Articles and whitepapers

A New Quality Standard for Projected Images (3/3/2005)

By Greg Jeffreys, Paradigm

That big projected image is the centrepiece of most installations. Yet it is often the least defined and woolliest aspect of any system. Large installations are often specified by AV consultants who use elevated science and methodology to specify and assess other elements, including audio and control systems, but why is this not the case with the projected image? Practical obstacles evidently exist. Let us consider them:

1. Quality standards
Relevant quality standards do exist, but only for discrete elements such as projector (ANSI/APM IT7.228-1997 etc) and screen (ISO11315 pt 3 etc). However, the real deliverable is the combination of the two, namely the projected image, for which no standard exists. Just as importantly, the image sits within its environment - what will be the impact of ambient light?

2. Calculating image deliverables
There are several easy-to-use formulae for calculating image deliverables such as brightness, contrast, uniformity etc, but unless accurate data is used, it is another case of 'garbage in, garbage out'.

3. Projector brightness
ANSI lumens typically provide a kind of 'best case' scenario, using the best lens and a bright lamp, although Presentations magazine made a depressing survey last year of the true out-of-the-box brightness figures in most projectors that they tested. Also, few users realise that lamp life is defined as the time it takes to reach 50% of its initial luminance, so if most users expect the end of lamp life to be lamp failure, luminance can decrease by more than 80%!

4. Projector contrast
There is little mention of ANSI here, although it is included in the standard. Current measurements involve the meter facing the projector - which would blind the viewer - and represent the difference between full projector white and the projector being switched off. The benefit of knowing this information has yet to be established. Leaving the meter facing the projector, but measuring a projected black and white pattern, most projectors struggle to get over 100:1. It gets worse. Most business front-projection installations usually only achieve in the range of 3:1 to 10:1, although home cinema installations, in relative darkness, achieve much more.

Spreadsheet showing the devastating effect of ambient light on front projection screens, which cannot distinguish between projected and ambient light. Black areas in images are shown on white material!

5. Screen gain and uniformity figures
This is a similar story to projectors - the figures provided are mostly for sales purposes and are often not accurate enough to use within predictive spreadsheets.

Groundwork

So, the issues to be addressed have to include a reality check - what can projectors and screens really achieve? Then there is the issue of acquiring reliable data that can be used in calculations and assessments. Shortcomings aside, existing quality standards actually contain most of the science and methodology required - although they arbitrarily separate front and rear projection, despite the science being the same and being able to be compared on a true like for like basis. They just do not go far enough. The issue is more about organising and structuring existing methodologies rather than any need for wholesale reinvention of the wheel.

In 2003 this situation led to an initiative for the creation of a quality standard for the specification and assessment of the projected image. This was initially founded by myself and Søren Weis Lindegaard, Head of R&D at dnp denmark A/S, and has now developed into a committee involving consultants and specifiers from the EU, including Arup and Mark Johnson Consulting, and the US, including ICIA which intends to incorporate relevant materials into its education materials.

We were initially concerned about possible contradictions between the need for a practical 'in-the-field' standard and maintaining scientific rigour. Fortunately however, as in auditory perception, perception of light levels is logarithmic - normally we only perceive a difference in levels if they double or halve. Therefore, although there will be greater potential latitude compared with measuring under lab conditions, it is possible to achieve rigour relative to what we can actually see.

Components of the standard

The standard breaks into three parts:

1. The standard itself, containing the methodology and science in obtaining useable figures, using available test devices and under realistically achievable test conditions. The projector elements will include brightness, contrast, uniformity, and lamp decay factors. Screen elements will include gain, gain angles, uniformity, contrast, and contrast reflectance factor (CRF). CRF concerns the impact of ambient light on rear projection screens, and is only needed for rear projection. Contrast is the key to good image quality - not brightness, and the biggest killer of contrast for front projection is ambient light. Gain figures alone can be used to predict the impact of ambient light on image contrast for front projection.

Demonstration unit showing the difference between front and rear projection, using the same projector and image. The front projection screen looks worse because the contrast level is low, not because it is not 'bright' enough

2. Appendices for recording test data. Only the headline details of the projector will be disclosed (published ANSI lumens, contrast etc). The make and model of projector will not be for wider dissemination - the objective is to avoid situations where, for example, competitors can misapply data from this standard against their own commercially-based data. Possible options include a private database being maintained for committee members, individual practitioners or organisations. Obviously, private databases can record as much information as required.

3. A guidelines section. This will be a separate section due to the possibility of subjective judgements that are inappropriate to a scientific standard. It will initially attempt to make guidelines by referring to other published standards such as CIBSE (Chartered Institute of Building Services Engineers - www.cibse.org) Lighting Guide 1. The proposal is to construct this in the form of interlinked spreadsheets with the option of the individual adjusting and amending performance criteria, according to individual professional judgement.

The white paper

Our initial committee meetings discovered a reassuring level of agreement and some reassuring results based on practical measurement. As it can take many years to get a standard recognised, as a first step, we are producing a white paper detailing best practice for both specifying systems and assessing them after installation. The first draft has just been circulated generally among consultants and specifiers for peer review. (Editor's note: the draft may also be requested by contacting the author via the website below).

The results and materials generated to date have been used in seminars presented at the Infocomm 2004 and ISE 2004 and 2005 exhibitions, and will be presented at Infocomm later this year. Initial feedback has been very positive. We remain optimistic that we can produce a set of guidelines that can be adopted internationally for the professional benefit of consultants and specifiers.

Greg Jeffreys is the Managing Director of Paradigm, designer, manufacturer and distributor of rear projection.

www.rearpro.com