nikonians

Even though we ARE Nikon lovers,we are NOT affiliated with Nikon Corp. in any way.

Select your language:

How-to's

How Many “Bits” Do I Need - 8, 12, or 14?

Darrell Young (DigitalDarrell) on December 9, 2013


Keywords: bit, bit_depth, format, jpeg, tiff, raw, nef, color, channels

Most advanced Nikon cameras offer JPEG, TIFF, and RAW (NEF) storage. The argument over which bit level to choose generally goes along with which image format should one choose? The image format question is addressed in a different article found here.

However, this particular article does one simple thing. It discusses the math of image bit-depth in a simplified manner, as follows:

An 8-bit JPEG can capture over 16 million colors. However, it can only capture 256 unique colors per RGB channel. That’s 256 reds (R), 256 greens (G), and 256 blues (B). When you combine all three channels (all the shades you have) you will see this: 256 x 256 x 256 = 16,777,216 colors.

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.

A JPEG file can contain a high degree of color compression while still maintaining proper luminance levels so that the image looks great. It just cannot be modified over and over and resaved without degrading from compression losses.

I would shoot in 14-bit RAW mode and take advantage of all those potential color storage bits (for subjects that need it) and later make a conversion to JPEG from my vastly superior RAW container. It is especially important to do so if you plan on using data from the low-order bits (darker areas in the image) for maximum dynamic range. I prefer 14-bit RAW files for those times I have the time and need to post-process images, a must when I shoot my commercial work.

As to which is visually better, 8-, 12-, or 14-bit files, this will always be highly subjective because we all see differently. I am speaking about "potential" color in this article. If there are not 4.3 trillion shades available, such as when one is shooting something like a red balloon against a black wall, would it really make any difference?

However, if one is shooting a magnificent landscape with colors of all sorts and a huge dynamic range potential, the number of bits can make a big difference in the final look of the image.

20131209_085207_color_depth_jrp_5842.jpg

A 14-bit RAW to 8-bit JPEG conversion, in computer, seems nicer to me than a straight 8-bit JPEG out of the camera. Using as many bits as your camera can possibly record with will give your images an edge. When you decide to make a masterpiece JPEG from your RAW file, you will have more color levels to choose from and, therefore, less compression banding to deal with in the final image.

Moral of the story: Shoot and store your images as RAW files; find the time, extra disk space, a good calibrated monitor, patience and will to learn good post-processing technique. You will be rewarded with great detail in the shadows and exquisite tonal range, even if your final image is a JPEG file.

 

Keep on capturing time…
Darrell Young

 

 

(10 Votes)
Darrell Young Darrell Young (DigitalDarrell)

Founding Member of the Nikonians writer Guild. Author of most of the NikoniansPress books. Donor Ribbon awarded for his generous support to the Fundraising Campaign 2014

Knoxville, USA
Team, 5973 posts

11 comments

Dr. Patrick Buick (profpb) on January 8, 2014

As always, Darrel, a nice simple, clear explanation for what I have been doing but did not understand why. Thank you.

Alan Wong (hyphotographer) on December 23, 2013

I have a D7000 and always shoot in 14-bit RAW mode. Using in camera NEF processing with white balance, D-Lighting picture and exposure controls, my JPEG images look fine. Most of my RAW files are deleted to save storage space, I only keep some valuable shots in RAW format. My computer is too slow to process NEF files, taking some 5 minutes for a minor alteration, whereas D7000 takes seconds!

Richard Walliker (richardd300) on December 12, 2013

Thanks Darrell, a useful and informative assessment. I tend to use 14bit for subjects requiring access to the minutest recovery of shadow detail and access to a greater dynamic range, such as Landscapes and Architectural subjects. So, I totally agree with your comment " the number of bits can make a big difference in the final look of the image". I switch to 12bit for just about everything else.

John D. Roach (jdroach) on December 11, 2013

Fellow Ribbon awarded. John exhibits true Nikonian spirit by frequently posting images and requesting comments and critique, which he graciously accepts. He is an inspiration to all of us through constant improvement in his own work, keen observations and excellent commentary on images posted by others. Donor Ribbon. Awarded for his very generous support to the Fundraising Campaign 2014

Thanks for the lesson!

