Reading mode: Light Dark

The truth about 14-bit

Go to last post
Amin Sabet

Amin Sabet

Veteran Member
Messages
6,781
Reaction score
182
Location
Boston, US
Canon says this:

"Adding to the improved virtuosity of the images captured by the EOS 40D SLR is the camera's 14-bit Analog-to-Digital (A/D) conversion process. Able to recognize 16,384 colors per channel (four times the number of colors recognized by the EOS 30D SLR's 12-bit conversion capability), the EOS 40D camera is able to produce images with finer and more accurate gradations of tones and colors."

Some here say that 14-bit A/D is worthless. Others seem to think this 14-bit business is the cure for all our problems. No one seems to have any evidence besides what they have "reasoned" in theory. Basically the usual blah blah. Does anyone really know? Anyone have enough experience with the 1D Mk III to whether tonal transitions hold up better, particularly after manipulation?

--
http://aminphoto.blogspot.com
 
Amin Sabet said:
Anyone have
enough experience with the 1D Mk III to whether tonal transitions
hold up better, particularly after manipulation?
Click to expand...
No, but some of us have seen the results of 14 bit conversions on medium format backs and are hopeful.

Brian A.
 
Here is a randomly generated texture (clouds filter in CS3) in 8-bit grayscale, inverted to put the texture in shadows rather than highlights. I removed the last significant bit on one of two copies (7-bit signal vs 8-bit signal) and added gaussian noise with a std dev of two (random noise in the last two bits) to each image. Which image has had the least significant bit removed?





I still found it hard to tell a difference after applying a strong contrast curve. Unless the dynamic range exceeds 11 stops or so, I doubt the extra two bits is going to make any difference whatsoever. You would do just as well taking a 12-bit image and adding two randomly generated bits.
--
emil
--



http://theory.uchicago.edu/~ejm/pix/20d/
 
It may not always be perceptible, but it does make a difference. Case in point - back in the day when I used to have to scan a black and white photo - conventional wisdom would have said to scan in greyscale - but I scanned in color - even though the photo was black and white. The color scan always produced better gradiation of the greys and thus noticeably more natural looking images. When the image was converted to greyscale, it looked no different than when it was RGB, but if it was scanned in greyscale to begin with and compared to the color scan you'd never scan in greyscale again.
 
Dynamic range (in digital) has two components -- range, and resolution. It sounds like the new sensor has the same range as the old one but better resolution. That makes sense if they've just put in a 14-bit ADC. Increasing the range would require some bigger changes to the sensor itself.

If so the extra dynamic range won't help out if you've got a scene with a lot of difference between the shadows and highlights (what most people think of when you say dynamic range, because they're used to the analog definition). You'll notice a difference if you have a scene with not much tone variation that you're cranking the contrast or doing similar processing on. Any manipulation that will give you gaps in the histogram on an 8-bit JPEG will do the same on a 12-bit RAW if you push it hard enough. The extra two bits in the 14-bit RAW will let you push that little bit harder.

--
Robb

 
It's not if you can tell the difference between unmodied images, it's the extra tones available for editing. The 14-bit option gives us 16K tones while the 12-bit gives us 4K. It is hard to belive that the entire extra 12K tones of 14-bit will get lost in noise. I expect the histogram to hold up much better under editing 14-bit
 
I think this is a misunderstanding. The 14 bit is not in post processing, but in ADC readout. What you're doing is manipulating some existing images.
 
Peter Rowe said:
It's not if you can tell the difference between unmodied images, it's
the extra tones available for editing. The 14-bit option gives us 16K
tones while the 12-bit gives us 4K. It is hard to belive that the
entire extra 12K tones of 14-bit will get lost in noise. I expect the
histogram to hold up much better under editing 14-bit
Click to expand...
Both an image recorded in 14-bit and one recorded in 12-bit
color depth get copied into CS3 and edited with 16-bit color depth
(or perhaps 15-bit; I seem to recall the last bit of data is a parity bit
for Photoshop?); anyway, CS3 will treat them the same way.
The question is whether there
is any extra info in the two bits for CS3 to play with, and I think the
answer is no. Though I am open to anyone showing how
to edit these two samples in a way which will show which
one has the extra bit of real image info under the noise.

