Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Alternative Gear & Lenses | Join Upload & Sell

1       2       3      
4
       end
  

Archive 2013 · Dynamic Range vs. Tonal Range
  
 
alundeb
Offline
• • • •
Upload & Sell: Off
p.4 #1 · p.4 #1 · Dynamic Range vs. Tonal Range


zhangyue wrote:
but I am not quite sure that once we reach mid or high tonal range, the step still is defined by this noise level? We can amplify the signal, though, without change S to N ratio, but amplified signal can be divided into more bits/step to process in digital domain to achieve finer tonal response.


and

zhangyue wrote:
I didn't go over your number here but you are right that for high key image, the step level's loss can be less.


It seems that I managed to make my point understood



Apr 24, 2013 at 09:05 PM
theSuede
Offline
• • • •
Upload & Sell: Off
p.4 #2 · p.4 #2 · Dynamic Range vs. Tonal Range


RustyBug wrote:
So, does that mean for "shot noise", a longer exposure (equivalent exposure) will "even out" the randomness of the distribution somewhat ... but, it comes at the risk of incurring greater thermal noise at exaggerated shutter lengths?



Yes! -But thermal noise is only really adversely affecting the image when you go REALLY long, like 10s+. At 60s, it gets rather bothersome in many cases.

Thermal noise is almost exactly time-linear (it's simplified as time * temperature), which means in turn that if we say that it is barely visible at a 5.0s exposure, and it's 100x lower at 1/20s exposure (0.05s), it's been pushed it down into very small fractions of a single electron in effect... Which is hardly a problem.

Getting more light where you want it / need it is usually the biggest problem...



Apr 25, 2013 at 12:52 AM
kwalsh
Offline
• • •
Upload & Sell: Off
p.4 #3 · p.4 #3 · Dynamic Range vs. Tonal Range


theSuede wrote:
Thermal noise is almost exactly time-linear (it's simplified as time * temperature), which means in turn that if we say that it is barely visible at a 5.0s exposure, and it's 100x lower at 1/20s exposure (0.05s), it's been pushed it down into very small fractions of a single electron in effect... Which is hardly a problem.


I think we want to be careful here... Thermal noise may be linear in the sense of total accumulated charge. If you don't do dark frame subtraction that's all there is to it, but for long exposures you really should do dark frame subtraction and most cameras do (often call long exposure NR or something, you know it is doing it if it locks up after taking a shot for the same amount of time as the exposure). If you do dark frame subtraction then the remaining noise is increased shot noise from thermal accumulated charge in both the main exposure and the dark frame exposure. And in this mode noise would go as square root of time rather than linear.

Unless I made some sort of mistake there...

Ken



Apr 25, 2013 at 01:42 AM
theSuede
Offline
• • • •
Upload & Sell: Off
p.4 #4 · p.4 #4 · Dynamic Range vs. Tonal Range


No, you didn't (make a mistake), but the point was that it gets noticeable in real-world images only at exposure times well into what most people would call "specialty usages".

Shot noise fully dominates thermal noise (and thermal PRNU) well up into the several-second ranges, making it a problem only for those into very long exposures. Which is kind of a user specialty, demanding some additional knowledge from him/her - just like very high-magnification macro does.

Regarding the exact nature of TN, it's a sign-strapped time constant vector. Easiest way to say this is "linear", never mind telling that it's a pseudo-random occurrence that can only have a charge-increasing (no negative values) effect. I thought that that was a bit beyond this thread.

Regarding BFS, if you're serious about long time exposures you prepare several black-frames for different times/temperatures before shooting the actual target. This makes the in-camera "long time exposure NR" void, you can disable it - and this doubles your frame rate on location... You make each of the BF versions from the average of 8-12 black-frame images, meaning that the shot noise from each is averaged into the final result. This makes shot noise effect approach zero, or at least very low levels, while making the thermal noise image more accurate...



Apr 25, 2013 at 12:49 PM
kwalsh
Offline
• • •
Upload & Sell: Off
p.4 #5 · p.4 #5 · Dynamic Range vs. Tonal Range


OK, thanks for clarifying! Just wanted to make sure I was thinking clearly. And yes, I completely agree it is a non-issue until you get to quite long exposures and sorry to drag the thread a bit off topic.


Apr 25, 2013 at 01:50 PM
RustyBug
Offline
• • • • • •
Upload & Sell: On
p.4 #6 · p.4 #6 · Dynamic Range vs. Tonal Range


Nope, you didn't drag it off topic ... you took it where I was trying to discern the diff's eventually anyway. Excellent stuff guys. +1 @ "specialty uses" vs. routine stuff, etc.

