I’ve been conducting some custom white balance color tests with my camera. I have been making it up as I go, and now want to reshoot al the tests with new color cards, but I thought I would go ahead and link up what I have done so far in case anyone is interested.
Here’s what I have done so far:
First, I painted up some random color cards for custom white balance tests. I didn’t really have a set idea of what colors I would paint here, except that I would do some warm (redish or yellowish) and some cool (blueish). I also used a few color cards I had inherited. My first test shots came out interesting (to me at least):
After I shot all the pictures, I used opened the raw images in lightroom and looked at the temp for each of them. As I noted in a previous post, I was surprised when I learned that two of them had the same temp (k value) but were different colors. this is because the camera also sets the tint of the photo.
So, I decided to repeat my experiment, but with some different color cards. At this point, I made a mistake – I painted up some color cards that were based on what I’d learned to be the color wheel: Red, Orange, Yellow, Green, Blue, Purple. I didn’t even do that good a job painting them – I’d have failed this assignment if it was art school. Besides each of the 6 colors, I painted a light version and a dark version of each.
I then shot a still life using the cards to set a custom white balance, under 4 different types of light:
- Cloudy Daylight: This was next to a large open window with cloudy, 3pm daylight outside.
- Alzo Daylight Bulbs: Main lighting was two 100 watt equiv bulbs (4 in room total).
- Cheap daylight lamp: I didn’t set this op up correctly, so I consider it a failure. oops.
- Incandescent: A single 60w incandescent bulb in a desk lamp.
After I shot each of these, I used lightroom to find the temp and k values of each of the photos. I also white balanced the color cards according to a white card. (I sure hope this is making sense to someone else.)
I also shot a set of photos color shifted using my camera’s white balance bracket / color shift feature, and took a photo of the screen of each of the settings. The camera was set to daylight for each of the shots, and I used the K and tint values to fake a shot for each of them by applying the camera’s readouts to a color balanced still life shot.
SO, after all that, I put together some test pages for each of the types of light where I can sort by temp or k values:
- Cloudy Daylight
- Alzo Daylight Bulbs
- Cheap Daylight Lamp
- Incandescent
- Camera Color shift
- All still lifes compared
- All color cards compared
I plan to reshoot the photos based on color cards pulled from the color shift shots, but am not sure when that will be since my free trial of lightroom ran out and I’m not going to buy it right away. (If anyone knows of a free program which will tell me the temp/tint values, let me know – The Canon camera raw software won’t, and my copy of photoshop is too old to open my camera’s raw files.)
What have I learned? Nothing definitive, but a few things I find interesting.
- The color shift pictures from the camera follow a pattern when sorted by temp and tint (this may make more sense if you look at the Camera Color shift experiment):
- I really need to rethink color if I am going to work digitally. It’s hard to unlearn/relearn everything you know about something, but I need to find a way.
- Temp seems to be a bit more consistently applied by the camera than tint, which can vary wildly.
- Using color cards seems like a quicker method of getting a desired color shift than setting a color shift in the camera, provided you already have some cards made up.
- I’m probably doing these experiments as a way to procrastinate applying principles, i.e. actually getting out there and taking photos.
- I am not really sure what to make of this all, but I am fascinated with the results.
My next step is to build a spectrograph and go back and create a graph for each of the light sources I tried (I won’t be able to get exact on the sunlight, but I hope to get close.) I am hoping this graph in addition to the photos will give me a better picture of the effect of different lights on photos. But, I fear I need to reshoot the photos with a better set of color cards first.