Page 1 of 1

How to Reduce Noise in-camera

Posted: 17.11.2006 04:33
by Ariel
This forum is titled "Technique" and "Noise Reduction", but since the forum is so small, there is nothing in here on how to reduce noise in-camera (though Helicon Filter does a good job fixing noise after the fact)! (Yes, I know this is a software website and "technique" could refer to editing technique, but I still think that photography technique should be a part of this forum, unless the moderators tell me otherwise.) I'll assume that you have manual control on your camera.

So basically, the lower ISO (film speed) one uses, the less noise (meaning the showing up of two colors which are similar in color or brightness where there should only be one solid color) there will be. In some situations, such as when there is a small amount of available light and the subject is moving, higher ISOs are needed and post-processing noise reduction is then often needed.

(Long time) exposures at night (with no flash) often produce high-noise levels, so the lowest ISO your camera has available is in order. As is usually the case with night-long-time-exposures, the subject is a landscape. If this is the case, the aperture should be small (like f/11) in order to increase DOF. Note that this low ISO/small aperture combination requires long shutter speeds.

Posted: 17.11.2006 09:21
by Dan Kozub
I would like to add that temperature of the sensor has very important influence on noise level.

As far as I remember the noise doubles every additional 10 C degrees. I've read that some professional photographers keep their cameras in the fridge before photo session (Do not do this with your camera!! It can be damaged by condensed water!!!)

Posted: 17.11.2006 23:18
by Ariel
LOL!! That's very funny.

(But just to make sure nobody is confused, temperature refers to blue/red color cast added in editing.)


EDIT: Sorry, I misunderstood while writing this post.

Posted: 18.11.2006 10:49
by Dan Kozub
Ariel, I am serious! : ) Part of the noise is caused by electrons that comes from neighbor materials powered by temperature (low-energy) fluctuations.

High level telescopes use special cooling systems to keep the temperature of the sensor as low as possible to get minimal noise.

Posted: 19.11.2006 04:16
by Ariel
:wink: You know, I thought about what you wrote and the way you wrote it, and a few hours later, I thought maybe you were serious. :|

Anyway, that's very surprising information, but I don't see how electrons cause noise when photons are what triggers the sensors.

Posted: 27.11.2006 05:36
by Visitor
Ariel, you have some pre-conceived ideas but don't let them cloud your judgement. Take a quick look into astronomy, physics, how CCD's work, etc. You'll find out what causes noise and why cool sensors do work better.

Posted: 28.11.2006 02:00
by Ariel
Thank you for your help. I'll admit I don't know everything about CCD chips, but this sounds interesting and I want to learn. I guess if I read about digital astronomy photography I'll find something that says what you are talking about, though I have seen many documents which discuss image noise, none of which mention temperature (until now, that is :wink: ). By the way, I know quiet a bit about general physics, and the only thing I can think of which you are referring to would be that the many loose electrons in metals is what causes their reflectiveness, but I don't know what that has to do with image noise. This is very interesting. Do you think you could give me a little bit more information on the actual atomic physics which leads to image noise?

Posted: 28.11.2006 20:47
by Dan Kozub
Ariel, I do not remember where I have read this exactly, you can find it in google quite easily I think.

The priciple is very simple. The energy of elections in silicon plate of CCD or CMOS matrix is proportional to the temperature. So the probability that electron can go through potential barrier increases (according to quantum physics). So electrons from neighboring CCD cells and even from the deep of the plate can get into CCD cell and increase the brightness of the cell/pixel.

That partially explains why small matrices have higher noise: the cells are closer to each other so electrons can migrate quite easily. The second reason is that cells on small matrix contains less electrons, let's say 100 instead of 1000 on APS-C matrix, so each "warm" electron produces more noise on small matrix.

This is somewhat simplified explanation : )

Posted: 28.11.2006 21:18
by Ariel
That's perfect! Now I know what you guys were talking about and why! :D
Maybe that's part of the reason why landscape photography produces clearer pictures in the winter than other times of year.

Posted: 05.02.2007 09:12
by mireille
Interesting!

I rarely have to post process noise on my winter landscape photographs and summer is very hot and damp in Quebec..and my summer shots are usually noisy!

Posted: 08.03.2008 11:38
by Adrian
you can actually reduce noise in noisey photos, by taking multiple photographs (static exactly the same), and then stacking them ontop of eachother with opacity.

this technique is applied in long exposure work commonly in amateur astrophotography, dslr/non-cooled sensor astrophotography, and quite rarely in long exposure photography.

this technique is utilized in various astrophotography processing softwares and freewares.

heres a site that explains it well, with some photographs. this is probly the best explanation ive seen for a beginner.

http://www.cambridgeincolour.com/tutori ... uction.htm

hope this helps.

adrian.

Posted: 22.04.2008 22:56
by Ariel
Helicon Filter actually recently got a feature to stack images to reduce noise.