Buyers of professional and technical products like to buy on what they hope will be a rational basis. Of course, there is always room for branding, marketing and the sales and marketing processes, but buyers want to buy products based on specifications, if possible, so that they can contrast and compare different products, ideally with a single metric or two. We’ve seen in the comments on the (you might have thought) simple question of resolution following recent Display Daily articles, that even that is open to controversy. Then, of course, there are lies, damn lies, statistics and ANSI lumens…
I’ve been editing Display Monitor – and now Large Display Monitor, Mobile Display Monitor and Display Daily – for 21 years. During that period, a constant source of controversy has been the brightness of projectors. In the early days, even getting brightness quoted in ANSI lumens was an improvement over lumens that were measured in an unmanaged way.
However, ANSI lumens as a way of measuring projector output had a number of significant disadvantages. In one of the early Infocomm shows that I went to (or it may have been the SID), display technology guru Fred Kahn (the SID Fellow and Karl Ferdinand Braun prize winner) said to me, “An ANSI lumen measurement that has no specified white point has no value”. ANSI lumens are still used by the industry, but very outdated. The ANSI lumen measurement system was developed in 1992, but it was deprecated in favour of IEC 61947-1:2002, which is basically the same (ANSI/NAPM IT7.228-1997 and ANSI/PIMA IT7.227-1998 – were officially retired on July 25, 2003).
Since then, the SID/VESA International Measurement Standard has become one of the main sources of learning and best practice in display measurement. That standard contains (Section 15.3) a method of generating luminance information based on the ANSI/IEC method. (Note – the ICDM is freely downloadable from http://tinyurl.com/p9xp89d – Man. Ed.)
However, the ANSI method was developed in the days when many projectors were based on CRTs and very dim and when the quest was simply for brightness, brightness, brightness. I wrote an article on a product described by Barco as a “Light Cannon” back in 1998 with 2,600 ANSI lumens (at the same time a new CRT projector from Barco, using 12″ CRTs, was just 410 ANSI lumens, but was still said to be twice as bright as “typical” CRTs).
So, when companies were really just trying to get white(ish) powerpoint presentations onto screens, ANSI lumens may have made some sense.
However, one way to get a relatively high lumen rating is to optimise the white performance, while not worrying too much about lower levels of luminance in colour. This is, effectively, what DLP projectors with white segments in their colour wheels do. Going back to my earlier point, this boosts the key metric that buyers of projectors often look for – ANSI lumens.
CLO measures each colour separately – add them together to get the total3LCD projectors, and others which do not use white segments or similar, including 3 chip DLP systems, can produce more of their output with saturated colours and the IDMS contains a second way of measuring brightness, known as Color Light Output (CLO), (Section 15.4). In this method, the output of each colour is measured separately and the results added together.
When DLP projectors with white segments are measured, their CLO is much lower than the ANSI lumen number, while for 3LCD projectors, the CLO and ANSI numbers can be similar. There are three chip DLP projectors (with no colour wheel) that do quote CLO. Some point out that CLO does not specify a white point, so the CLO output could be an “off white”.
None of this is new, so why am I covering it now?
First, without any real study, I had written off CLO as a 3LCD marketing gimmick, but after a long (very long) discussion and meeting with a US colour specialist, Karl Lang, brought over from the US by Epson to talk to the European specialist press on this topic, I have bought into the argument. Lang has measured over 200 projectors with the results listed at http://www.colorlightoutput.com/. Lang told us that one 5,000 lumen projector produced a CLO of just 850 lumens. Lang emphasised the inclusion of CLO in the IDMS, a standard which I have always been supportive of and he convinced me that it has some merit.
Second, at ISE, Gitex and IFA recently, Epson has been showing side by side comparisons of DLP and 3LCD projectors that we had trouble taking seriously (as did others). They made the colour performance of the DLP projectors look very bad. We couldn’t get a really clear understanding of the set up of the DLP units, so we dismissed the comparison. However, we have now been able to establish that the DLP units were set up to achieve the brightness level of the published luminance specification. i.e. a 3000 lumen projector was set in the “bright” mode that reached that level. Epson compared with 3LCD products that have the same headline specification. That’s not so unreasonable and maybe I have to do some of my own testing, just to get comfortable with the idea.
We took this picture of Epson’s LCD vs DLP demo at ISE
Third, at the Display Summit, before Infocomm, I reported that BenQ’s Felix Pementel said that it might be time to measure brightness in sRGB (BenQ Wants Brightness Measured with sRGB (DS9) – subscription required). I raised this idea with Lang at the Epson event and he was very positive about this as it would bring numbers much closer to CLO. He said that to meet the sRGB requirement would mean disabling the white segment on the colour wheel or changing the colour wheel (some DLP makers provide alternative wheels for high luminance or for good colour).
The IDMS is currently being reviewed, so it would be good to see this issue being addressed. Going back to my first point, the real challenge is to summarise a complex issue of luminance, colour and brightness (in a 3D colour space or volume) into a single or perhaps two numbers that can be recognised by buyers. Lang told us that US retailers are acknowledging CLO (Best Buy lists 3LCD projectors with both CLO and “White Brightness”, while DLP units only have “White Brightness” quoted on its website). – Bob Raikes
