Blur Busters Forums

I'm testing two different DLP projectors. On the left is 1080p240hz from a single laser phosphor color wheel light source, on the right is 1080p240hz from a 4LED (RGB+Boost B) light source. I'm trying to understand why the sync track rainbow effect is aligned RGB on the right, but alternating BGR, RGB on the right. Is this as simple as a color wheel with BGRWRGBW configuration, or is there something I'm not thinking of?

TheHookUp wrote: ↑

30 Apr 2024, 08:26

I'm testing two different DLP projectors. On the left is 1080p240hz from a single laser phosphor color wheel light source, on the right is 1080p240hz from a 4LED (RGB+Boost B) light source. I'm trying to understand why the sync track rainbow effect is aligned RGB on the right, but alternating BGR, RGB on the right. Is this as simple as a color wheel with BGRWRGBW configuration, or is there something I'm not thinking of?

Yes.

There are a few possibilities:
- Yes, Different colorwheels. One projector may use an RGBRGB wheel, and another projector may use a BGRWRGBW, or other odd pattern.
- Or refresh rates are not always fully integer synchronized to colorwheel. For example, some refresh rates means blue is filter is still in place at the end of one refresh cycle and the start of a new refresh cycle. However, I do see RGB followed by BGR followed by RGB. This is suggestive of different colorwheels in this case.
- Or you may have a situation where the positions of pixels during temporal dithering may be spatially in different positions during mid-pursuit camera.
- Combination of above.

But it is definitely a colorwheel that reverses color sequence because I see another point of evidence: The double black lines in the red UFO body.

I know why that happens.

The reversing of temporal color sequence every other refresh cycle, also leads to mispositioning artifacts of pixels especially since R is late every other refresh cycle, and early every other refresh cycle.

As your analog moving eyes move along the axis, your eyes are in different positions early in refresh cycle versus late in refresh cycle. So the black pixels versus red pixels are quite significantly offest because of early R versus late R, in alternating refresh cycles. This definitively (for me) confirms colorwheel does have an RGB and a BGR sequence (reversing).

This creates the double-black-lines effect in the red body of the UFO. I am not a fan of changing temporal color sequence in consecutive refresh cycles -- they should have designed colorwheel to be RGBRGBRGB.

It's a very interesting artifact. However, for determining pursuit camera tracking accuracy, ignore the color (you can make the image monochrome to make it easier to interpret, if necessary) and worry about the overall black-versus-notblack tickmark consistency, as seen in the advice in other threads;

Perfect, thank you.

Since you say this is a single-laser phosphor projector, it's possible it's a 4-segment color wheel where you've got a long R phosphor segment, short G phosphor segment, long B phosphor segment (or clear if it's a blue laser), short G phosphor segment, in this proportion: RRGBBG, basically, R as 1/3 of colorwheel, then G as 1/6th, then B as 1/3rd, then G as 1/6th.

This would match the artifacts I am seeing. This would do RRGBBGRRGBBGRRGBBGRR. Generally, laser phosphor projectors omit the W segment (as it does not use a pure white light source).

There must be a technical reason why they did -- but most certainly did not benefit motion quality. Best avoided if motion handling is critical. It may have been been an unforseen side-effect.