jaques Posted February 6, 2014 Share #1 Posted February 6, 2014 Advertisement (gone after registration) I have asked this qustion a few times- but so far I havn't got an answer- so I will start a thread: Does 'dark frame subtraction' have to occur immediately after a shot is taken? And if so: why? It seems to me that if an option could be added- so that you could choose to apply the 'subtraction' as a bulk action- on all shots- after you are finished shooting (thereby doubling your effective long exposure shot rate over a given period)... I for one find it quite annoying to stand at night in the dark in a sudden lightning storm- for a full minute- waiting for the dark frame business to end. Especially when there are great spectacles lighting the sky- 'potential photos' escaping my grasp as I wait... leaving me only this exposure test shot as a momento of what might have been: Welcome, dear visitor! As registered member you'd see an image here… Simply register for free here – We are always happy to welcome new members! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/221596-dark-frame-subtraction-a-question/?do=findComment&comment=2526980'>More sharing options...
Advertisement Posted February 6, 2014 Posted February 6, 2014 Hi jaques, Take a look here Dark Frame Subtraction- a question.... I'm sure you'll find what you were looking for!
mjh Posted February 6, 2014 Share #2 Posted February 6, 2014 Does 'dark frame subtraction' have to occur immediately after a shot is taken? And if so: why? The dark frame should have the same hot pixels as the actual shot, the likelihood of which diminishes with the time passing between the real exposure and the dark frame capture. Also, for a ‘bulk action’ to work all exposures would need to be the same length which isn’t guranteed. Link to post Share on other sites More sharing options...
jaques Posted February 6, 2014 Author Share #3 Posted February 6, 2014 The dark frame should have the same hot pixels as the actual shot, the likelihood of which diminishes with the time passing between the real exposure and the dark frame capture.In the real world would this matter much? How much is actually going on in this process? What would/do images look like if there was no dark frame subtraction at all? I think I have read that some DSLR's allow the user to disable this function? Could similar results to the in-camera process be achieved by software in post-processing? If this was the case I would like to have the option to turn the DFR off if I want. Also, for a ‘bulk action’ to work all exposures would need to be the same length which isn’t guranteed. In that case couldn't it be applied to one after another each according to their own? But set in motion as a bulk action. Link to post Share on other sites More sharing options...
jto555 Posted February 6, 2014 Share #4 Posted February 6, 2014 Hi Jaques, I feel your pain. I think what is happening at Leica is they have a great engineering solution to a photographic problem. The Leica engineer will look at the picture and marvel at how low they got the noise and the lack of hot pixels, while a photographer will complain about the missed photographs. Link to post Share on other sites More sharing options...
jaques Posted February 6, 2014 Author Share #5 Posted February 6, 2014 The Leica engineer will look at the picture and marvel at how low they got the noise and the lack of hot pixels, while a photographer will complain about the missed photographs.this is exactly my fear. That they impose limitations on the camera in search of a certain high image IQ standard- at the expense of greater functionality. I am concerned that this is also the cause of the 60 second exposure limitation- rather than some technical limitation of the hardware. I would prefer to judge for myself what is acceptable IQ v Functionality- at least in some cases... One is reminded of the story of the German Tiger tank in WW2. It could beat anything on the field of battle: but it was so superbly designed, so thoroughly over-engineered, and so expensive and slow to manufacture never-mind next to impossible to service in the field: that supply never matched demand... (sounds just a little like Leica, eh? ;-). For every one that Germany produced Russia could produce 40 T-34 tanks. The US Sherman could be knocked out of action one day but be repaired on site by the next... Link to post Share on other sites More sharing options...
george + Posted February 6, 2014 Share #6 Posted February 6, 2014 I wonder if we could make it optional? User to select noise or delay. Link to post Share on other sites More sharing options...