--
emil
--



http://theory.uchicago.edu/~ejm/pix/20d/
 
exactly !

even with 12b RAW we have plenty of informations to play with exposition correction and levels, so for 99% usuall photos will 14b RAW have no visible effect
 
herr_bob said:
I think this is a misunderstanding. The 14 bit is not in post
processing, but in ADC readout. What you're doing is manipulating
some existing images.
Click to expand...
Yes, to demonstrate that if you (1) take the same signal, (2) convert
to digital using two different bit depths, (3) add noise to the mix
whose amplitude makes the last few bits pure noise; that the result
is indistinguishable between the two bit depths. You might as well
try to read the message that is encoded in flipping coins.

The demonstration gives an idea of the difference in image quality
between one of the new cameras as it is being produced, and
what the images would look like with a 12-bit ADC. That is
what my removing the least significant bit by hand is simulating.
Adding the noise is meant as a crude approximation of all the
sources of noise that get added to the image signal in the
course of recording the image in the camera. The result is
indistinguishable between the two treatments of the signal
as far as I can tell.

I'm not the one trying to bring post processing latitude into
the discussion; others are trying to claim that 14 bit images
will magically allow greater PP latitude. That is a red herring IMO.

--
emil
--



http://theory.uchicago.edu/~ejm/pix/20d/
 
Most probably it will not make a difference.

I happen to design integrated circuits for this industry. Although my designs currently are used in high speed color copiers , the technology is the same. In fact my last product was specifically for CMOS sensors.

The marketing guys have a term...."marketing bits" Typically these A/D designs do not have true 14 bit performance, but are usually closer to 12 bits best case and output an extra two noisy least significant bits. Whether you are talking SNR, or linearity of the design the 14 bit A/D will most liky be about 72dB SINAD or about 12 bits.

I also think if you look at the sensor, chances are good that the sensor may have noise in the same range. Since I cannot speak specifically to this sensor, I could be wrong.

In a normally exposed image, no difference...marketing bits. I would not use this "feature" as a reason to buy. I will be interested to see the detailed noise measurements on this camera.

Ron
Amin Sabet said:
Canon says this:

"Adding to the improved virtuosity of the images captured by the EOS
40D SLR is the camera's 14-bit Analog-to-Digital (A/D) conversion
process. Able to recognize 16,384 colors per channel (four times the
number of colors recognized by the EOS 30D SLR's 12-bit conversion
capability), the EOS 40D camera is able to produce images with finer
and more accurate gradations of tones and colors."

Some here say that 14-bit A/D is worthless. Others seem to think
this 14-bit business is the cure for all our problems. No one seems
to have any evidence besides what they have "reasoned" in theory.
Basically the usual blah blah. Does anyone really know? Anyone have
enough experience with the 1D Mk III to whether tonal transitions
hold up better, particularly after manipulation?

--
http://aminphoto.blogspot.com
Click to expand...
 
shot identical images using the same lens to compare a 1DmkII (12-bits) to a 1DmkIII (14-bits). So far more than a few folks that have looked at the two pictures have NOT been able to see "any" advantage with with 14-bit file.

I suggest you read the entire discussion to understand why a polarizer was used and the thought behind a reduction in Mach banding should have occurred with two additional bits of resolution. Here is a link (note, registration is required at this forum to see any of the threads).
http://www.1dseries.com/1d/index.php?topic=139.0

On page "2" of the thread look for the post by "Fredman"; he provides links to two RAW files, one from the 1DmkII and the other from the 1DmkIII.

Regards,

Joe Kurkjian, Pbase Supporter

http://www.pbase.com/jkurkjia/original



SEARCHING FOR A BETTER SELF PORTRAIT
 
maybe the old 12bit ones also were a bit dodgy in the last bit or two? then wouldn't this be better since we'd at least get a perfect 12bits, even not 14?
wsucoug said:
I happen to design integrated circuits for this industry. Although my
designs currently are used in high speed color copiers , the
technology is the same. In fact my last product was specifically for
CMOS sensors.

The marketing guys have a term...."marketing bits" Typically these
A/D designs do not have true 14 bit performance, but are usually
closer to 12 bits best case and output an extra two noisy least
significant bits. Whether you are talking SNR, or linearity of the
design the 14 bit A/D will most liky be about 72dB SINAD or about 12
bits.

I also think if you look at the sensor, chances are good that the
sensor may have noise in the same range. Since I cannot speak
specifically to this sensor, I could be wrong.

In a normally exposed image, no difference...marketing bits. I would
not use this "feature" as a reason to buy. I will be interested to
see the detailed noise measurements on this camera.

