Tutorial

8 bits vs 10 bits vs 12 bits footage

 

So, pretty much everyone knows the theory behind 8 bits, 10 bits and 12 bits RAW footage, but what does that mean to us filmmakers? Does it actually matter? My answer is… it depends.

If you are new to this, understanding color depth is a requisite. So color depth is separated by channels, Red, Green, and Blue. So, 8 bits has 256 values per channel, which is 256*256*256 = 16 million colors. 10 bits has 1024 values per channel and 12 bits has 4096 values per channel. In theory, 12 bits has more tones and shades than 8 bits, and this is how you can stretch the color information without creating banding and degradation.

8bits vs 10bits colors

I did this test a few times in the pasts but camera technology is changing so much and having the Atomos Shogun on hand with the Sony FS5 right now calls for a redo. Things I learned is that not a single camera will wield the same results to this test, compression, and codec is also a big part of it, so your equipment may behave better or worst than what we have here.

So for the test, we have recorded the footage with the FS5 and Atomos Shogun. The 8 bits footage is the internal 4K XAVC, the 10 bits is in Prores and the 12 bits is in CinemaDNG.

Here is our starting point

8bits-untouched footage8 bits footage

10bits-untouched footage10 bits footage

12bits-untouched footage12 bits footage

 

Now, let’s throw a heavy curve on them to see which image will fall apart first.

8bits-touched footage8 bits footage

10bits-touched footage 10 bits footage

12bits-touched footage12 bits footage

With the heavy curve on, we can see the 8 bits footage becomes “blocky”.
As for the 10 bits footage, we can see there is a little bit more latitude less pixelization in the colors than 8 bits.
The 12 bits RAW footage, in this case, is smooth when it comes to color gradient transition. There is no noticeable banding or pixelization.

Regardless of the differences between the footage, I am surprised how the 8 bits footage is handled by the FS5, it holds up pretty well. From my previous tests a few years ago, I thought the gap between the 8 bits and 10 bits would have been bigger.

 

From those test, in my opinion, I don’t think that recording in 10bits or RAW is worth the hassle and the storage management it creates for most projects. Not that it is useless, far from it, I love grading raw footage, its make it for a breeze. And if you do have the equipment or a 10 bits native camera like a C300 it is a nice bonus to have, but 90% of the time, a great end result can be achieved at 8 bits.

Anyhow, I hope you liked the article and if you haven’t watched the video yet, I would suggest you do so as it shows this test a little bit more in depth.

Leave a Reply

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