Yes, I recommend 16-bit over 8-bit tiff if you can. If you wont be post processing the image, then 8-bit TIFF might give you the same result, but as soon as you start manipulating the file, the benefits of higher bit depths becomes important.
Yes, I recommend 16-bit over 8-bit tiff if you can. If you wont be post processing the image, then 8-bit TIFF might give you the same result, but as soon as you start manipulating the file, the benefits of higher bit depths becomes important.
Just like many others I really enjoyed this post, but I was wondering how Nikon compairs in this subject? I guess what im saying is if Canon is shooting 10 bit is Nikon really 14bit?
Noise is complicated.
There may be situations where the signal contains enough data that 14 bits is warranted, such as ISO-100 on bright days, maybe with long shutter times, I
On Flickr - Namethatnobodyelsetook on Flickr
R8 | R7 | 7DII | 10-18mm STM | 24-70mm f/4L | Sigma 35mm f/1.4 | 50mm f/1.8 | 85mm f/1.8 | 70-300mm f/4-5.6L | RF 100-500mm f/4-5-7.1L
Originally Posted by Matt.s.Maneri
This is kind of interesting.A few years ago, Nikon was only a little better on this issue, but with the release of the D3X and D7000, Nikon has pulled *way* ahead of Canon. Nikon is mopping the floor with Canon when it comes to dynamic range.
When Canon went from 12 bits to 14, they improved exactly nothing (in fact, some of their cameras got slightly worse). When Nikon went from 12 to 14, they did the same thing -- with one exception. Nikon knew that you're not supposed to waste an additional 2 bits unless you actually improve the camera to utilize those bits. But they didn't improve the camera, so they decided to implement a dirty hack: they forced the camera to operate four times slower to reduce noise (either by quadruple read-out averaging or lower ADC clockspeed). This improved the noise by about one stop (1 bit), which was about enough to fully utilize a 12-bit camera.
In other words:
* Canon had 12 bit cameras that only needed 11 bits (if you don't count pattern noise). Then Canon added 14 bit cameras that only need 11 bits.
* Nikon had 12 bit cameras that only needed 11 bits. Then Nikon added 14 bit cameras that only need 12 bits (in super slow mode).
Of course, both Nikon and Canon could have stayed with 12 bits and Nikon could have renamed their 14-bit feature as "reduce frame rate by 4X to get less noise", but I guess the Marketing department wanted 14-bit.
Now fast forward to the D3X and D7000. Here Nikon has actually built cameras that use *all* 14 bits of the 14-bit raw file. You can accidentally underexpose a D700 ISO 100 photo by *SIX* stops (or more), boost it back up 6 stops in post, and still get a nice looking image in post. If you tried the same thing on Canon the result would be total garbage.
<meta content="text/html; charset=utf-8" http-equiv="content-type" />
<meta content="text/html; charset=utf-8" http-equiv="content-type" />
As of the D3X and D7000: yes. Hopefully Canon will put up some competition in this area with their next camera.
<meta content="text/html; charset=utf-8" http-equiv="content-type" />
Could you explain that for the dummies out here? I don't see how processing speed changes anything... unless you are saying that by lowering computing speed the read noise goes down?
Umm, Daniel- can you remind me why it was I decided to go with Canon? []
Because of the detailed reviews and friendly atmosphere right here on TDP, wasn't it?
Jon
I bought a Canon because I have owned several Nikon Riflescopes over the years, and they were all crap. I figured if they couldn't make a rifle scope right how could they get a camera lens right. Call it product disloyalty. But once you invest so much in Canon it gets harder to change.
Daniel
Your comment about underexposing is the first comment in this thread that seems to make your point straight foward and direct. Most people want to see how somthing effects them in the real world. We could talk about bits all day, and they just as well be the kind horse's use if we don't see what effect it is going to have on us when we go out and take a picture. Myself I would thinkunder exposing six stops a phtograher deserves to loose his picture, but things happen I guess. What other ways might this affect us?
Rick
Larger image files than necessary, for one thing. But then, storage is cheap...
That one is true. Maybe if I only worked off the laptop and was space limited it would be annoying. As is its not a problem.
<meta content="text/html; charset=utf-8" http-equiv="content-type" />
<div>
Well, the observed behavior is that "14-bit" mode is four times slower and two times less noise. I have two theories about what they might be doing:
1. Read out the pixel nondestructively (i.e. read without reset) four times (instead of just once like normal) and average the four values together for less noise.
2. Operate the ADC at a slower clock speed, which I've read allows them to have less noise.
</div>
In addition to exposure latitude (being able to recover accidental underexposure or overexposure) and smaller file sizes, it also allows the photographer to use more dynamic range in a single image. For example, Canon (e.g. 60D) photographers that shoot in low light tungsten around ISO 3200 will often find some of the skin tones blown out. With the D7000, you could shoot ISO 200 and then boost to 3200 in post. The noise level will come out the same (actually, Nikon is a little lower right now), but Nikon will have *four* stops more highlight headroom -- so no blown out cheeks or foreheads.
A lot of film photographers relied on having 5 stops below middle gray and 5 stops above (film cinematographers today use 12-14 stops of dynamic range for low contrast hollywood films). Canon defaults to only 3 stops above. If we had 14 usable bits, then we could easily have 6 above and 6 below.
Posting Permissions
Reply With Quote