Exercise- Sensor Linear Capture

Take a TIFF or JPEG image and open it in Photoshop. Convert it to 16 bits per channel because you will be making some strong adjustments that might create some banding in regular 8-bit. Make a curve that looks like the example. You will need to create several points along the curve to keep it smooth. The resulting image is close to what the photo looked like as it was captured- before the camera’s processor got to work on it. Open this image along with the original and compare the histograms for each. How do they differ?

This was a really interesting exercise which showed how a digital camera’s sensor captures an image. Digital sensors respond to light in a very different way to the human eye and also photographic film- which as opposed to digital cameras, are non-linear. When the camera captures the light for an image, it is treated as linear which results in most of the captured light being towards the left of a histogram to the darker areas. To make up for this, the camera applies a gamma correction to lighten the image by dramatically boosting the highlights as well as the midtones. This then works in shifting the histogram to the right- which results in an image closer to what the human eye sees.

I have chosen to use the image below from a previous assignment. I loved this image, so I was therefore really interested to see how this specific photo would have looked prior to the camera’s gamma correction process.

Original Image Histogram

I applied a similar curve to the one shown in our handbook in order to bring the image back to what it would have been like prior to the camera applying the gamma correction. This resulted in a very dark image- with the white background appearing blue and pink! The histogram for this image is now mostly bunched up to the left, with very few values to the right, as seen below.

Dark image histogram

Most of the levels in the histogram above are used in the brighter part of the image (to the right of the histogram). This results in there being much less available for the darker tones to work with (all scrunched up to the left of the histogram). This then introduces artificial effects such as noise.

I then applied another curve to my darker image in order to try and revert it to it’s original state. This worked well, however, it has introduced additional noise in the shadow areas because it tries to lighten these areas so strongly. Below is my image brought back to as close to the original as possible. The shadows in the histogram are now noticibly reduced, and therefore this introduces unwanted effects.

Image returned to normal

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s