or Connect
AVS › AVS Forum › News Forum › Latest Industry News › Epson Color Light Output Demo at CES 2013
New Posts  All Forums:Forum Nav:

Epson Color Light Output Demo at CES 2013

post #1 of 23
Thread Starter 

 

Epson presented an interesting demo at CES this year. It involved a relatively new video-measurement standard adopted by SID (Society for Information Display) called color light output (CLO). Most of us are familiar with white light output (WLO), more commonly called peak white level. And of course, white is simply a mixture of red, green, and blue, so if you were to measure the maximum level of the three primaries and add them together—which yields the CLO—you should get the same value as the peak white level.

 

However, this is not always the case. In particular, single-chip DLP projectors sometimes include a white (clear) segment in the color filter wheel to increase the peak white level, and some filter wheels also include segments with one or more of the secondary colors (yellow, cyan, magenta). Therefore, when you add the maximum levels of red, green, and blue, the result is lower than the measured peak white level, because those extra segments are not used in the RGB measurements, whereas they are used in the white measurement.

 

The effect of CLO was clearly evident in the side-by-side demo shown in the photo above. Two 720p projectors with the same rated WLO were fed the same signal from one Blu-ray player using an HDMI splitter, and each projector was set to its brightest out-of-the-box mode with no other tweaking. They were firing onto identical screens (Screen Innovations Lunar Gray, 16:10, 120 inches diagonal, 0.85 gain).

 

On the right is the image produced by an Epson 750HD, while the image on the left was produced by an Optoma GT750E. Both have a WLO rating of 3000 lumens, and measurements made in the venue were quite close to this value—2920 for the Epson, 2840 for the Optoma. However, the measured CLO was quite different—2950 for the Epson, 980 for the Optoma. As you can plainly see, the Epson's image is a lot brighter than the Optoma's, and the colors are not the same.

 

You might think the CLO measurement is designed to make the specs of LCD projectors look better than those of single-chip DLPs. But the measurement methodology was developed by SID, which has no skin in that game. Also, any 3-chip design, including LCD and LCoS as well as 3-chip DLP, is likely to have equal WLO and CLO. Finally, there are single-chip DLP projectors—those with only red, green, and blue segments in their color wheels—that have the same WLO and CLO measurements.

 

For more on this, visit www.colorlightoutput.com, where you will find lots of interesting info, including the results of independent third-party tests of many projectors. Those results include the manufacturers' WLO specs and independently measured CLO; I'd prefer to see independent measurements of both.

post #2 of 23
Hey Scott, does this in anyway affect contrast?
post #3 of 23
This information is extremely biased by these points:

1. No one (on AVS) runs an HT projector "wide open" for lumens as that hammers contrast and black level
2. No one (on AVS) runs a serious HT projector that puts out 3000 lumens (and probably 150:1 ansi contrast)
3. The CLO number is much less impressive if we're talking 500-800 lumens, which is a more usable output for HT (who uses a 120" 0.85 gain grey screen unless your have too many lumens and zero contrast?)
4. Measurements are meaningless, if a DLP color wheel is designed well enough to trick the human eye but it "measures badly", who cares?

Sorry, just another biased "research report" that proves only that someone is trying to sell ussomething.
post #4 of 23
Thread Starter 
Quote:
Originally Posted by e39mofo View Post

Hey Scott, does this in anyway affect contrast?

I would think so, at least when viewing real program material. One could calculate "CLO contrast" by dividing the CLO by the black level; if the CLO is less than the WLO, the contrast would be lower than the peak-white contrast as well. I'll be talking about this subject on my podcast in February (tentatively 2/11) with a color scientist, and I'll be sure to ask this question.

post #5 of 23
Thread Starter 
Quote:
Originally Posted by pottscb View Post

This information is extremely biased by these points:

1. No one (on AVS) runs an HT projector "wide open" for lumens as that hammers contrast and black level
2. No one (on AVS) runs a serious HT projector that puts out 3000 lumens (and probably 150:1 ansi contrast)
3. The CLO number is much less impressive if we're talking 500-800 lumens, which is a more usable output for HT (who uses a 120" 0.85 gain grey screen unless your have too many lumens and zero contrast?)
4. Measurements are meaningless, if a DLP color wheel is designed well enough to trick the human eye but it "measures badly", who cares?

Sorry, just another biased "research report" that proves only that someone is trying to sell ussomething.

