In this article, I'm going to talk about bit-depth and why I use 14-bit files, also known as "shooting in RAW." If you aren't into the technical stuff, skip to the pretty pictures at the end!
One of my major goals as a wedding photographer is this: no matter how bad/weird/unexpected the lighting is, I will always produce great photos. Probably the most valuable way to accomplish this is to understand bit-depth, and to understand bit-depth, you have to know a little bit about camera sensors.
The pixels on your camera's sensor are only capable of recording a numeric value, like 1, 2, 128, etc. The value that a pixel records is based on the intensity of light that hits that pixel.
That *number* of possible values is referred to as "bit depth." JPEG files have a maximum bit depth of 8, or 2^8 (two to the eighth power, or 256). This means pixels in a JPEG image can have 1 of 256 different values* (expressed as 0-255). The darkest value is 0 and the brightest is 255.
*really meaning 256 possible values for each channel (there are three channels: red/green/blue, or RGB for your monitor to display). But let's not get into the complicated way that we make color images for a sec, okay?
Each pixel in my D700, however, can record 16,384 possible values! (a bit-depth of 14). As you might imagine, having all that extra information means that filesizes are going to be much larger, like 10x larger for a 14-bit file.
Now, basically every monitor you use, whether it's on your phone, tablet, or pc, displays images in a bit depth of 8, just like JPEGs. So why in the heck do I waste my time and energy shooting in RAW?
Because of this picture that I took of Stephanie and Josh last year:
One of my major goals as a wedding photographer is this: no matter how bad/weird/unexpected the lighting is, I will always produce great photos. Probably the most valuable way to accomplish this is to understand bit-depth, and to understand bit-depth, you have to know a little bit about camera sensors.
The pixels on your camera's sensor are only capable of recording a numeric value, like 1, 2, 128, etc. The value that a pixel records is based on the intensity of light that hits that pixel.
That *number* of possible values is referred to as "bit depth." JPEG files have a maximum bit depth of 8, or 2^8 (two to the eighth power, or 256). This means pixels in a JPEG image can have 1 of 256 different values* (expressed as 0-255). The darkest value is 0 and the brightest is 255.
*really meaning 256 possible values for each channel (there are three channels: red/green/blue, or RGB for your monitor to display). But let's not get into the complicated way that we make color images for a sec, okay?
Each pixel in my D700, however, can record 16,384 possible values! (a bit-depth of 14). As you might imagine, having all that extra information means that filesizes are going to be much larger, like 10x larger for a 14-bit file.
Now, basically every monitor you use, whether it's on your phone, tablet, or pc, displays images in a bit depth of 8, just like JPEGs. So why in the heck do I waste my time and energy shooting in RAW?
Because of this picture that I took of Stephanie and Josh last year:
That is how the picture looked straight out of camera. It has that cool light cast effect that is in vogue right now, but you basically cannot distinguish any of Stephanie or Josh's features due to the incredibly low contrast.
Had I been shooting in JPEG with a bit-depth of 8, my pixels would've only stored a very narrow range of values. The lightest spots in this photo would've been around 255 and the darkest, say 206 (that's 50 shades of grey! *rimshot*). Trying to regain detail by adding contrast to an 8-bit image is an exercise in futility:
Had I been shooting in JPEG with a bit-depth of 8, my pixels would've only stored a very narrow range of values. The lightest spots in this photo would've been around 255 and the darkest, say 206 (that's 50 shades of grey! *rimshot*). Trying to regain detail by adding contrast to an 8-bit image is an exercise in futility:
Yeah, you can start to see some detail, but it looks like this photo has been abused by every Instagram filter in the book, and it is noisy beyond repair. That's because I've stretched a very narrow range of values across the full range of displayable values.
Now, because I used a bit-depth of 14, I captured a much wider range of values. Instead of only 50 values (206-255), I captured 3200 values:
Now, because I used a bit-depth of 14, I captured a much wider range of values. Instead of only 50 values (206-255), I captured 3200 values:
That extra information (64x the amount in an 8-bit image) allowed me to take the values which are supposed to be dark and actually make them dark--sort of like 'stretching' the image to occupy a wider range of dark and light values, thus revealing all the detail a 14-bit image can contain. There is a real art to editing in this way, and I have spent thousands of hours fine-tuning my skill in this area (I also use a special monitor that can display 10 bits of color information per channel). Here is a slideshow of the original in-camera image, my original edit, and then a recent "remaster" I made after tacking on another year's experience of editing 14-bit images.
So, to sum up: 14-bit images are awesome, and they're a great reason to trust an experienced professional to take your pictures! If you have any questions, leave me a comment!
RSS Feed