Creative Communities of the World Forums

The peer to peer support community for media production professionals.

Activity Forums DSLR Video Color saturation issue with theatre lighting

  • Color saturation issue with theatre lighting

    Posted by Tim Neighbors on November 9, 2018 at 3:56 am

    I’ve had this issue here and there over the years when shooting video with theatre lighting and I’ve never been sure of the cause and what I can do about it -both to prevent it and to try to fix it in post. I think it is caused by overly saturated colored light hitting the subject but I notice that some cameras seem to handle it better than others and I’m not sure what this undesirable effect is called. What camera specs should I look for to avoid this in the future if I buy a new camera, what settings can be tweaked to handle it better with my current camera? This was shot with an A7s (first gen) going into a shogun 4K recorder and being recorded at 1080 10bit (I believe) prores. Meanwhile my little 8 bit osmo with x3 camera handled this exact same moment fine. Thanks for any help with this.

    Tim Neighbors
    invisibleharness.com
    FB.com/invisibleharness
    instagram.com/invisibleharness

    Steve Crow replied 5 years, 11 months ago 3 Members · 8 Replies
  • 8 Replies
  • Chris Wright

    November 9, 2018 at 4:16 pm

    that may be infrared taking over. you can try color mode transfer and levels white output to clip the chroma, but the cri levels probably washed out the rgb codes. might have to do some masking or even painting. probably better to key to white and add glow if low on time. channel mixer might help overall balancing too. blue channel blue-blue

  • Tim Neighbors

    November 9, 2018 at 7:23 pm

    Thanks for the response. Is there a certain type of light that I should try to discourage them from using? Why are some cameras able to handle this situation so much better than others? What is meant by “blue channel blue-blue”?

    Tim Neighbors
    invisibleharness.com
    FB.com/invisibleharness
    instagram.com/invisibleharness

  • Chris Wright

    November 10, 2018 at 5:37 am

    yea, stay away from single colors and low cri lights. Colors come from high cri and broad spectrum colors. the blue saturation clipped before the white did. the blue-blue is in inside channel mixers. sometimes ir filters can help, sometimes nothing can if using solid gels. there’s just no info reflecting into the camera.

  • Steve Crow

    November 11, 2018 at 6:43 pm

    I’m not sure but maybe you could try shooting with a flat color profile or even a log. It appears that those areas which are deep purple are essentially blowing out. (It makes total sense to me that the forehead and nose are two areas being overexposed due to sweat) Because of the colored gels being used for the lighting what would have been blown out whites is now blown out purples. This is a theory. I wonder if you could possibly get an actor to stand under the lights in the same lighting conditions to try different color profiles in your camera.

    Failing that, maybe shooting a little underexposed to protect the highlights and then bringing up exposure carefully in post. This is an interesting problem and one that I’ve actually wondered about, like how to do video shooters successfully film bands during concerts when the lighting is often so challenging.

    Please post back if you find a working solution – otherwise you may just have to live with it or have the actors use heavier makeup on the forehead and nose areas.

    Steve Crow

  • Tim Neighbors

    November 11, 2018 at 7:01 pm

    Thanks for the response. This was actually shot in SLog2. One thing I’ve tried to do in the past is to turn down color saturation in camera, though it’s difficult to know if it helped because I’ve never been able to do a controlled test. As of now I’m guessing that it is clipping the color saturation early in the chain so adjusting color saturation probably isn’t helping. I think making a request for more makeup might be a little out of line for me since the show is primarily done for the audience and not for camera.

    Tim Neighbors
    invisibleharness.com
    FB.com/invisibleharness
    instagram.com/invisibleharness

  • Chris Wright

    December 4, 2018 at 5:08 pm

    I saw before and after pictures of this technique and it looks like it worked. ( think he used resolve though)
    It appears he shifted the clipped gamut into a larger gamut and then re-balanced the white balance.

    “I decided to change the input colour space to the timeline (Rec.702/2.2) and just rebuild the gamut/space transformation from slog3 to rec709/2.2 myself. It changed the wb to get things safe so I lost the blue in the haze but I can build that back in. A little bit of tweaking on the log wheels and primary bars and I’ve now got something with safe luma and chroma that I can start to grade!”

  • Tim Neighbors

    December 4, 2018 at 8:10 pm

    Interesting. Something else to experiment with.

    Just to follow up with this for the benefit of others, I was able to do a quick controlled test with different cameras in this same lighting. Some cameras clearly handle it much better than others. In fact for most of the cameras we were using, it wasn’t a problem. Only my A7s (first gen.) had this issue noticeably bad. I was able to improve it a bit in camera by picking a different “color mode” in the camera’s picture profile. I found “Movie” worked the best of the different color modes. Also, underexposing the shot helped. Reducing color saturation did not help.

    In post, using Vegas Video, I was able to fix the problem surprisingly well using the “saturation” plugin. I isolated the color of the unwanted patches on his face and pulled most of the saturation out of that color. Since that particular color didn’t exist elsewhere in the shot it didn’t affect the rest of the frame. With more work, one could even go further and try to isolate that color and then change it into the desired color by rotating it’s hue and reducing saturation.

    As of now, that is the best I can do with this. Maybe a better solution exists. …or maybe this is just a shortcoming of this particular CMOS sensor.

    I hope that helps someone.

    Tim Neighbors
    invisibleharness.com
    FB.com/invisibleharness
    instagram.com/invisibleharness

  • Steve Crow

    December 4, 2018 at 8:13 pm

    Oh, that’s a cool solution too – thanks for the follow up, it’s always nice to when people come back to the forum to post the outcome/solution – so often you never hear back.

    Steve Crow

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy