Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Mysterium-X dynamic range...

Each stop of increase represents a doubling of the amount of light the camera can resolve without clipping highlights or losing dark details in noise. If you express this as a contrast ratio 11-1/2 stops is a 2896:1 contrast ratio between white clip and noise floor. 13 stops is an 8192:1 ratio. So 1-1/2 stops is a big deal. Much more than 12%.
 
David Rasberry said:
Each stop of increase represents a doubling of the amount of light the camera can resolve without clipping highlights or losing dark details in noise. If you express this as a contrast ratio 11-1/2 stops is a 2896:1 contrast ratio between white clip and noise floor. 13 stops is an 8192:1 ratio. So 1-1/2 stops is a big deal. Much more than 12%.
Click to expand...

Exactly. If your toe EV remains the same, adding one full stop of latitude effectively doubles your DR as the amount of luminosity in your range is double what it was before. If your sensor at 100 ISO is capable of registering from 0 EV at about 0.25 foot candles and everything in between up to 13 at about 1900 foot candles, you have 13 stops of DR (simplistic example here). If you add 1.5 stops of range and your base EV remains at 0.25fc, you now have a top end of 14.5 or about 5700fc. You have effectively increased your luminosity range by 3X or 300%.

So while that is "only a 1.5 stop increase" it's still 3X the DR when viewed on a linear scale.

David Mullen ASC said:
The thing is, as a cinematographer, if I have to transpose numbers like "12%" or contrast ratios in my head, then they are not particularly useful for me. Which is why most cinematographers talk in f-stops.
Click to expand...

That's true. And I agree with Stephen saying that he would just say "it's a 1.5 stop increase". But that still does not coincide with his other statement above about a 12% increase in DR, or perhaps it does, if you consider EVERYTHING as measured by stops. But you can't do that. Stops are NOT a linear scale, so a linear percentage calculation DOES NOT apply. You can't simply divide 13 stops by 14.5 stops and say it's about 12%.
 
Last edited:
Eki Halkka said:
It's the opposite - linear gives much more precision in the highlights than in the shadows. Half of the available brightness values are spent on the one brightest stop, half of the remaining values to the next stop and so on.

There's 4096 brightness values total in 12 bit linear, assigned as follows if i've understood correctly - one stop each:

White

2048
1024
512
256
128
64
32
16
8
4
2
1

Black
Click to expand...

Wouldn't this be more effective if it could follow a "bell curve" gaussian distribution? I mean, is it that difficult to push sensitivity/precision to the midtones?

Is this inherent in all systems like this? Audio too? (actually it makes sense to me in audio.)
 
Since you guys kept going: I understand that the dynamic range of a sensor is the range between the noise floor of the system and the size of the "photon bucket". I guess what surprises, or don't understand, is why we can't create a sensor capable of reading "unlimited" values on the + side... What gives? (besides my lack of electronic engineering knowledge). Is it that the current necessary to express higher values becomes too strong for such compact and delicate electronics? Some sort of "pollution"?

Jeff Kilgroe said:
That's true. And I agree with Stephen saying that he would just say "it's a 1.5 stop increase". But that still does not coincide with his other statement above about a 12% increase in DR, or perhaps it does, if you consider EVERYTHING as measured by stops. But you can't do that. Stops are NOT a linear scale, so a linear percentage calculation DOES NOT apply. You can't simply divide 13 stops by 14.5 stops and say it's about 12%.
Click to expand...

It makes sense to talk in stops when you paint as all we see is contrast, right? But I agree with Jeff.

BTW, thanks Jeff for the graphical representation of stops.

Patrick Tresch said:
I think I'll open a light renting house... I've got some matches to rent!

Pat
Click to expand...

:claps:
 
Roberto Lequeux said:
Since you guys kept going: I understand that the dynamic range of a sensor is the range between the noise floor of the system and the size of the "photon bucket". I guess what surprises, or don't understand, is why we can't create a sensor capable of reading "unlimited" values on the + side... What gives? (besides my lack of electronic engineering knowledge). Is it that the current necessary to express higher values becomes too strong for such compact and delicate electronics? Some sort of "pollution"?



It makes sense to talk in stops when you paint as all we see is contrast, right? But I agree with Jeff.

BTW, thanks Jeff for the graphical representation of stops.



:claps:
Click to expand...

The phtons strikinghte substrate induce an electrical charge.

There is literally a finite charge photons can induce on a given material of a given size.