Of course, you are correct that, for home theater applications, a calibrated projector is not run "wide open" at 3000 lumens, and a serious HT enthusiast would not use such a screen as that used for this demo, which was intended only to demonstrate the delta between WLO and CLO. Even under ideal HT conditions, a significant delta between WLO and CLO means that real program material would exhibit less brightness and contrast than the peak-white level and contrast measurements would indicate (not to mention inaccurate colors). I disagree that measurements are meaningless, though they certainly do not tell the whole story. The bottom line is picture quality of real program material, but properly conducted measurements can reveal things that are worth knowing. And don't forget that this measurement standard comes from SID, whose only concern is picture quality.

post #6 of 23
Hmm very interesting. Could be a handy spec to compare unless companies whore the spec out the way all companies do to contrast ratio. Scott its cool to see our posting here I talked to u many times over the years over at HT, under the screen name Jarod, and a big fan! What can I say I love HT geeks podcast. I never miss an episode. Maybe you could get a Dolby rep to come talk about Atmos and it's future?
post #7 of 23
Quote:
Originally Posted by Scott Wilkinson View Post

Of course, you are correct that, for home theater applications, a calibrated projector is not run "wide open" at 3000 lumens, and a serious HT enthusiast would not use such a screen as that used for this demo, which was intended only to demonstrate the delta between WLO and CLO. Even under ideal HT conditions, a significant delta between WLO and CLO means that real program material would exhibit less brightness and contrast than the peak-white level and contrast measurements would indicate (not to mention inaccurate colors). I disagree that measurements are meaningless, though they certainly do not tell the whole story. The bottom line is picture quality of real program material, but properly conducted measurements can reveal things that are worth knowing. And don't forget that this measurement standard comes from SID, whose only concern is picture quality.

I don't know if it's relevent to this conversation but many, if not most home theater 3D projectors are ran wide open to suck out as much light output as possible to compensate. My Ben Q W7000 only runs 3D in high lamp with dynamic iris turned off and iris wide open. Most reviews of other ones ive seen seem to do the same. So is CLO relevent to 3D?
post #8 of 23
Thread Starter 
Quote:
Originally Posted by Reddig View Post

Hmm very interesting. Could be a handy spec to compare unless companies whore the spec out the way all companies do to contrast ratio. Scott its cool to see our posting here I talked to u many times over the years over at HT, under the screen name Jarod, and a big fan! What can I say I love HT geeks podcast. I never miss an episode. Maybe you could get a Dolby rep to come talk about Atmos and it's future?

Hey Jarod/Reddig, thanks so much for the kind words! Glad to see you here! As it happens, my podcast guest on Jan. 28 is scheduled to be Dolby's Atmos guy, Ioan Allen.

post #9 of 23
Thread Starter 
Quote:
Originally Posted by Reddig View Post


I don't know if it's relevent to this conversation but many, if not most home theater 3D projectors are ran wide open to suck out as much light output as possible to compensate. My Ben Q W7000 only runs 3D in high lamp with dynamic iris turned off and iris wide open. Most reviews of other ones ive seen seem to do the same. So is CLO relevent to 3D?

You are correct that in 3D mode, most projectors do run wide open because they need to pump out as much light as possible to combat the drastic light loss through 3D glasses, especially active glasses. My podcast guests on Feb. 11 are scheduled to be a couple of CLO experts, and I'll be sure to ask this question about CLO and 3D. Thanks!

post #10 of 23
RGB Separation is a key measurement for any display, and a good Separation is required if a display is to accurately portray the images presented to it.

As a consequence I fail to see how this 'new' measurement standard is new?

It sounds to me like previous calibration checks have just not been performed correctly when comparing displays or projectors?

Steve
post #11 of 23
Just the headline of this makes me a little crazy. I drove all the way from south of Tucson to Las Vegas to go to CES, and my primary reason for doing so was to visit Epson and see what they were up to. When I got there, I was shocked and dismayed that they were NOT LISTED in the official list of exhibitors! Where in the heck were they, Scott?
post #12 of 23
Thread Starter 
Quote:
Originally Posted by Light Illusion View Post

RGB Separation is a key measurement for any display, and a good Separation is required if a display is to accurately portray the images presented to it.

As a consequence I fail to see how this 'new' measurement standard is new?

It sounds to me like previous calibration checks have just not been performed correctly when comparing displays or projectors?

Steve

Not sure I know what you mean by "RGB Separation" in this context; that is, as a measurement. Perhaps I know it by another name. Can you please elaborate? Thanks!

post #13 of 23
While it is marketing, it sounds like the CLO number is a more truthful indicator of the capabilities of a projector than the WLO number is, so in that sense I think this is a useful measurement to have published for projectors. It reminds me of the CRI number of fluorescent light bulbs, which indicates the accuracy of the rendering of colors of objects when illuminated by such a bulb. Cheaper tubes typically have a lower CRI. It's useful to have that number published in case you care about it (think of art galleries or people who work with colors for a living). I used it to help decide which tubes to put in my home office for example.
post #14 of 23
Quote:
You might think the CLO measurement is designed to make the specs of LCD projectors look better than those of single-chip DLPs. But the measurement methodology was developed by SID, which has no skin in that game.