This is actually a part of my reason for the OP to begin with. In routine images, it isn't much of an issue even shooting on my antiquated sensors. The world would have you believe you "must have" the latest & greatest. Not to say there aren't gains to be realized, but I just would like to better know how/where those gains will actually benefit vs. be inconsequential in practical application.

Thanks to this thread, I can now see where my most likely gains would be in "read noise" by upgrading to more efficient technologies, but shot noise isn't going to be a significant gain with regard to pixel size @ well exposed/base ISO.

I've yet to go down the BF road (on my list to do), but I'd been wondering at the real world impact it'll have. I've been trying to shoot with my SLR/C (ISO 6 capable) at extended times with "long time exposure NR" and just haven't gotten the kind of results I'd like to achieve yet. Part of me thinks I should pack my camera in dry ice first.

I figure I'm the weak link in things, at least trying to discern where those limits of gear vs. operator exist a bit. Again, very helpful stuff. If you've got any specific insight into the SLR/C @ sub-base ISO usage, I'd love to hear about that as well. Chasing K25 in digital form , but just not getting there yet. I thought the broader discussion @ DR vs. TR, SNR,CS would be more helpful to others as well, but as long as we've gone into specialty ... I'm still listening.




Apr 25, 2013 at 02:16 PM
Guari
Offline
• • •
Upload & Sell: Off
p.4 #7 · p.4 #7 · Dynamic Range vs. Tonal Range


tag for read and re-read

Edit: Very interesting discussion here.

Suede, just to confirm if I understood correctly, if long exposures are your thing, then you could shoot a series of black frames (lens cap on) for various exposure times and let them be at the computer. Then once needed, you use the black frame that has the closest time to the one shot in the field and then substract them? I understand this would not be Sstrictu as there will be differing temperatures from "lab" and "field" conditions, but it's a good aproximation though..

And without trying to go off topic, how long do you consider a safe bet with latest generation snsors (d800 and such) for long exposures without black frame substraction??

Thanks for the help and this good discussion




Apr 25, 2013 at 02:26 PM
theSuede
Offline
• • • •
Upload & Sell: Off
p.4 #8 · p.4 #8 · Dynamic Range vs. Tonal Range


Yes, the reason the camera makers even HAVE the function directly in camera is to allow users with "no specialty knowledge" to get good long-exposure results. Doing it right after the "real" image has only very small gains compared to using "pre-cooked" blackframes.

A good approximation is often good enough, since the in-camera option adds sqrt(2) noise to the image anyway as Kwalsh mentioned... Which is like doubling the ISO, but without any shutter speed gain. This is often more damaging than the small differences between do-it-now and already-done-it blackframes.

I haven't done long exposure with the D800 so far, but if it's like the D7000 (very likely) the crossover would be around 5.0s. This depends on the target of course. It's less important on flowing water scenarios than on astro-targets - since the thermal noise is easily mistaken for star detail.... This doesn't happen in the same way on more down to earth targets.



Apr 26, 2013 at 02:57 PM
 

Search in Used Dept. 



cwinte
Offline

Upload & Sell: Off
p.4 #9 · p.4 #9 · Dynamic Range vs. Tonal Range


This has been a great read, so I actually signed up to jump in the water...
My thoughts were how the issue I always hear is the visible noise we can PP treat etc in the ways we have developed but there is little said about the other, high key, end.
In particular a film user in a camera shop declaring all digital cams "could not see through the mist", which I was not sure made much sense to me initially.... Now I am unsure.

But I just had a few days in very cloudy mountains and will not get home to see what I have available in raw but certainly in jpeg quick laptop (wife's so no software on board) viewing there is not much to see in the high tones. So I am wondering if I will be able to pull out subtle wisps of shapes in & through the white misty cloud.
The barely visible is something I always have liked - less is more sometimes!

I did expose some to left to see how they compared as the Sony generally has lots in the dark and anyway my interest and composition ideas were around the higher tones and I am happy to squash the dark down.

Thoughts and comments would be most welcome.



Apr 26, 2013 at 03:35 PM
AhamB
Offline
• • • • •
Upload & Sell: Off
p.4 #10 · p.4 #10 · Dynamic Range vs. Tonal Range


@cwinte: I think you'll need to process it with curves to get more contrast in the mist to bring out details.

Negative film typically has a more gentle rolloff (a "shoulder") in the highlights compared to digital (which typically clips highlights more abruptly), but if you have captured the highlight information digitally in RAW, you should be able to process it to come out in a similar way.