To use your bucket analogy, they're full.

-sc
 
Roberto Lequeux said:
...why can't we create a sensor capable of reading "unlimited" values? [...] Is it that the current necessary to express higher values becomes too strong for such compact and delicate electronics? Some sort of "pollution"?
Click to expand...

It's because photodiodes only contain a finite number of electrons. Once they are all knocked loose, that's it.

It's like a cup. It's only possible to fit so much water in it. If you put a cup out in the rain and it fills up half way, you know you have half a cup of rainfall. But if it fills up to overflowing and all the excess rain spills out, you don't know how much rainfall there was; all you know is that it was more than one cup.

Now, there are many ways to increase the dynamic range. For example, when the cup fills up with rain, you could empty it out and place the cup in the rain again. With current sensors, that takes too long, so it can't be used within a single exposure. There are lots of other methods, but none are in common consumer cameras that I know of.
 
my pixel runneth over

my pixel runneth over

Roberto Lequeux said:
I guess what surprises, or don't understand, is why we can't create a sensor capable of reading "unlimited" values on the + side... What gives?
Click to expand...

When you put too much water in a bucket, it pours out over the edge. I know on some CCD sensor designs, a very similar thing happens; if one pixel receives too much light and passes "full well" capacity, the excess photoelectrons slop over into an adjacent pixel, and this causes blooming or streaking in your image. I'm not sure if it's the same process with CMOS sensors or not.

There are CCD anti-bloom circuits that act like little rain gutters to drain away overflows without affecting neighbor pixels. There are also applications that use blooming to advantage, for example if you're trying to measure the exact brightness of a star and you already know there aren't other stars within several pixels distance, you can just add up the total charge in all neighboring pixels to get the total light accurately, even though the central pixel was badly overexposed.
 
Blah, blah, blah... the question is whether or not you can DO something with this sensor. Beating the numbers to death is making me numb.

Jim
 
Daniel Browning said:
Would you describe the difference between 1/48 and 1/108 as less than 12% increase in shutter speed? How about ISO 500 to ISO 1125? T/2 to T/1.2? Those are all only 1.5 stop increases like 13 -> 14.5 stops. It is a 12% increase in the number of stops, but the increase in Dynamic Range is far greater than that.
Click to expand...

Yes. Stops are exponentials. 3 stops = 2x2x2 4 stops = 2x2x2x2, etc.
 
Sarah C. said:
Alrighty then! As one of the women that visits this fine web house, I suggest we all stop pecking at the keyboard and go out and shoot stuff! Lord knows our best work has to be earned, not theorized in a forum!

hugs,
Sarah
Click to expand...

I for one feel that learning and reading and understanding is also important - as important as hands-on experience. You need both.
 
Jannard said:
Blah, blah, blah... the question is whether or not you can DO something with this sensor. Beating the numbers to death is making me numb.

Jim
Click to expand...

lipsrsealed.gif


This is me taking a break from bleeding bullets to do that Jim.
Bud if you want I'll go hang out somewhere else.
I know I ask a lot of questions. : )
 
Roberto Lequeux said:
I see. Thanks. What is the instantaneous dynamic range of human sight?
Click to expand...

You may find this interesting, though not of much practical value for filmmaking:
http://www.clarkvision.com/imagedetail/eye-resolution.html

This may be more interesting:
http://www.luminous-landscape.com/columns/eye-camera.shtml (Looking at it, though, I think when he says that sensors have more dynamic range than film, he must be referring to color slide film.)

That Wikipedia article that Cail posted a link to is interesting:
The retina has a static contrast ratio of around 100:1 (about 6 1/2 f-stops). As soon as the eye moves (saccades) it re-adjusts its exposure both chemically and geometrically by adjusting the iris which regulates the size of the pupil. Initial dark adaptation takes place in approximately four seconds of profound, uninterrupted darkness; full adaptation through adjustments in retinal chemistry (the Purkinje effect) are mostly complete in thirty minutes. Hence, a dynamic contrast ratio of about 1,000,000:1 (about 20 f-stops) is possible.
 
Last edited:
Roberto Lequeux said:
lipsrsealed.gif


This is me taking a break from bleeding bullets to do that Jim.
Bud if you want I'll go hang out somewhere else.
I know I ask a lot of questions. : )
Click to expand...

Maybe I should have put one of these at the end?

:-)

Jim
 
You must log in or register to reply here.
Back
Top