«

Hi,

Can anybody explain to me if there is a connection between the dynamic
range of a digital camera/CCD-/CMOS-Sensor and the bit width of the
analog-to-digital converter - does a 12 bit converter lead to a larger
dynamic range than a 8 bit converter - and if so why?

Thanks a lot for your input!
Johannes

Johannes wrote:
> Hi,
>
> Can anybody explain to me if there is a connection between the dynamic
> range of a digital camera/CCD-/CMOS-Sensor and the bit width of the
> analog-to-digital converter - does a 12 bit converter lead to a larger
> dynamic range than a 8 bit converter - and if so why?
>
> Thanks a lot for your input!
> Johannes

You might find these articles and their references interesting reading:

http://www.clarkvision.com/imagedeta...nge/index.html

http://www.clarkvision.com/imagedeta...gnal.to.noise/

David

David J Taylor schrieb:

> You might find these articles and their references interesting reading:
> http://www.clarkvision.com/imagedeta...nge/index.html
> http://www.clarkvision.com/imagedeta...gnal.to.noise/
> David

Thanks for the reference but my english is not good enough for this
kind of technical matter. My question is, to be more precise, if I
loose dynamic range when the information containing, say, 12 stops of
range is digitized by a converter that uses only 8 bit.

Best regards!
Johannes

Johannes wrote:
> David J Taylor schrieb:
>
>> You might find these articles and their references interesting
>> reading:
>> http://www.clarkvision.com/imagedeta...nge/index.html
>> http://www.clarkvision.com/imagedeta...gnal.to.noise/
>> David
>
> Thanks for the reference but my english is not good enough for this
> kind of technical matter. My question is, to be more precise, if I
> loose dynamic range when the information containing, say, 12 stops of
> range is digitized by a converter that uses only 8 bit.
>
> Best regards!
> Johannes

Johannes,

If the digitising is only eight bits linear, then a lot of dynamic range
will be lost. As one f/stop is half the light, full scale on 8 bits is
256, 1 stop down is 128 and so forth, 2 stops is 64, so the total range is
only 8 stops. There is no tonal contrast in the lowest level, it is
either on or off, so the dynamic range may be considered as well less than
8 stops *depending how you define dynamic range).

With the JPEG images, a non-linear encoding is used so that the digital
value is more nearly the square-root of the light level: 1, sqrt (0.5),
sqrt (0.25), i.e. 255, 181, 128 etc., so that there is a greater dynamic
range possible, at the expense of a less accurate representation of a
particular light level (in the highlights).

Does that help?

Cheers,
David

David J Taylor wrote:
> Johannes wrote:
>> David J Taylor schrieb:
>>
>>> You might find these articles and their references interesting
>>> reading:
>>> http://www.clarkvision.com/imagedeta...nge/index.html
>>> http://www.clarkvision.com/imagedeta...gnal.to.noise/
>>> David
>>
>> Thanks for the reference but my english is not good enough for this
>> kind of technical matter. My question is, to be more precise, if I
>> loose dynamic range when the information containing, say, 12 stops of
>> range is digitized by a converter that uses only 8 bit.
>>
>> Best regards!
>> Johannes
>
> Johannes,
>
> If the digitising is only eight bits linear, then a lot of dynamic
> range will be lost. As one f/stop is half the light, full scale on 8
> bits is 256, 1 stop down is 128 and so forth, 2 stops is 64, so the
> total range is only 8 stops. There is no tonal contrast in the
> lowest level, it is either on or off, so the dynamic range may be
> considered as well less than 8 stops *depending how you define
> dynamic range).
> With the JPEG images, a non-linear encoding is used so that the
> digital value is more nearly the square-root of the light level: 1,
> sqrt (0.5), sqrt (0.25), i.e. 255, 181, 128 etc., so that there is a
> greater dynamic range possible, at the expense of a less accurate
> representation of a particular light level (in the highlights).
>
> Does that help?
>
> Cheers,
> David

I should also have added that the typical ADC is 12-bit accuracy, followed
by the non-linear encoding to 8-bit JPEG.

David

David J Taylor schrieb:
> Johannes,
>
> If the digitising is only eight bits linear, then a lot of dynamic range
> will be lost. As one f/stop is half the light, full scale on 8 bits is
> 256, 1 stop down is 128 and so forth, 2 stops is 64, so the total range is
> only 8 stops. There is no tonal contrast in the lowest level, it is
> either on or off, so the dynamic range may be considered as well less than
> 8 stops *depending how you define dynamic range).
>
> With the JPEG images, a non-linear encoding is used so that the digital
> value is more nearly the square-root of the light level: 1, sqrt (0.5),
> sqrt (0.25), i.e. 255, 181, 128 etc., so that there is a greater dynamic
> range possible, at the expense of a less accurate representation of a
> particular light level (in the highlights).
> Does that help?


