Are "LED lumens" a Real Thing?
Trying to figure out what a projector's lumen rating means is getting increasingly harder. Some manufacturers are now quoting brightness for their LED projectors in "LED Lumens" instead of ANSI lumens. Some are putting an asterisk next to the lumen spec on LED and laser projectors with a footnote saying that their rating is based on lamp-based models with the same perceived brightness. The fact that a light meter would report a much lower reading, is implied but never actually stated.
We've asked companies how they've come up with these brightness numbers, but none have been able to supply details. We decided to take a look at the issue ourselves.
A Word About Perceived Brightness vs. Lumen Measurements
It is commonly claimed that LED and laser projectors look brighter than they measure. There are two common arguments for this. The first is the Helmholtz-Kohlrausch (HK) effect, an observation that dates back to the 19th Century. It basically says that the human eye perceives more saturated and purer color as brighter. Since LEDs and lasers may offer improvements in both areas compared with lamps, it makes sense that the HK effect might indeed raise the perceived brightness of pictures coming from LED or laser projectors.
A second argument some manufacturers have made is that most light meters used for measuring projector brightness don't measure LED and laser light accurately. Konica Minolta, which manufactures some of the most widely used light meters, agrees that there is at least some truth to this.
Most of us use light meters designed for lamp-based projectors. The Konica Minolta T-10A is a good example. It has one sensor that measures the intensity of the light coming from the projector over the entire visible spectrum. It then calculates brightness based on the known distribution of light energy at each frequency for a standard lamp and the sensitivity of the human eye at each frequency. The advantage is that this design keeps costs down. The problem is that LEDs and lasers have different distributions of light energy than traditional lamps. So the meter may give a lower reading for those light sources.
Because of this, some projector makers argue that for accurate measurement of LED and laser light engines, you need a different kind of light meter like the far more expensive Konica Minolta CL-500A Illuminance Spectrophotometer. The CL-500A not only measures total light intensity, it measures the distribution of light energy across the visible spectrum instead of making assumptions about how the energy is distributed. And since the human eye is more sensitive to some wavelengths than others, it can then calculate the brightness based on the actual distribution of energy at each wavelength.
How much difference do the meters actually make?
In order to sort this out, we ran two sets of tests. First, we used both the T10 and the CL500A to measure the brightness of nine projectors with different types of light sources, so we could compare the results. Second, we used these results to pair each LED and laser projector with a lamp-based projector with the same measured brightness, so that we could visually compare images side-by-side for any differences in perceived brightness.
More precisely, we used the same lamp-based projector in each case--which we'll refer to as our control projector--to compare with each LED and laser projector. The control projector's manual iris allowed us to adjust its lumen output to the measured brightness of each LED and laser projector for the side-by-side comparisons. Note too that although all of the laser projectors actually had laser-phosphor light engines, we generally refer to them in this article simply as laser projectors.
Our nine projector test units included three with lamps, three with LEDs, and three with laser engines. All three of the LED projectors were DLP. Of the three lamp-based projectors, two were 3LCD and one was DLP. Of the three laser projectors, we had one DLP, one 3LCD, and one LCoS.
Regarding the meters, we used a Konica Minolta CL-500A and a T10, which is the model that the T-10A replaced in Konica Minolta's line. As it turned out, using the T10 led to a discovery about lamp-based projectors that we wouldn't have found if we had used a T-10A.
Meter Reading Differences on Lamp-based Projectors
When using these two meters on the lamp-based 3LCD and LCoS projectors there was little difference in lux measurements between them-they generally measured within 1.5% of each other. On the other hand, with the lamp-based DLP projector, the meters differed by 3% or more in most readings with the T10 reading consistently higher.
When I asked a Konica Minolta representative about this, he said that the rapidly pulsing light caused by the color wheel leads to the T10 measuring DLP projectors as a little brighter than the human eye sees them. The CL-500A is designed to measure that kind of pulsing light correctly and thus give a slightly lower reading.
In short, some meters--including most current models in the $100-$200 range according to Konica Minolta--misread lamp-based DLP projectors, adding a few percent to their actual brightness. The difference isn't enough for the human eye to notice, so it is not significant in any practical sense. But when you see measured lumens in a review, keep in mind that most reviewers use T10 type meters that will often give lumen readings for lamp based DLP projectors that are 3% to 4% high.
Meter Reading Differences on LED and Laser Projectors
Of course the big question is whether the T10 and CL-500A give consistently different readings with LED and laser projectors and, if so, how much different? The answer is a resounding no for the models we tested. Across the board, for the three laser projectors and three LED projectors, the two meters gave brightness readings within 1.5% or less of each other in most instances. There were a few outlier readings but none of any practical significance. There was, in short, no meaningful difference between the two meters in how they detected LED and laser light.