Apr 26, 2013 at 03:46 PM
RustyBug
Offline
• • • • • •
Upload & Sell: On
p.4 #11 · p.4 #11 · Dynamic Range vs. Tonal Range


theSuede wrote:
Doing it right after the "real" image has only very small gains compared to using "pre-cooked" blackframes.

Would this be due to the amount of thermal noise generated by the heat has not yet dissipated. Thus doing "pre-cooked" (at intervals to allow heat dissipation) blackframes would have less thermal noise involved. In creating blackframes, would it be preferable to create them in lower ambient temperatures as well, or is that simply getting theoretically "silly" ... or do they need to be created in the same (reasonably) ambient as the image shot itself to make the subtraction process more accurate?



Apr 26, 2013 at 04:09 PM
theSuede
Offline
• • • •
Upload & Sell: Off
p.4 #12 · p.4 #12 · Dynamic Range vs. Tonal Range


In the ideal case, the camera has the exact same conditions in the "real" exposure as in the black-frame. Same temperature, same exposure time.

Unfortunately doing it "like the camera" directly after the real shot - with one BF sample - adds in the normal electronic noise from the black frame with the same strength as in the real shot - and adding two equal strength random noise sources together increases total electronic noise (that you mostly see in deep shadows) by sqrt(2). So you remove the thermal noise (with so-so accuracy) and increase overall shadow noise.

A higher accuracy blackframe (several blackframes average) removes random noise so that your real shot is undistubed by additional electronic noise, and averages several thermal noise results for a better BF accuracy.

The blackframe is pseudo-random. This means that it HAS a "perfect solution", but also that each and every BF you take will have some random deviation - noise.



Apr 26, 2013 at 04:36 PM
theSuede
Offline
• • • •
Upload & Sell: Off
p.4 #13 · p.4 #13 · Dynamic Range vs. Tonal Range


cwinte wrote:
This has been a great read, so I actually signed up to jump in the water...
My thoughts were how the issue I always hear is the visible noise we can PP treat etc in the ways we have developed but there is little said about the other, high key, end.
In particular a film user in a camera shop declaring all digital cams "could not see through the mist", which I was not sure made much sense to me initially.... Now I am unsure.

But I just had a few days in very cloudy mountains and will not get home to see
...Show more

Jpg has almost ten times lower tone resolution in the highlights, meaning that it's much harder to pull in highlight detail. You often see this in cloud formations and so on, where it's fairly easy to get posterization if you pull them down.

The expression "see through the mist" is actually a VERY good analogy for "tone resolution"... But I must say that I think (know) the salesman to be wrong. He might be correct for compact cameras with small sensors and lots of noise reduction going on - but not for larger sensor cameras. You can pull astounding amounts of image detail from digital highlights - as long as they're not clipped.



Apr 26, 2013 at 04:41 PM
RustyBug
Offline
• • • • • •
Upload & Sell: On
p.4 #14 · p.4 #14 · Dynamic Range vs. Tonal Range


Would there be any benefit from using an AC power module vs. internal DC batteries for long exposures ... with regard to thermal noise/heat generation?


Apr 26, 2013 at 06:34 PM
kwalsh
Offline
• • •
Upload & Sell: Off
p.4 #15 · p.4 #15 · Dynamic Range vs. Tonal Range


RustyBug wrote:
Would there be any benefit from using an AC power module vs. internal DC batteries for long exposures ... with regard to thermal noise/heat generation?


In theory, perhaps slightly, in practice I doubt it. The internal resistance of the battery is about 0.1 ohm. The load with the camera on would be probably at its lowest 10 ohms or so. So the battery is contributing at most 1% more heat to the system than an external source would.

The battery chemistry itself for Lithium-ion is usually exothermic during discharge so it also contributes some heat, but apparently this is lower than the Joule heating from internal resistance unless the battery is in a fault condition.



Apr 26, 2013 at 07:02 PM
RustyBug
Offline
• • • • • •
Upload & Sell: On
p.4 #16 · p.4 #16 · Dynamic Range vs. Tonal Range


Gotcha. So if I can't feel the heat, likely it isn't an issue of merit. Other reasons exist to use the AC module, but battery inducing thermal noise @ long exposure not necessarily one of them.

Shooting multiple blackframes @ extended exposures for general power consumption however might be worth going the AC route.



Apr 26, 2013 at 07:43 PM
1       2       3      
4
       end




FM Forums | Alternative Gear & Lenses | Join Upload & Sell

1       2       3      
4
       end
    
 

You are not logged in. Login or Register

Username   Password    Reset password