Rss

Best of the Tetons

Working in 16 Bit Mode

Memory is Cheap — Memories are Priceless!

I typically shoot in 14 bit and process in 16 bit in Photoshop as long as I can. Here’s why…

16 Bit Clobber and Recovery

The issue is not what you can see, or what your monitor can display, or what your printer can print—but what is under the hood of the file! I believe you will be amazed by the examples! (For this article, 8 bit vs 16 bit refers to Color Bit Depth while using Lightroom and Photoshop.)

14 Bit Capture 16 Bit Image

16 Bit Export

The image above was captured with a Nikon D810 in 14 bit mode. I set that in the camera’s menus long ago and never looked back! The files are much larger, so they fill cards faster, fill the buffer quicker, and possibly slow down the frame rate on some cameras. You might consider these issues up front. You can always “downgrade” a capture during your workflow, but you can’t “upgrade” one. As seen in the screen grab, I export images from Lightroom to Photoshop by selecting the 16 bits/component option.

5 percent Output

For this example, I am going to CLOBBER the image by bringing the output levels to 5, darkening it to almost black. I’ll be doing the same commands to an 8 bit image in a minute.

16 bit at 5 Output

It doesn’t look like there is much data here. Now comes the magic!

Adjust Input to 5 percent

I adjusted the Levels on the “black image” by dragging the white slider in the Input Levels to 5 and hit OK.

Results After Input

Wow! It looks darned good! The tiny bit of data in the “clobbered” image’s histogram has been “stretched” to fill the entire histogram.

5 percent output

If I open the Levels box again, I still have a good histogram. Personally, I think this is amazing! Now watch what happens with an 8 bit file.

8 Bit Clobber and Recovery

8 bit File

8 bit Export

This is the same 14 bit image, exported this time at 8 bits. (You won’t be able to see any differences in an 8 bit JPG web image).

5 percent Output

Just as before, I dragged the Output Levels to 5, which darkened the entire image.

8 bit at 5 Output

Same 8 bit image at 5 Output. It looks that same as the first 16 bit image.

5 percent Output

To recover the dark image, I opened the Input Levels adjustment tool and dragged the white slider all the way to the left to 5 and hit the OK button.

8 bit File

This shows the results of the Input slider at 5.

Results on 8 Bit File

The Histogram on this recovered 8 bit file looks terrible and the sky is severely posterized. This 8 bit file NEVER had the same amount of data that was stored in the 16 bit version.

The Real World

Underexposed Image

No one would ever “clobber” an image like I did in the two previous examples, but everyone takes an underexposed image once in a while…or often. It might look a lot like the one above.

Levels Adjustment

The common, and easy fix, is to drag the white slider in the Input Levels to the left until it touches the right edge of the histogram. (Note: you can do this with Curves, too)

Adjusted Image

After the quick Levels adjustment, the underexposed image looks correct again, and if you followed along with the previous examples, it’s histogram will have been slightly “stretched”, but still have plenty of data if I had been working with a 16 bit file.

Should you shoot in 12 bit or 14 bit for your raw files? This debate can get hot and hostile!

First, let me grab a quote from John Sherman at Photography Life. “12-bit image files can store up to 68 billion different shades of color. 14-bit image files store up to 4 trillion shades. That’s an enormous difference, so shouldn’t we always choose 14-bit when shooting RAW? Here’s a landscape I snapped, then found out later I had shot it in 12-bit RAW. Better toss this one out, right?”

14-bit vs 12-bit RAW – Can You Tell The Difference?

John’s conclusions seem to indicate he couldn’t find much difference. I found sites that landed on both sides of the answer.

The Wrap Up

Each day, there are countless numbers of photos taken with point and shoot cameras, iPhones, and even professional level camera bodies in (8 bit) JPG format. Some shoot in RAW + JPG, giving them quick access to their process JPG files and full control later with RAW files. JPGs are small in file size, and in most cases have been processed using the camera’s software for hue,  saturation, contrast and sharpening. But, they are almost always JPGs with 8 bits.

