Help: Color rendering w/ Gels

[Arthur Woo]

Recently shot a project, using a turquoise gel over an ARRI fresnel. To the eye, it looked turquoise. To the Scarlet sensor at 3200k, it looked more blue than turquoise. Looking for an explanation for this, and any way to overcome in the future. No r3ds at the moment to share.

Edit: Red/orange gels and anything toward that part of the color spectrum render just fine. Purples/blues/dark greens seem to get washed out and are much lighter on camera than their gels look in real life. Something to do w/ the fact that blue channel is starved when camera set at 3200k?

 

[Ivan McCullough]

It could be any number of things Arthur.

I believe the weak blue channel may be a contributing factor but so could IR pollution too.

FYI colour gels had the same problem with film and would have to be checked/tested prior to production for this very reason.

A gel may colour a light a certain way to your eye but Film or a sensor may be extra sensitive to particular wavelengths of light as a result of the colourtemp of the light and the filter from the gel.

 

[Arthur Woo]

It could be any number of things Arthur.

I believe the weak blue channel may be a contributing factor but so could IR pollution too.

FYI colour gels had the same problem with film and would have to be checked/tested prior to production for this very reason.

A gel may colour a light a certain way to your eye but Film or a sensor may be extra sensitive to particular wavelengths of light as a result of the colourtemp of the light and the filter from the gel.

Thanks Ivan. I've read on the forums people mentioning IR pollution - can you elaborate on this a bit more? Is my sensor/camera not behaving as it should?

 

[Ivan McCullough]

Thanks Ivan. I've read on the forums people mentioning IR pollution - can you elaborate on this a bit more? Is my sensor/camera not behaving as it should?

No, the MX and other CMOS sensors have IR pollution issues, the Dragon seems to have mostly eradicated this but I need to test myself to be sure.

IR comes from heat in general so using tungsten lights will bring the problem on quicker than using other types of lighting, it is still an issue especially in sunlight.

To be honest its a complete PITA.

On the RED it is advised to use a Hotmirror filter either as a combo ND or in front of other filters. (Never stack hotmirrors and always go shiney side out)

I usually use a hotmirror and IRND's together. Most people do not do this but I prefer it. The slight overall green tint that sometimes occurs is preferable and easier to remove thn a jacket being dark brown in one scene and magenta in the next IMHO.

 

[Arthur Woo]

No, the MX and other CMOS sensors have IR pollution issues, the Dragon seems to have mostly eradicated this but I need to test myself to be sure.

IR comes from heat in general so using tungsten lights will bring the problem on quicker than using other types of lighting, it is still an issue especially in sunlight.

To be honest its a complete PITA.

On the RED it is advised to use a Hotmirror filter either as a combo ND or in front of other filters. (Never stack hotmirrors and always go shiney side out)

I usually use a hotmirror and IRND's together. Most people do not do this but I prefer it. The slight overall green tint that sometimes occurs is preferable and easier to remove thn a jacket being dark brown in one scene and magenta in the next IMHO.

Thanks, this helps a lot. I'm going to pick up a hot mirror filter and see if that solves the issue.