dynamic range of ccd

[Mike Brennan]

Just to add another angle to this subject

 

apparently the larger the size of the individual pixel the greater the dynamic range.

 

Anyone who knows what they are talking about care to elaborate:)

 

Mike Brennan

[Luke Prendergast]

As is the case with higher speed film having bigger grains, a larger sensing area will be more sensitive to light, ie., will receive more photons in a given time but that does not change the dynamic range. 50 ISO stock is not necessarily going to see more into the dark and more into the light than 500 ISO and likewise 1/4" chips do not inherently have a lesser dynamic range than 2/3" of the same resolution.

[Phil Rhodes]

Hi,

 

Actually, size of photosite does slightly modify the dynamic range. One of the parameters of a CCD photosite is the full well capacity, normally stated in number of electrons (each photon motivating the generation of an electron). For some real-world numbers compare Kodak's AF-1400 CCD with sites measuring 6.8 micrometres square, with a full well capacity of 45,000 electrons. The 1000 version has a stated capacity of 630,000 electrons in its 24 micrometre sites.

 

This is modified by the noise level, which is generally some number of electrons in two digits. Clearly if you quantise the signal to any greater fineness than the full well capacity divided by the noise level, your extra resolution is just going to be noise (not that it stops you doing it anyway then putting big numbers on your spec sheets, ahem, paging camera manufacturers). Larger photosites are more likely to be impacted by non-light energy, cosmic rays, heat up, etc, and so generally have a higher noise floor resulting in overall similar dynamic range but often increased sensitivity.

 

The combination of these characteristics, and the advanced physics that goes with poking them about, determine characteristics such as dynamic range.

 

Phil

[John Pytlak RIP]

Film achieves it's great dynamic range by using a variety of grain sizes, from tiny sub-micron grains to grains as large as about 5 micrometers across (the fast yellow layer of the fastest color negative films). In many cases, a "fast", "mid", and "slow" emulsion are used, sometimes blended, sometimes coated as separate layers:

 

http://www.kodak.com/US/en/motion/support/h1/structure.shtml

 

http://www.kodak.com/US/en/motion/support/...sureP.shtml#din

[danny bartle]

Hello Windman..

Kako si?

Another Croatian here too....

[Filip Plesha]

Dobro sam Hvala :)

 

I think it wouldn't be polite to speak on our language, so I'll continue on english.

[Mike Brennan]

"50 ISO stock is not necessarily going to see more into the dark and more into the light than 500 ISO and likewise 1/4" chips do not inherently have a lesser dynamic range than 2/3" of the same resolution."

 

Yes larger pixels means more potential dynamic range. Something about binning wells and noise.......

 

 

 

Mike Brennan

[Luke Prendergast]

Yes but in the real world

Larger photosites are more likely to be impacted by non-light energy, cosmic rays, heat up, etc, and so generally have a higher noise floor resulting in overall similar dynamic range but often increased sensitivity.

- thank you Phil

[Mike Brennan]

John wrote

 

"Yes but in the real world

 

QUOTE

 

"Larger photosites are more likely to be impacted by non-light energy, cosmic rays, heat up, etc, and so generally have a higher noise floor resulting in overall similar dynamic range but often increased sensitivity."

 

- thank you Phil"

 

It was the "real world" that lead me to ask the question!! The larger format digital motion picture and stills cameras appear to have greater dynamic range than 2/3 inch!

 

This is a post from Martin Euredjian on CCow that seems to get nearer to the point of discussion. The conscept of greater dynamic range is also alluded to in various white papers from CCD manufacturers. Also born out in greater dynamic range of larger format digital stills cameras, which is what is prompting me to ask the question.

 

Martins response on Creative Cow

"A CCD sensor is a photodiode. Photodiodes convert photons into electrons. For each photon a number of electrons is released. Quantum Efficiency (QE) is a measure of how well a particular photodiode can perform this conversion. QE depends mostly on the quality of the material used to make the sensor. In addition to this, QE is measured in electrons per unit-area per second. It's a great little number because it defines most of what you want to know in order to determine exposure. For example, if a photodiode has a QE of 50 electrons (e-) per square millimeter per second it means that one photon strike will release 50e- into a square mm of silicon. I'm simplifying a little here, of course.

 

Now, as I said, CCD's are special versions of photodiodes. Instead of having one big sensor it is sliced-up into many. This is done by a clever arrangement that creates these "buckets" that accumulate all electrons released into a certain area --the pixel. It follows from the above that a larger pixel will have the ability to store more electrons.

 

Given equal exposure settings, a larger pixel (greater electron capacity) can offer greater dynamic range. I say "can" rather than "does" because this, of course, depends on whether or not the camera is designed to take advantage of this."

 

 

 

Mike Brennan