What all this translates to is that which meter one uses can make a difference, but not enough to matter--at least for standard white brightness measurements. When measuring color brightness we experienced one anomaly. All color brightness measurements on eight of the nine projectors were essentially the same for both meters. However on the DLP laser model, the CL-500A measured color brightness at 56% of white brightness, while the T10 measured it at only 31%. That is a huge difference considering these two meters measured white brightness on this model at essentially the same. This aberration remains unexplained, but it may be related to the particular phosphor that this projector was using.
We would have to test more laser and LED projectors to know how common this result is. However, we can say that if a projector has a lower color brightness than white brightness, there are at least some models for which the T10 and similar meters may report a substantially lower color brightness than meters like the CL-500A.
The HK Effect - How do they actually look?
Testing for the HK effect is more of a challenge than measuring brightness with meters. Visual perception is notoriously subjective and several factors can influence it including exactly what colors two projectors show for a given image and any differences in their contrast ratios.
Nevertheless, if the concept of "LED lumens" is at all meaningful in real world use, LED projectors have to leave you perceiving a brighter image most of the time, even with other factors in the mix. With that in mind, we decided to judge brightness using a miscellaneous collection of 52 images, including both graphics and photos, without trying to pick images that would specifically show off the HK effect.
We did a side-by-side viewing of each LED and laser projector against a 3LCD lamp-based projector with a manual iris. The iris let us adjust the 3LCD projector's brightness to within about 2% of the measured brightness of each of the LED and laser projectors we were comparing it against. For each comparison, we connected both projectors to the same source by HDMI using a splitter.
The test consisted of looking at each of the 52 images, one at a time, and recording which projector looked brighter, or noting that they appeared equal in brightness when they did. We ran this test in a dark room because in theory the HK effect should be most obvious in the dark since colors get washed out with ambient light. We also retested with lights on using a light level typical of a living room at night. However, there was no significant difference in results with the lights on. For the sake of simplicity, the table below shows the test results with the lights off.
The table below shows the results for all six LED or Laser projectors. Each row shows the count of how many images looked brighter with the LED or Laser test unit, how many looked brighter on the Control Unit (which was a 3-chip lamp-based model with 100% color brightness), and how many times the images looked equally bright on both. The results list each of the LED and Laser projectors in ascending order of color brightness as a percentage of white brightness.
Comparative Perceived Brightness of 52 Test Images
|LED DLP A|
|LED DLP B|
|LED DLP C|
As the table shows, for all the projectors with a lower color brightness than white brightness, any potential advantage gained from the HK effect is completely swamped by the control projector's higher color brightness. As a side note, most DLP projectors can be calibrated to put out 100% color brightness at the option of the user. However, for this test we used factory default calibrations that maximize white light output at the expense of color brightness. Even with the default settings, these numbers don't mean that you won't get a suitably bright image from a DLP projector. They just mean you won't get a boost in perceived brightness from the HK effect.
As the numbers also show, the lower the color brightness as a percentage of white brightness, the more images look brighter with the control projector. There's also no evidence for the HK effect with the two laser projectors with matching color and white brightness. In both cases, the vast majority of images were judged to be equally bright on both projectors.
The LED projector with matching color and white brightness is the obvious standout. It was judged as appearing to be brighter than the control projector for more than 75% of the test images and equally bright with nearly 15% more. That score is almost certainly due to the HK effect.
With only three LED projectors and three laser projectors in this overview, we need to be careful not to draw overly broad conclusions. However, there are three that we're confident of.
First, with LED and laser projectors, more accurate meters can indeed give higher readings than the most widely used meters for measuring brightness. However, the difference in many, if not most, cases is minor. The only significant difference we saw in this group was for color brightness, and we saw that on only one projector. We saw no reason to doubt reported measurements for white brightness regardless of the type of meter.
Second, for LED and laser projectors with lower color than white brightness, the actual lower brightness for color images will almost certainly be far greater than any increase in perceived brightness that you might get from the HK effect. For DLP projectors, in short, the difference between white and color brightness is the more important consideration by far.
Third, for LED projectors with equal white and color brightness, the HK effect can significantly increase the perceived brightness. However, not all LED projectors deliver matching levels for both, so look for comments in reviews that tell you whether they do. Keep in mind too that although you shouldn't see any benefit from the HK effect in the bright ambient light of an office, we saw no significant difference in perceived brightness between viewing in a dark room and in a room with low to moderate ambient light.