rscheffler Posted February 6, 2014 Share #7 Posted February 6, 2014 Advertisement (gone after registration) IIRC, Canon's more recent DSLRs do this 'delay' of dark frame subtraction, giving priority to shutter release. It used to be with the original EOS-1D, that you'd sit there for x period of time like with the Leicas waiting while photo opps came and went. I don't do a lot of long exposure shooting, but the few times I shoot fireworks now, the Canons are the ones I use, because of this dark frame delay (if they even do dark frame subtraction any more...). Link to post Share on other sites More sharing options...
marknorton Posted February 6, 2014 Share #8 Posted February 6, 2014 What we don't know of course is what the frame looks like without the dark frame subtraction - you might not like the effect of all those errant pixels... Link to post Share on other sites More sharing options...
farnz Posted February 6, 2014 Share #9 Posted February 6, 2014 With my M8 I could simply power off to cancel the Dark Frame, which was very useful when inadvertently photographing the inside of my lens cap. The pictures weren't noticeably that much noisier than those where the Dark Frame had completed. Pete. Link to post Share on other sites More sharing options...
jaapv Posted February 6, 2014 Share #10 Posted February 6, 2014 You can see what it looks like, The uncorrected shot is displayed momentarily on the LCD on first playback. Link to post Share on other sites More sharing options...
CheshireCat Posted February 8, 2014 Share #11 Posted February 8, 2014 The dark frame should have the same hot pixels as the actual shot, the likelihood of which diminishes with the time passing between the real exposure and the dark frame capture. Temperature also affects the sensor, so it makes sense to capture the dark frame at the same sensor temperature, i.e. immediately after the original shot. Link to post Share on other sites More sharing options...
Tonki-M Posted February 8, 2014 Share #12 Posted February 8, 2014 I wonder if we could make it optional?User to select noise or delay. there has been a few report by M users who have experienced no dark frame subtraction. This is not a normal phenomenon, and from what i have heard, it is not pleasant! Without the dark frame subtraction, the noise is unbearably detrimental to image quality. His opinion on the matter is leave the dark frame be as it is! Link to post Share on other sites More sharing options...
Jared Posted February 8, 2014 Share #13 Posted February 8, 2014 Two things are critical for dark frame subtraction to work properly. First, the temperature of the CCD/CMOS chip must match quite closely between the dark frame and the light frame. With most chips, the amount of dark current doubles with a delta T of just six degrees Celsius or so. That's the primary reason the dark needs to be taken immediately after the light--to match temperatures. The second thing that needs to match is exposure length. In astrophotography, dark frame management becomes a bit more sophisticated. Most astrophotographers use cameras that are cooled to minimize dark current and have thermostats so dark and light frame temperatures can be matched. That way, a large number of dark frames can be taken at a time of your choice and averaged together to provide an improved "master dark" that can be applied against the light frames in a batch mode. In this scenario it is even possible to handle different exposure lengths between the darks and lights by incorporating bias frames into the workflow and "scaling" the darks. All this is total overkill for most reasonable exposure lengths, but when you start talking about integration times measured in hours this stuff matters. - Jared Link to post Share on other sites More sharing options...
farnz Posted February 8, 2014 Share #14 Posted February 8, 2014 ... With most chips, the amount of dark current doubles with a delta T of just six degrees Celsius or so. ... 3 dB per 6C: is that over the linear portion of the gradient or will it increase exponentially at some point owing to decreasing internal resistance and lead to thermal runaway? Pete. Link to post Share on other sites More sharing options...
Jared Posted February 8, 2014 Share #15 Posted February 8, 2014 3 dB per 6C: is that over the linear portion of the gradient or will it increase exponentially at some point owing to decreasing internal resistance and lead to thermal runaway? Pete. Since it is a doubling function it's not linear, but it is quite consistent in its behavior for all CCD's/CMOS detectors that I am aware of--to the extent that most manufacturers just list a single delta T doubling value without any associated graphs for variations. I have used chips between about -40C and 40C and the relationship holds up fine. I'd love to try some deeper cooling for some of my narrow band astrophotography, but most amateur CCD's are quite limited in how low they can go. My current astrophotography camera can maintain -35C year round without resorting to water cooling. Getting significantly better numbers requires liquid nitrogen and is not readily available for amateurs. - Jared Link to post Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.