Just adding fuel to the fire...

"3. The White Brightness data in Table 2 was obtained through Projector Central, as specified by the manufacturers. These manufacturers do not provide Color Brightness (Color Light
Output) data. Color Brightness was determined in compliance with IDMS 15.4 by 3rd party laboratory testing of a single unit of each model. Serial numbers available on request at
Tim.Anderson@3LCD.com"

http://www.colorlightoutput.com/assets/Color_Brightness_Buyers_Guide.pdf
post #15 of 23
Because I work in visual simulation I am very happy to see the CLO spec and I hope it becomes widely adopted. We use a variety of projector products and technologies, including single chip DLP. When we do use a single chip DLP projector, we stipulate a color wheel without a white segment, sometimes called a "simulation" option. Providing CLO along with WTO is a welcome improvement in the spec game.

Do you know if the CIE color coordinates are also part of the information to be specified?
post #16 of 23
This spec is an interesting development.

Many years ago (late 90's) when developing the color space specifications for digital cinema, we made saturated red, green, blue, magenta, cyan, yellow, and white patches on Kodak Vision film. We projected these and measured the chromaticity and luminance of each patch. We found that the luminance of R+G+B was considerably dimmer than the luminance of white (less than half). Of interest is that if one wishes to replicate colour dynamics of film on the digital projector, a projector with color lumens = white lumens will be throwing away bitdepth in the saturated primary colors. On the other hand, if the primary colors match the luminance of film (relative to peak white), the projector will use the full bitdepth to represent the colors. This of course implies that quality 3D color processing is enabled (which is available in the DLP projectors, through P7 color processing.) Our first color models for the digital cinema projectors enabled P7 processing and developed a profile where we depressed the luminance of RGB relative to the CYM and W primaries, and we were able to achieve much better and smoother replication of the film colour relationships.

Matt Cowan
post #17 of 23
1) Most of us know the whole brightness measuring thing is a marketing tool, but it is useful - I prefer the screenshot of the Optoma on the left, but in a sunny room, the Epson's picture might be more viewable

2) While colour accuracy can be measured, ultimately it all depends on perception, not measurements.

The average between red and green peaky filters are interpreted by our brains as YELLOW when the projectors or displays are outputting a spread between red and green that might not even have the yellow wavelength, or might not have it as the brightest wavelength. The viewer's retina processes the intensities, and our brain fills in the gaps.

It is possible to produce displays that have better gamut or CRI by using fill-in colours (e.g. yellow, emerald green, violet), and even to have better colour storage systems than RGB (IR,R,G,Y,B,V,UV anyone?), but unless you have butterfly or cephalopod eyes, you can't tell the difference!



Maybe evolution will product super AV eyes in a few generations, which will flock to buying super-high-CRI displays smile.gif
post #18 of 23
> www.colorlightoutput.com
Wow! Very interesting.
It is good to see a technical topic such as this discussed with hard data.
However, the projectors they have chosen to test, the way that the data is presented, and their failure to specifically disclose who is running this website is problematic.
Firstly, why have they chosen to only test white segment DLP projectors in their Table 2? Some manufacturers sell both DLP projectors with an RGB colorwheel, and DLP projectors with an RGBW colorwheel (and other variants too). Some manufacturers sell both LCD and DLP projectors, but again only DLP projectors with a white segment colorwheel have been tested. This omission and the failure to explain the selection clearly represents a bias. Without adequate explanation, the average consumer would be influenced by the fancy website and the incomplete data to assume that Epson, Hitachi, Sony = GOOD; and Acer, BenQ, Casio, etc = BAD. But this is an incorrect conclusion based on the incomplete data, hence the accusation of bias.
Secondly, although they try to suggest that this is all open handed and fair by pointing to the SID IDMS, their failure to clearly disclose who has funded and develop this website makes it look like they're got something to hide.
I only found two mentions of the 'owner': The 'Buyer's Guide' lists an email address for a 3LCD.com employee , and the very last page on the site 'Feedback' lists this at the bottom of the page: "Copyright © 2012 3LCD Business Center All Rights Reserved"
For those who know about projectors, and knowing who is behind this website, it becomes clear what is going on.
In Australia I think this site would sail very close to breaching consumer protection law - perhaps this would be the case in other regions too.

On the positive side, hopefully this will encourage all manufacturers to publish the CLO values of their projectors, plus clearly specify the types of colorwheel used (RGB, RGBW, RGBCYW), the colorwheel speed, and also the brightness (luminance) of projectors when they are switched into 3D mode.

