Grid like flare from OLPF

[David Wyatt]

I've seen this phenomena on other cameras - not just a Red problem.

 

[Dylan Macleod CSC]

I think the back reflection is actually from the OLPF not the back of the lense. I believe this because when RED moved the OLPF the character of the flares also seemed to change. Can anyone verify this?

IBloom

Yep, that's what I meant...the OLPF being reflected off the back of the lens, then back in to the sensor...

Dylan Macleod
Cinematographer
Toronto, Canada
www.dylanmacleod.com

 

[Dylan Macleod CSC]

I've seen this phenomena on other cameras - not just a Red problem.

Absolutely but we count on RED to be better!

This problem has been discussed in other threads (in the form of ghosting).

Still waiting for RED to pipe in.

Maybe they can incorporate a fix in the new sensor next year?

Dylan Macleod
Cinematographer
Toronto, Canada
www.dylanmacleod.com

 

[Casey Green]

We've also had this occur in some direct sunlight shots.

It would be nice for RED to let us know exactly what the cause is and how best to work around it. (perhaps there are already ansers in other threads on the subject, but I haven't been able to find them yet.)

Also, if they have any plans to prevent this in future builds or if it is hardware related and only can be addressed in SCARLET and EPIC.

 

[Ed Watkins]

Anyone from RED have any feedback on the 'grid flare'?

 

[David Wyatt]

Yep, that's what I meant...the OLPF being reflected off the back of the lens, then back in to the sensor...

Dylan Macleod
Cinematographer
Toronto, Canada
www.dylanmacleod.com

Personally I always thought it was the sensor being reflected onto the OLPF and then back again onto the sensor??? Moving the distance between the OLPF and the sensor might account for the different look on the old models compared to the new ones? Just idle speculation but would it be practical in the future to have the OLPF on a filter-wheel type rig that you could swing out the way for these tricky flare shots? It'd also be interesting to see a side by side of how different lenses handle these kind of shots (I'd guess lenses with superior anti-flare coatings like the S4s and the MasterPrimes would fair much better than most other lenses?)

 

[Dan Hudgins]

Filter wheel idea...

Filter wheel idea...

Personally I always thought it was the sensor being reflected onto the OLPF and then back again onto the sensor??? Moving the distance between the OLPF and the sensor might account for the different look on the old models compared to the new ones? Just idle speculation but would it be practical in the future to have the OLPF on a filter-wheel type rig that you could swing out the way for these tricky flare shots? It'd also be interesting to see a side by side of how different lenses handle these kind of shots (I'd guess lenses with superior anti-flare coatings like the S4s and the MasterPrimes would fair much better than most other lenses?)

I like your filter wheel idea, but the blank spot would need a thick glass close to the same index to make up the backfocus.

There are just 9 o'clock and 3 o'clock OLPF flters "two point" rather than four point, so the filter wheel could have more than one kind of OLPF to pick from.

I think the OLPF could be replaced by a fast rotating wobble plate behind the lens, but that would have to be in sync with the frame rate and would make noise, the wobble plate would not have the same kind of OLPF artifacts.

While coatings on the lenses help, with a light as bright as the sun you are going to get something even if the coating goes from 1% to 0.5% reflection. The true problem is the radius of the lens elements, since all elements reflect a little the coating just reduces the brightness of the reflections, but their shape and location are more or less the same with or without coating on the lenses. Like I said before, uncoated lenses took into account the reflections more than todays lenses because you could see them all the time, now the reflections are just obvious when there is a bright point light source, so the modern lens designs can show very bad in focus and upside down reflections that would have made an old uncoated lens desiginer look for a new job...

 

[Ed Watkins]

So, having had a chat with the RED techs, the flare is caused (as suspected) by the light bouncing between the OLPF and the back of the lens elements.
It was cut back by some firmware changes and may still be reduced further in future builds.

Currently the best advice they could give was:
1) to try different lenses, some are worse for the flare than others
2) try changing the position of the light source in the frame
3) use heavy ND/IR cut filters and/or try to keep the iris down (i.e reduce the amount of light getting into the lens to try and reduce the reflections.

Guess I'll just have to experiment a bit more to get those shots.

 

[Poi Boy]

Thanks for the info Ed.
Aloha
-A

 

[I Bloom]

Personally I always thought it was the sensor being reflected onto the OLPF and then back again onto the sensor??? Moving the distance between the OLPF and the sensor might account for the different look on the old models compared to the new ones? Just idle speculation but would it be practical in the future to have the OLPF on a filter-wheel type rig that you could swing out the way for these tricky flare shots? It'd also be interesting to see a side by side of how different lenses handle these kind of shots (I'd guess lenses with superior anti-flare coatings like the S4s and the MasterPrimes would fair much better than most other lenses?)

Right, off the sensor to the OLPF and back. So the back of the lense doesn't matter so much.

 

[Julio Quintana]

Just shot yesterday with the new M-X sensor and Master Primes, and I got the grid of red dots while looking into the sun directly. What's the status on this problem?

Julio

 

[Jeff Kilgroe]

Just shot yesterday with the new M-X sensor and Master Primes, and I got the grid of red dots while looking into the sun directly. What's the status on this problem?

There is no "status". It's just an internal reflection. You can duplicate internal OLPF pattern reflections with any CMOS camera... Unless they don't have an OLPF. You can duplicate it with any camera, actually, to some extent. Even film cameras can bounce an internal reflection from the film plane back to the rear element of a lens or other potentially reflective elements in the optical path. Without a digital sensor comprised of "pixels" and an OLPF, you don't see the geometric patterns. You just see some crazy, washed-out flare.

Proper filtering and light handling will go a long way toward eliminating these internal reflections. If you're shooting directly into the sun, be warned that you're putting your sensor at high risk. The lens is like a magnifying glass -- it's not too difficult to scorch your sensor when intense sunlight or laser light shines into your lens. Ever seen SLR mirror shutters get melted when shooting a sunset time-lapse? ...It happens.