Shot Anatomy: The Blood Orange

January 24, 2007 in fruit

In my previous post I mentioned that I would talk a bit about how I shot the blood orange image. I will talk about the set up and then the post processing today.

A note about my equipment for this shot:

  • Canon 30D
  • EF 100mm 2.8 USM macro lens on manual focus
  • manfrotto tripod
  • ambient light through a sheet of white sheer polyester fabric (bright sunny mid-day sun)
  • Various styrofoam blocks to use as structural light bouncers
  • small mirror (very clean)
  • blood oranges
  • plastic ice from Wal-Mart (They call them “table jewels”, I call them cheap and great ice models)
  • water
  • poster-sized white foamcore board (for movable light bouncing)
  • Photoshop CS2, PS Camera RAW, and Bridge
  • Flickr-based DxO modifier – get the software or a demo here

The thought process:

I didnt plan this out ahead of time like an art director might. My intent was to see how my photos come out after having re-read the manual. I have all sorts of odds and ends that I have bought just for food photos so I walk around the room and pick out things that seem right at the time. I only knew that I wanted to shoot the blood oranges. I ended up with the black tray as a background and to catch the water. The mirror was something new for me but it fit in with my mood for reflections and playing with depth of field. Lately, I have been unhappy with the way my 50mm was giving me so much soft very shallow DOF.

Here is an example of that.

I read the manual to get more precise control over DOF (something easy to lose if you let yourself shoot only in program mode).

What happened:

I mount the camera to the tripod. In this case I have it flipped over into portrait orientation (see shot above).

I always shoot in RAW.

I shot a 18% gray card on manual focus, auto white balance and spot metering. In the menu, I toggled to custom WB and hit “set”. The gray card image is then used as a reference. I get out of the menu and set the white balance to the custom setting. I set the metering to evaluative (perhaps there is a better method?) and keep the manual focus.

After setting up the shot and dripping additional juice onto the half you see here from the other half of orange, I shot a variety of aperture settings.

I had the ISO set to 100 (exposure time 1/100 sec) . The EXIF data on this blood orange image says I ended up with f/6.3.

I used the RGB histograms in the informative display in my camera LCD to determine how the camera is capturing the image. The little image on your LCD is not a good way to evaluate whether you are getting good colors or if you are blowing or over-saturating any highlights or colors. In this mode you can see little red boxes in the image preview that suggest where the focal plane is, thats helpful.

This is from a different shot that day but shows what I am talking about.

Once I have bracketed the f-stops all around what looks like the best setting its time to download and process.

I opened Photoshop CS2 and Bridge and then navigated to the best looking RAWs in the preview.

I clicked on it in Bridge and it opened it in PS using Camera RAW. There I set all the autosettings to off and set the color noise correction and sharpening to zero. (My camera has a zero sharpening set and also one stop down in saturation, -1). I futz with the exposure and other settings in Camera RAW. See the following image for those settings.

I saved this file as 8-bit (thinking of going to 16 bit) TIF file with an AdobeRGB color space. I closed this and went to Bridge and opened the TIFF folder that I saved this image in and clicked on it to open it in PS CS2.

I have an action that I execute at the beginning to speed things up and to standardize what I do. It can change and does as I learn more.

This action does the following after I hit Control-F2 (user defined):

Make a –

  1. background copy
  2. second background copy
  3. Curves adjustment layer
  4. Threshold adjustment layer
  5. Color balance layer
  6. “dodge” adjustment layer that is set as “overlay” blending mode and “50% gray”

Here is what the desktop looked like when I hit this first action:

I hid the threshold layer to see if I needed to crop the image at all. If I did (though I did not in this case), I would have selected the first background copy layer and cropped. I then unhid the Threshold layer and double clicked the Threshold icon. I moved the slider to the left (as shown above) to get the blackest black on the image.

Getting a handle on white balance in PS CS2

I selected the color sampler tool (see below) and clicked in the section that turns black in the Threshold screen the earliest (furthest left on the slider).

When you click with this tool on your image you get a little bulls eye sort of icon.

You then do the same for the earliest appearing white zones in the Threshold view.

After I selected the white point, I threw the Threshold layer in the trash. I always double check that I am not selecting a blown out highlight like reflections in the ice. Here I chose the white of some of the center of the orange.

I then opened the Curves layer and clicked the little black eyedropper below the curve plot. With that selected, I clicked inside of the black-specific bulls eye in the image. I did the same with the white eyedropper. If you get funky color casts using these eyedropper selections, go back and make new color picker selections that give you a better balance.

In this image you can see I have selected the black and the white. It makes some in the image.

Color balancing

Next you can use the color balance layer to modify the various colors in your image. In this example I did two things – I created two different color balance layers and I worked on two different colors in those two different layers. I also used the masking option in each of those layers to selectively remove some of the changes I made so that only certain parts of the image would have a color change.

After you have done your color balancing, you can choose to lighten or darken parts of the image with the dodge layer. Select that layer and select a brush (choose the opacity and brush type according to your preference)

I lightened up the white pith a bit.

Cloning away problems

You can also remove imperfections. I named a new adjustment layer “spot” and selected the clone tool. I chose a desirable part of the image and hit alt click to pick up image to clone with. Below is a before and after shot to show you what I changed with the clone tool.



I removed the annoying background stuff.


The next thing I did was to sharpen by a method I learned as a tip over at the Adobe Photoshop site (PDF). I went back to the third background layer (third from the bottom) and changed it to “overlay” (see image below).

And then I went to Filter>Other>High pass.

This brought up the following window:

As you slide the bar at the bottom, the “radius” changes. Less sharpening on the left, more to the right. You want to be sure to not oversharpen as it increases noise. I did not do a whole lot of sharpening in this case. There seems to be a million different ways of sharpening. This is my way, for now.

I saved the file as a TIF (I actually did all along the way so if my toddler pulled the plug I would not lose much) with a suffix “-post” to let me know its not the TIF I saved right from the RAW file. Save the RAW, the original TIF and this -post TIF in a safe place.

I applied copyright info in the EXIF data by going to File>File info and entered that data.

I saved again. You can now save as this large size as a JPG if you want a huge JPG file or you can reduce the size and save as JPG for your blog, etc. Before you save as a JPG make sure that you change this AdobeRGB colorspace to sRGB (AdobeRGB renders partially on the web and looks washed out – crappy).

To do this I went to Edit>Convert to Profile.

You will see this window:

I clicked “ok” and then saved as JPG.

Thus, the photo has changed from:

Finally, I did an additional modification to this photo.

I used a Flickr toy called DxO Lighting that allows you to upload (or use an image uploaded into flickr previously) an image and run the DxO software on it.

Before DxO and after.

Whew, that was a long one. I hope you learned something from this tutorial of sorts.