Paul F Austin (paustin) on December 11, 2013

Good comments all. I would add a couple of notes. Regarding RAW, there is no real need for proprietary file formats for "off the focal plane" image storage. JPEG-2000 includes lossless storage as one of the storage options. Lossless is just that. Even the LSBs are recoverable, unchanged. The use of an industry-standard storage format from front to back would not only ease Adobe (and others) task but allow the use of proven, bug-free open source software packages. As far as 14- vs 12-bit storage, a question arises regarding the noise floor of the read-out electronics string. The latest generation of sensors have superb noise performance but the industry noise characterization presents the results as a qualitative measure. I know from experience that designing a 12-bit data acquisition system that doesn't insert artifacts into an image is extraordinarily difficult, never mind a 14 bit one. 14-bit storage seems to give a more pleasing image but the monitors and especially printers have difficulty rendering images in a meaningful way with that kind of color depth. Finally, in processing, I installed a 512GB SSD as my C: drive, where speed critical applications and my files for editing reside. The improved performance is breath-taking. In a balanced system (a reasonably capable graphics card, CPU and adequate RAM), a large SSD is really worth it in post processing.

Todd ONeil (toneil14741) on December 11, 2013

I shoot in 14 bit raw. I have a side question and since I cannot get an answer anywhere else, I'll ask here. I use Nikon Capture NX2. I was hoping for NX 2 but that is obviously not coming. In your opinion(s), what is the best replacement especially since I do shoot with RAW. I do not like Photo-shop.

Linwood Ferguson (Ferguson) on December 10, 2013

I think the emphasis on the article on colors is a bit distracting, as people will rightly point out that humans can distinguish relatively few colors. It is really more about light intensity levels. What is really captured as each pixel in raw is just the intensity of light after passing through a color filter. Each extra bit stores twice the information. Then recall each F-stop is twice or half the light. So as a concept, think of each bit as being one more f-stop of dynamic range. Scientifically there's lots to take issue with in that statement (especially since 8 bit images are not linear post conversion), but as a conceptual framework to help new camera owners decide -- any modern camera can capture more than 8 stops dynamic range. Use 8 bits and you are throwing away either highlight or shadow detail. Use 12 and it depends on your sensor. And new sensors are starting to knock on the 14 bit door a bit also, though there is still some headroom.

N. Kent Lewis P. (nkcllewis) on December 10, 2013

Daryl, thanks. I have heard that in addition to capturing in RAW 14bit mode, to capture the maximum amount of information, one must seek to also overexpose slightly. This ensures maximum information possible is captured by the camera sensors which in turn provides needed information when opening up the shadow areas. Just wondering if you had heard this as well. Kent in VA

Eric Carlino (ecarlino) on December 9, 2013

a quick google search produced this fine article that does a great job explaining why more Bits are better: http://thegreyblog.blogspot.com/2011/10/tones-and-dynamic-range-why-you-should.html

Eric Carlino (ecarlino) on December 9, 2013

I always prefer more bits primarily to get more detail out of the shadows. The explanation of how light is mapped in RAW files is a bit involved, but Lloyd Chambers has several great write-ups and RawDigger shows it with their Histogram: http://www.rawdigger.com/usermanual/histograms

Stephen Blakesley (lajolla) on December 9, 2013

Thanks again for the very clear and important explanation. The increase in color info is the main reason I shoot at 14-bit with all Nikon DSLR's. I shoot NEF+JPEG originals, with the NEF menu setting at "uncompressed" on the d800 and "lossless compressed" with the d600. Glad to hear that the Df also offers the "uncompressed" menu shooting option. I also shoot the NEF+JPEG separate original pair with JPEG set to Large image area and JPEG Compression set to Optimal quality. While this leaves me with smaller buffer sizes, I normally do not shoot high-frame rate sporting events with NEF. In any case, I deal with huge files that are imported via Aperture to Mac platforms. If one has the memory, speed, and storage that the latest Mac OSX can provide, the large NEF images files are truly quickly processed (the latest vogue-word on the west coast is 'thrashed') into life - even by Aperture version 3.5.1 :) I shoot and import via Aperture both separate NEF and JPEG images - in order to compare the baked-in Nikon camera (and lens) JPEG effects with the NEF (RAW) that is untouched by Aperture. Hard to believe there still is such a proprietary, image- algorithm wilderness out there? I was told that both Adobe software folks and Apple software folks need to buy the latest Nikon DLSR's off-the-shelf and then reverse engineer their own RAW image processing algorithms for the proprietary NEF image :)

Back to top