles) to be a square metre (it’s actually 1.12 m2), and this is the likely source of the 1,000W/m2 figure.

What actually happens in practice is that the array of MicroTiles self calibrate down to the lowest performing LED light source, and Christie’s brochure refers to 30% off the typical power figure. Measurements by the factory show the power consumption to be about 70W per tile unit at their stated brightness. That and the corrected screen area figure means this works out at about 560W/m2.

The comparison with large format projection is rather more difficult to quantify, where any number of parameters would need to be taken into consideration. Screen surface, projection throw distance, and other environmental constraints mean that trying to quantify a general figure for comparative power consumption would be difficult without looking at specific cases, so for now that figure has been left (by us) untouched.

Below is a new example table based on the revised figures:

Total Energy Consumption LPD LCD LED Plasma LED Rear Projection Tile
Total Energy Consumption (kW) 2.5 4.0 5.9 5.4 5.6
Total Energy Cost ($/lifetime) 21,900 34,953 51,119 46,750 48,935
Additional Cost to Operate vs. LPD 13,053 29,219 24,850 27,035
% Saved with LPD 37.3 57.2 53.2 55.2
Display Energy Consumption Amount of energy used while in operation
Typical Energy Consumption (W/m2) 250 400 585 535 560
Display Energy Consumption (kW) 2.5 4.0 5.9 5.4 5.6
Energy Cost ($/lifetime) 16,425 26,280 38,435 35,150 36,792
Additional Cost to Operate vs. LPD 9,855 22,010 18,725 20,367
% Saved with LPD 37.5 57.3 53.3 55.4
Cooling Energy Consumption Amount of energy used while cooling down
Typical Cooling Consumption (W/m2) 83 132 193 177 185
Cooling Energy Consumption (kW) 0.8 1.3 1.9 1.8 1.8
Energy Cost ($/lifetime) 5,475 8,673 12,685 11,600 12,143
Additional Cost to Operate vs. LPD 3,198 7,210 6,125 6,668
% Saved with LPD 36.9 56.8 52.8 54.9

Now that the picture has been re-balanced – no pun intended obviously, Ed – two main results can be seen.

  1. Prysm’s LPD is still the most energy efficient of all the display technologies, but not by the same large margin that was previously claimed. The energy saved over running LDP against LCD is just over 37%, and in the 50-60% range for the others.
  2. The maximum cost of energy saved by using LPD over others during the example 5 year lifetime has reduced to $30,000, or $6,000 per year.

Other parameters like overall display brightness might also be an overriding factor for some projects. Although not reflected in this energy cost analysis, brightness figures claimed by manufacturers can be notoriously difficult to compare, and to actually achieve a consistent brightness to allow energy figures to be compared accurately would really need proper controlled and independent testing. Otherwise, although a useful illustration, Prysm’s Eco calculator can only be treated as a method of modelling a likely scenario.

If we take cost of ownership to include initial purchase cost and energy consumed, in the case of LCD and Plasma panels, the lifetime cost of energy is very significant, accounting for over half of the total cost (assuming purchase price per square metre is in the $10,000’s range).

For large screen modular based displays, like LPD, LED and MicroTiles, with the purchase price an order of magnitude greater (cost per square metre in the $100,000’s), the total cost of energy amounts to 5-20% of purchase cost, rather less significant when considering the total cost of ownership.

Costs such as installation and on-going maintenance and repairs have been ignored, of course, and each technologies’ strengths and weaknesses would need to be factored in when considering a true total cost of ownership.

Other emerging technologies yet to make an impact are LCD panels with LED backlights (instead of cold cathode fluorescent tube) and OLED – for example the Mitsubishi 149″ prototype seen at ISE this year which is competitively bright, and 4mm pixel pitch. Both of these are also claimed to have reduced power consumption when compared to existing technologies – some claim 50 per cent less for OLED, and it will be interesting to see what impact they will have in the power stakes when actually in production form.

While Prysm’s LPD has low energy consumption and the scalable, modular large screen approach on its side, end users will only be able to find all the answers to their questions once the production models are available and the pricing has been announced. All eyes will be on #InfoComm2010, then!

One Response to “Prysm Readies For Launch”

  1. Harry Filer Says:

    Good points on power consumption. Lies, damn lies and marketing comparison tables!

    As you rightly point out, brightness level is not taken into account and so LED in particular is arguably unfairly marked down in power comparison terms. Prysm claim a 1000NIT brightness, LED would typically be 2000-3000NIT. We run most indoor screens at around 30% brightness which would of course mean that the power consumption comes down to a similar amount and neatly promotes it nearer the top of the energy efficiency table.

Leave a Reply