The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

LCC in mixed/strongly coloured light

I'm shooting architecture/interiors with the SK35 + P45, and I often end up in situations with areas of strongly coloured/mixed light (e.g. a tungsten interior with a window with bright daylight streaming in). After looking at a few threads here and on forum.phaseone, some people have argued for exposing your LCC plate with a consistently coloured light source -- for instance, firing a speedlight at low power at your LCC plate for your LCC exposure. I've been trying both exposing an LCC "conventionally" using room light and exposing an LCC with a speedlight, and I'm noticing that there are a few subtle differences--particularly, the corners seem to get a bit brighter on the speedlight-exposed LCC, and colour casts on white areas seem more technically accurate but possibly less pleasing.

Are there any good rules of thumb for what's likely to produce the best results?
 

Louis Novak

New member
How exactly are you using the LCC? Mixed light shouldn't matter when you are using the LCC plate to correct for color casts in the lens. If you are trying to color balance that is different than removing casts in the lens.
 
Not attempting to do colour balance--just correct for lens colour and falloff. Routine is: set up shot, dial in final shifts, expose final exposure, then put the LCC plate directly against the lens (or center filter if I'm using it), open up 1 stop (or two) and shoot the LCC frame.
 

weinlamm

Member
I think mixed light is not possible to correct via a normal LCC. This kinds of pictures I
- develope from Raw
- make a white-balance on the light which is in the room ( most pictures are shooten with "daylight" in the back )
- make a second picture, corrected the white balance on the problem-areas
- use this second picture as layer for the first
- then correct the strongest mistakes via hand

I think mixed light is a problem since ever. With film you never had a chance - with digital now you can use your software
 

torger

Active member
It will depend a bit on how the LCC algorithm is implemented, I'm assuming you are using capture one.

What the raw developer typically does with LCC correction (flatfield correction) is only to correct for deviations from the "average", so if your LCC shot is shot in yellowish or blueish light will make no difference, as only deviations are corrected, and they will be the same regardless of average color.

However when light sources are mixed it could be the case that the LCC card will not even out the light to a smooth average over the whole surface, which means that the algorithm can be confused, so it can see casts where it actually are only different light sources contributing. So for mathematically perfect results you should probably use a fixed light source so you can light the card with one uniform light.

However, in practice maybe it is better to let LCC algorithm also even the mixed lighting slightly? More pleasing results? Mixed lighting always look more extreme in a picture than in real life, so in theory it sounds like it could be a good idea to shoot LCC with the original light sources. I have little practical experience of these type of scenes myself so I don't know.

How to interpret mixed unnatural light sources (ie mixed white balance) is much a matter of taste as cameras are generally quite bad at reproducing what the eye/brain experience at the scene. Going for pleasing rather than accurate would be my bet.

Note that not a flatfield correction algorithm does not necessarily need to pick the average over the whole surface as reference color, it could pick just from the center as it is typically most neutral there, but I doubt C1 does that as it is supposed to work with shifted lenses. My guess is that it takes average over the whole image as reference.
 
Top