I seldom shoot anything but 14 bit RAW files—knowing I can convert them to JPGs at any time. I also know I have options to export my 14 bit captures to 8 or 16 bit images. Again, I can downgrade the files but not upgrade them after capture. Also consider…at this moment in time, some people cannot see the difference in files shot at 12 bit and 14 bit, but technology is changing fast! I sometimes look at some of my oldest images captured as JPGs and wish I had taken them as RAW files. I have a tendency to think of them as primitive images, or not up to par. A few years from now, I might think the same thing about 12 bit RAW captures?

While I process a lot of the images seen on Best of the Tetons with only Lightroom adjustments, my “important” images always go through Photoshop, even if the heavy lifting was executed in Lightroom. Photoshop is definitely a player if I plan on adding textures or compositing images.

Many people work in 16 bit in Photoshop “as long as they can” before being forced to work in 8 bit. Sometimes, converting to 8 bit is the very last step when a web sized JPG image is required. Other times, third party filters require the conversion to 8 bit. I watched a web cast a while back featuring many of the Photoshop filter packages that were available for purchase. Throughout the web cast, web audience members were writing in to find out if the filters were 16 bit. Many are 16 bit of course, but unfortunately not all of them! Interestingly, even with Adobe Photoshop CC 2017’s Filter Gallery functions only on 8 bit files.

I definitely shoot in 14 bit for sunrise, sunset, and night shots. In other words, the tough captures may need the benefit of all of the data, while captures in bright, sunny days may not. I might switch back to 12 bit if it looked like I was running out of space on my memory cards. Memory cards are relatively cheap and hard drives are constantly getting cheaper, so staying in 14 bit when I can makes sense to me. My current lineup of cameras don’t seem to be adversely affected by shooting in 14 bit, but my older D300 dropped from 7 FPS to 2.5 FPS when shooting in 14 bit. Your camera may or may not have similar issues.

Out of curiosity, I dug through my Lightroom Catalog and found a similar photo to the example image, shot in 2011 using a Nikon D300 in 12 bit. I put that shot through the same Photoshop routine as the example image. That file looked fine—without gaps in the histogram nor did it have a posterized sky. Worth mentioning!

Lastly, you can do Internet searches for “14 bit vs 12 bit” and “8 bit vs 16 bit”. If someone tells you that 14 bit files can contain 4 trillion “shades” of color and 12 bit files can contain “only” 68 billion colors, you might be impressed, or not?

12-bit RAW file can contain 4096 colors per RGB channel or over 68 billion colors. When you combine the RGB channels you have: 4096 x 4096 x 4096 = 68,719,476,736 colors.

14-bit RAW file can contain 16,385 colors per RGB channel or over 4 trillion shades. When you combine the RGB channels you have: 16385 x 16385 x 16385 = 4,398,851,866,625 colors.

Impressed by the math?  If not, scroll back to the top of this page and go through the images again!  Think of the differences as “potential” color data, along with the possibility of more dynamic range, less banding, and more detail in the darkest part of the image. Maybe you will see it? Maybe you will need it? For me, for now, it’s only a matter if file size and storage costs.

If you like Charts and Graphs, check out this site. And, check out Thom Hogan’s page on RAW Conversions.

MJ

20th Time Results

July 27th Update: I was impressed with my original test image after cutting the output levels to 5 and then adjusting the input levels to 5 to fill the histogram. Out of curiosity, I put the results of that “clobbering” through the test again. It still looked good. I ran that file through the routine and it looked good. The image above is #20! There’s a little loss of detail in the clouds, but overall the image still looks pretty good. After doing a screen grab of #20, I reverted the file to its original, then ran the test again, but this time setting the output level to 1 (the scale goes from 1 to 255). When adjusting the input levels, I received a notice the entry had to be between 2 and 255, so I reverted the image and set the output setting to 2 and then the input slider to 2. I copied the results to the clipboard, reverted the file, and then pasted the clipboard contents on top of the original image. I could turn off the visibility of the pasted layer and never see the difference when flipping back and forth.


 

Teton Photo Excursions

For inquiries, send an email to info@tetonimages.com.

Like This Post? Share It

Comments (2)

  1. Thanks for this, Mike. Really good explanation with clear examples. I’ve always shot in 14-bit RAW simply because it just makes sense to me to always start with as much data as possible but I’d never really sat down and worked through the math. Well done.

  2. Stephen M

    Thanks for the explanation!

Leave a Reply

Your email address will not be published. Required fields are marked *