8-bit vs 16-bit editing guide

The 8-bit vs 16-bit dilemma made easy.

Should I be working in 8 or 16-bit? I have been asked this question more than a few times by various people, who felt unsure, if they should be working in 8 or 16-bit mode. Should they develop always to 16-bit and then convert to 8-bit in Photoshop, if 16-bit of information is unnecessary for the image, or maybe always go with 8-bit? I had asked myself this question many times in the past, but only when I really sat down and thought about it, had all the ins and outs become obvious. So if you are one of those people read on!



Don’t get discouraged by these numbers in the beginning, in fact it’s all pretty simple.
In an 8 bit image (all jpgs for example) every pixel has 256 possible shades of Red (R), Green (G) and Blue (B), which we will refer to as 8-bit RGB. This totals to 16.8 million possible colors for combined RGB, whereas a 16-bit image has 65,536 possible shades in each channel, which gives us 281 trillion colors.

When we shoot in RAW our cameras usually capture 12-bit or 14-bit color. Most medium format backs work in 16-bit, which gives us a touch more data to work from while developing, yet the color fidelity difference compared to high-end DSLRs is much smaller than it used to be.

During development, amongst many other decisions, we have to pick weather to develop to 8-bit or 16-bit (JPG, TIF, PSD). Which should you pick? Well it depends…


What are the files for?

The destination of the output files in fact determines to what bit depth you should be developing.

A. Just to get this out of the way: If you are developing to the web, or a bunch of family snapshots for your loved ones they will surely be jpegs, which anyway have to be 8-bit.
B. You are developing TIFs for printing or to give directly to your client, who will edit them.
C. You are developing TIFs or PSDs for further editing in Photoshop or other post-production software.

As we see we have two alternatives which we should consider, B and C.

B. TIFs for clients
Let us break this down into two categories:

B1. You are letting the client post-produce, retouch the image before using it.
In other words you are sure, that it will be manipulated and changed. In this case ALWAYS give 16-bit files. Since you have no control over how the photographs will be changed, it is essential that you give 16-bit, which allows for much greater changes without showing visual degradation in tone transitions. Normally big changes should be made during development.
The only reason not to give 16-bit in this situation, is if you know that the client is very unprofessional. If so he may have trouble with the higher bit depth, just stick to 8-bit and ask them to show you the final image for approval.

B2. You are not letting the client edit the images in any way.
In this case 8-bit is recommended. If someone inexperienced will be preparing the images for web or print use, he might get lost when seeing the 16-bit image and not convert it properly, especially if he is using some cheap software.

Yet if you know your client, know that his prepress department is professional and your image might be difficult to output, then why not give them a 16-bit image. Sometimes prepress departments, for example in ad agencies, need to output to very different media and if your image contains lots of subtle gradients and they need to modify the image to look well in a desired color space, then 16-bit might save the day. We almost always give out 8-bit files at our retouching studio, but when dealing with a knowledgeable client you might as well ask. It will probably just improve your reputation.

C. You are developing TIFs for further editing
Again lets break this option down into parts, as there are many options to consider

C1. When you developed your files you set the colors, contrast like you wanted in Lightroom or Capture One and now are only cosmetic retouching or doing any other operations which are not altering the parameters of the image. I mean by this not placing a zillion adjustment layers, which are changing the look of the image considerably.

Ideally, if you develop to 8-bit, you should not be doing any adjustments to the image apart from cleaning, cloning, merging images together, etc. This way, if you leave the values of the pixels unchanged, an 8-bit file is enough for about any usage you might consider. From the web to the most demanding printing 8-bit serves it all. Its 16.8 million colors exceed about every media that there is anyway, maybe except for some very very high-end devices, for which you would anyway know you are outputting.

/ Note
Still if you have developed to 8-bit and are forced to make changes afterwards without the possibility of reprocessing, although you did not intend to and your image consists of smooth gradients, then switch to 16-bit while editing. This way all of your corrections will be calculated in 16-bit and that might be enough to not have any banding show up. The extra bit data is lost anyway yet adjustment layers applied in 16-bit mode have many times saved the day for us, although in the end for output we switched back to 8-bit again.

C2. You are developing TIFs and know that you will be heavily retouching the look of them. You guessed it, 16-bit.

C3. You are developing Tifs for heavy retouching, but know that the file size will be so big, that it will be at the limit of your computer system, yet you are afraid that working in 8-bit will introduce too many problems: banding, edge artifacts, posterization… In this case we usually recommend to develop to 16-bit, merge everything together and apply the colors in 16-bit and if your machine is too slow to work, then convert to 8-bit then. It’s always worth a try in such a case.

If this is what your computer looks like, please stick to 8-bit.

If this is what your computer looks like, please stick to 8-bit.

Otherwise when working with very large files, with multi image comps often there is no sense in sticking to 16-bit. Unless you really have a very powerful computer (SSD raid disks, lots of RAM, strong graphics card and an adequate processor are always helpful) 16-bit can just be too much for your counting machine when editing a multi GB file.

The cost

It might not be at first obvious, but working in 8-bit is cheaper. If your only objective is to achieve the highest possible quality and you don’t care about what it costs you, then skip to the next section, yet it must mean you either are only shooting for your own pleasure, or your client pays you way too much. Otherwise it’s good to be aware that 16-bit editing and archiving just costs us more.

You will spend more time on every step of the editing to archiving process; that’s a fact. If you are running the latest high end model of PC or Mac and do not develop your images too large and then do not make your psd files too complicated, then the time difference will only be marginal, if not, it might be drastic, on occasions you might expect more then twice the time spent at looking while your computer is counting or even overloading your machine to the point, where you will want to treat it with anything close at hand.

The space on your hard drives will also be eaten up twice as fast. If you archive your work psd files and work a lot, you will eat up terabytes much faster. Although the cost of a TB of data has fallen greatly in the past few years, it is still a cost worth considering, especially if you are properly backing up your data.

So which should I pick?

For me it comes down to two simple sentences:
Do all of the contrast and coloring work during development and develop to 8-bit.
If you intend to manipulate a bit more in Photoshop develop to 16-bit and work in it till the end, if your computer permits it.

End note

One day we will have the same discussion regarding 16 and 32-bit. People will be afraid that 16-bit might not be enough to achieve top quality and at the same time images of the masters shot on film will still hang on the walls of museums and we will quietly sigh while looking at their beauty. So don’t worry too much, we’re moving forward, or maybe not…


  1. […] This post has been also published on House of Retouching blog […]

Leave a Reply