Ron
Amin Sabet said:
Canon says this:

"Adding to the improved virtuosity of the images captured by the EOS
40D SLR is the camera's 14-bit Analog-to-Digital (A/D) conversion
process. Able to recognize 16,384 colors per channel (four times the
number of colors recognized by the EOS 30D SLR's 12-bit conversion
capability), the EOS 40D camera is able to produce images with finer
and more accurate gradations of tones and colors."

Some here say that 14-bit A/D is worthless. Others seem to think
this 14-bit business is the cure for all our problems. No one seems
to have any evidence besides what they have "reasoned" in theory.
Basically the usual blah blah. Does anyone really know? Anyone have
enough experience with the 1D Mk III to whether tonal transitions
hold up better, particularly after manipulation?

--
http://aminphoto.blogspot.com
Click to expand...
Click to expand...
 
A 10MP image from a Rebel Xti, a 40D and a 1DmkIII would look pretty similar, especially if they used equivalent lenses. But the reason to get the 40D over the Rebel, or the mkIII over the 40D is the extra features that allow the photographer to push their photography to the limit, from extra fast frame rates, to rough weather protection.

Similarly, the 14bit A/D converter might not seem to affect the image much at first glance, but the ability to get greater tonality when you push into the shadows is a definite bonus. Arguably, you probably don't have to do that if you properly exposed your photo, but it might not always be that possible for example this photo from my old Canon Rebel (the grandpa of cheap DSLRs):



Obviously not everybody would need this feature, just like not everybody need 10fps or full frame, but these features open more options for the photographer to execute his/her vision.

--
The Secret to Life is... Calcium!!
Recent Gallery- http://www.pbase.com/supperman
Older Gallery- http://max-fun.fotopic.net
 
What? you don't have any banding. How can that be?

If you read this:
Amin Sabet said:
"Adding to the improved virtuosity of the images captured by the EOS
40D SLR is the camera's 14-bit Analog-to-Digital (A/D) conversion
process. Able to recognize 16,384 colors per channel (four times the
number of colors recognized by the EOS 30D SLR's 12-bit conversion
capability), the EOS 40D camera is able to produce images with finer
and more accurate gradations of tones and colors."
Click to expand...
The obvious impact of not having it means fewer gradations, which means bigger jumps between gradations, This will be perceived as banding.

In case I am not being obvious enough. I think this is pure marketing all the way. Like when the Sony Cybershots with micro sensors had 14bit ADCs, or when Pentax went to a 22bit ADC. Or even when the Canon G series went from 10bit to 12bit when those tiny sensors don't really have the S/N to support even 12bit, let alone the 14bit that Sony was using on the Cybershots.

There is a slight possibility that this will make an infinitesimal difference on 1Dmk3 with it's fat sensitive low noise pixels. None at all that it will make a difference with the 40d.

But it is certainly paying off in increased tongue wagging for Canon.
 
I think its pretty obvious that 14 bit would be very effective at fighting the color banding (posterization) that i get every time im doing heavy post processing, especially when i convert to b&w.

im looking at 14-bit as one of the most important features in the 40d

---
Rosti
http://www.photoforum.ru/11012
 
Max Fun said:
Similarly, the 14bit A/D converter might not seem to affect the image
much at first glance, but the ability to get greater tonality when
you push into the shadows is a definite bonus.
Click to expand...
Your statement above is 14-bit hype. The problem is you can't see the "greater tonality" in the shadows when the 1DmkII and 1DmkIII pictures are identical and converter tone curves normalized to produce equal intensity levels throughout the picture. Don't take my word for it, download the two picture files and see (or not as the case may be) for yourself - there is no visible difference when you compare 16-bit TIFF files.

Regards,

Joe Kurkjian, Pbase Supporter

http://www.pbase.com/jkurkjia/original



SEARCHING FOR A BETTER SELF PORTRAIT
 
You must log in or register to reply here.

Similar threads

B
New Member, Help with a DCS 14/n
Replies
2
Views
3
benhutch
B
M
The crushing reality of RAW Profiles - an introduction to Linear RAW
3 4 5
Replies
95
Views
4K
Zeee
Z
flyinglentris
Tonality in Terms of Possible Gray Scale Values (Bit Depth).
Replies
13
Views
326
57even
5
S
  • Question Question
PS: 32 bit mode for 16 bit images & RAW/smart objects question
Replies
0
Views
365
Scrollop
S
T
Myths about RAW and JPEG -- A thorough (read: long) list.
6 7 8
Replies
150
Views
9K
Bob A L
B

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top