Best,
Andrew.
post #19 of 23
The only place I know of where they issue RGB separation graphs in their reviews is ixbt.com. I agree this aspect of projectors is very important and it is very good it's now formalized and known as CLO. However I also believe that this particular site > www.colorlightoutput.com < is biased. They clearly play against DLPs. For example, look at those numbers:
BenQ W7000: WLO - 2000, CLO - 1500. W7000 can achieve 2000 lumens in High Lamp dynamic mode with Brilliant Color On. However it is known what BC does to CLO:

BC on (W7000, graph taken from ixbt.com)


BC off (W7000, graph taken from ixbt.com)


As you can see, the separation is almost perfect and CLO should be very close to WLO.

Epson 5010 (profile Dynamic)


Epson 5010 (profile Cinema)
post #20 of 23
Quote:
Originally Posted by Matt Cowan View Post

Of interest is that if one wishes to replicate colour dynamics of film on the digital projector, a projector with color lumens = white lumens will be throwing away bitdepth in the saturated primary colors. On the other hand, if the primary colors match the luminance of film (relative to peak white), the projector will use the full bitdepth to represent the colors.
I'm not quite following this. Could you perhaps explain this a bit more please?
post #21 of 23
Elix, thank you for your post that info is extremely enlightening, the graphs in particular. It's sad that a typical consumer would need to do so much of their own research and seek out websites such as avsforum with largely amateur individual contributors to find more accurate and useful information.

I consider myself to be a very anal retentive consumer when it comes to researching before purchase, but I find that it is extremely difficult to find reliable, independent sources to provide information on many types of products, ESPECIALLY televisions and projectors. It's horrible to think companies are making it even harder with blatantly misleading marketing information on websites trying to appear as if they are independent.
post #22 of 23
Thread Starter 
Quote:
Originally Posted by Donny Bahama View Post

Just the headline of this makes me a little crazy. I drove all the way from south of Tucson to Las Vegas to go to CES, and my primary reason for doing so was to visit Epson and see what they were up to. When I got there, I was shocked and dismayed that they were NOT LISTED in the official list of exhibitors! Where in the heck were they, Scott?

Actually, Epson was not an official exhibitor at CES, which you could have determined by going to the CES website and searching for it in the exhibitor list. (This would have returned no results, indicating it was not an official exhibitor.) The company puts more emphasis on CEDIA in September, especially with regard to its home-theater projectors. They did set up a meeting/demo room at the Aria hotel, mostly to meet with retailers. However, they did invite journalists such as myself to come see the CLO demo. I hope you saw enough other cool stuff to make the trip worth it!

post #23 of 23
The charts provided above for the BenQ W7000 are great…I’ve never seen the concept communicated in exactly that manner. It’s a great visualization, and clearly illustrates the gap between white light output and color light output.

As pointed out, once BC is turned off the CLO and WLO should be equal…and the data supports this.

We had a W7000 tested. It measured 1500 lumens of color in its brightest mode. Much lower than its peak white level spec’d at 2000lm. But, in the other modes the CLO matched the WLO:

Standard: 1224 white lumens & 1231 color lumens
Cinema: 908 white lumens & 910 color lumens

Obviously brightness is only one metric, but a useful one if there is any ambient light or if you’re going after a large screen.

You might recognize my name…it’s the one on the four page document highlighted above and available at www.colorlightoutput.com I’m a product manager for 3LCD.

I’m a little surprised by the comments suggesting we were trying to hide the identity of 3LCD behind the site. Clearly the site doesn’t scream 3LCD…it wasn’t supposed to. The “Hero” of the site is Color Light Output. The purpose is to provide information about this new measurement methodology…not present the technical details of 3LCD. I thought the ‘feedback’ page fairly well spells out who was behind it. That said, I will take these comments and make adjustment so that’s it’s clearer who is supporting the site.

Regarding the projectors selected for testing in table 2 of the document. It is true that all of these projectors are single chip models with color wheels. Why is that? As Scott points out above, an RGB 3-path projector will always have equal parts of WLO and CLO. I know already how an NEC LCD projectors is going to perform. Only single chip projectors were tested in order to better understand how each Color Wheel design impacted CLO. I do admit that the list is heavily leaning towards the biz/ed side of the projection market…that’s due to the makeup of sales volumes; only about 10% of projectors are sold into home theater.

I hope, regardless of the company on my business card, that you’ll agree with me that providing the customer this additional data is a good thing. My aim here is to get all manufacturers to list CLO as a supported metric.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Latest Industry News
AVS › AVS Forum › News Forum › Latest Industry News › Epson Color Light Output Demo at CES 2013