Originally Posted by tlalotoani
Looking for information regarding the DLP Rainbow Effect, found this page in which there is a test, as the disclaimer in the page says:
Legal Disclaimer: This review contains statements of pure opinion and fair comments made in good faith in the interest of the public by AIM Digital Imaging. To ascertain the facts please do your own research. There are also quotes from and links to external sources please see their disclaimers.
this is the page´s link:
And this is the test itself:
You can download the test to see it Full Screen.
For those who KNOW are Rainbow Effect Sensitive, would be nice if you check and see if this test shows what you are actually seeing with a DLP projector, this could be a starting point to those who don´t know if they are Rainbow Effect Sensitive.
And, if this is not a reliable way of testing it also would be useful to say it.
Your input would be usefull.
I've just run this test through my projector (a W1070). I've been RBE-sensitive since the days of rear-projection DLP TV's, and I can attest that it's absolutely reliable.
Thanks for sharing!
It comprises of three consecutive, plain-white-on-plain-black (ie extremely high-contrast) tests; each, I found, more likely than the last to yield RBE. The first is relatively slow-moving, mainly-large objects; the second is much smaller, faster-moving objects; and the last is a single very-high-speed object (a pong-ball!) against a fixed high-contrast background.
It's a great test because it fully demonstrates the scenarios in which rainbows are most likely to present themselves; i.e., high-contrast scenes with fast-moving objects; especially if those objects are small. (Car headlights in the dark are a common real-world-video example of visible RBE that comes to mind).
We've noticed in the past (see here
) that a DLP projector's color-wheel effective speed will often vary quite wildly according to refresh rate; so I ran this at three refresh rates for which the speed is known. It corroborates those color-wheel speed measurements.
- At 24hz: I saw noticeable rainbows on every test.
- At 60hz: The first test was clear; and I saw minor rainbows on the second test; and very noticeable rainbows on the third test.
- At 50hz: Both the first and second tests were clear; and I saw minor rainbows on the third test.
In practice, I'd say the second test is the most realistic in terms of what a worst-case-scenario video will look like; and that's the one I'd recommend to get an idea of a viewer's real-world susceptibility to rainbows. The third test is unrealistic enough that it could probably elicit rainbows from a 20x color wheel
- and as such, I'd ignore it unless you're planning on watching a lot of black-and-white films!
As a side-note, these tests also present that irritating engineering decision on BenQ's part to go with a rather slow speed at 24hz (the default framerate of a Bluray!) when they could have gone faster
Still, at 50hz (which is what I now use for everything), in practice, the W1070 doesn't yield a significant amount of rainbows with regular video at all - at least not to me.