Dear David,

As I´m more deeply involved into analog photography than into digital
I define dynamic range as the range of f-stops between the brightest
and the darkest area that still shows perceptable details. As far ist
this definition is concerned I fully understand your explanation: There
is no visible detail in the lowest level and in the highest level
(which are white and black) - that´s a true loss of valuable
information - and so the dynamic range is remarkable decreased after 8
bit conversion. That helped a lot!

Johannes

Johannes wrote:
[]
> Dear David,
>
> As I´m more deeply involved into analog photography than into digital
> I define dynamic range as the range of f-stops between the brightest
> and the darkest area that still shows perceptable details. As far ist
> this definition is concerned I fully understand your explanation:
> There is no visible detail in the lowest level and in the highest
> level (which are white and black) - that´s a true loss of valuable
> information - and so the dynamic range is remarkable decreased after 8
> bit conversion. That helped a lot!
>
> Johannes

Johannes,

I'm glad that helped. Certainly 8-bit linear conversion would produce an
image with notably degraded dynamic range, but 8-bit gamma 2.2 corrected
encoded data (such as found in JPEGs or analog/digital/analog TV) can
produce greyscale images which exceed the limits of the human eye (in
terms of dynamic range, under typical display conditions). Once you get
into colour, there is some evidence that 24-bit RGB images can be seen to
limit the displayed colour differences, but these effects are quite
subtle.

For processing, of course, you may want 12-bit raw data if you are going
to alter the levels appreciably. The very best DSLRs are now showing that
12-bit ADCs are not quite enough, and 14 bits would be required to capture
the full range available from the sensor.

Cheers,
David

David J Taylor wrote:
> Johannes,
>
> If the digitising is only eight bits linear, then a lot of dynamic range
> will be lost. As one f/stop is half the light, full scale on 8 bits is
> 256, 1 stop down is 128 and so forth, 2 stops is 64, so the total range is
> only 8 stops. There is no tonal contrast in the lowest level, it is
> either on or off, so the dynamic range may be considered as well less than
> 8 stops *depending how you define dynamic range).
>
> With the JPEG images, a non-linear encoding is used so that the digital
> value is more nearly the square-root of the light level: 1, sqrt (0.5),
> sqrt (0.25), i.e. 255, 181, 128 etc., so that there is a greater dynamic
> range possible, at the expense of a less accurate representation of a
> particular light level (in the highlights).

Even if all the encoders were linear, surely there's a simple
interplay between dynamic range and accuracy?

A single bit (potential values 0 and 1!) can encode
a massive dynamic range (e.g. 10 f-stops worth).
It just won't have much accuracy within that range

The same logic applies to 8 bit versus 12 bit.

There's no reason they shouldn't have the same dynamic range,
with different accuracy.

Of course, one could use 8bit to have the same accuracy
as 12 bit, but a smaller range.

Or any compromise between these 2 extremes.

(non-linear encoding improves perceievd accuracy,
since (IIRC) the eye is a log (ish) device,
but I believe this is a separable issue)

BugBear

Johannes wrote:
> Hi,
>
> Can anybody explain to me if there is a connection between the dynamic
> range of a digital camera/CCD-/CMOS-Sensor and the bit width of the
> analog-to-digital converter - does a 12 bit converter lead to a larger
> dynamic range than a 8 bit converter - and if so why?
>
> Thanks a lot for your input!
> Johannes

There are many factors that in total determine dynamic range. Bit
width is ONE of them. The quantization (number of bits) sets a hard
limit- the dynamic range cannot be greater than the range allowed by
the quantization steps of the number of bits.

However, signal-to-noise-ratio is another big factor. If the S/N is
less than appropriate for the quantization steps, then you will merely
be digitizing the noise in greater detail

The S/N, on the other hand, depends on the chip, the quality of the
preamp electronics and the quality of the A/D chip. It will also vary
with exposure.

Johannes wrote:
> David J Taylor schrieb:
snip
>
> Dear David,
>
> As I´m more deeply involved into analog photography than into digital
> I define dynamic range as the range of f-stops between the brightest
> and the darkest area that still shows perceptable details. As far ist
> this definition is concerned I fully understand your explanation: There
> is no visible detail in the lowest level and in the highest level
> (which are white and black) - that´s a true loss of valuable
> information - and so the dynamic range is remarkable decreased after 8
> bit conversion. That helped a lot!
>
> Johannes

There is another type of dynamic range that is sometimes called the
small-signal dynamic range. Sometimes there is a big floor to the
black level, which would seem to be a limit on DR. However, if the
noise is low, that floor can be subtracted out. What is more
bothersome is the noise in medium exposures- variations in level in
medium intensity. So one definition is the ratio of that noise in the
middle of the exposure range to the numerical value of the maximum
possible signal.