![]() |
|
|
#1 |
|
Member
Join Date: Jan 2012
Location: auburn alabama
Posts: 615
|
MY T3i gets insanely bad noise when I shoot at iso 12,800. Is there a ratio of dark frames I would need in order to remove iso 12,800 noise from my images? Or is this just to much noise?
__________________
Wes ----------- Gear: canon t3i / 600d, 18-55mm kit, tokina 11-16mm, reversed: pentax-A 28mm, sigma APO 70-300, Sigma UC 70-210mm, Carl Zeiss 8x30b |
|
|
|
| sponsored links |
|
|
#2 |
|
Member
|
it's not so much the dark frames (although it helps a little), but the light frames you need to take. With ISO 12,800...probably a lot of light frames.
Deep Sky Stacker has an example here using 32 lights and 20 of everything else. The added darks, bias and flats didn't seem to affect the noise, but did bring out the details (it looked pretty washed out without it). They also explain the purpose of each (darks, bias, and flats) in the FAQ off the main site. I'd say try 50 light frames and see how much better it looks. Maybe you can compare your images (at 100%) with the noise of the single light frame in the link. If your images at 12,800 look about the same, then maybe you can get away with 32 (if that's acceptable for you) |
|
|
|
|
|
#3 |
|
Cream of the Crop
|
Your sensor will produce two types of noise - random and systematic. Systematic noise is when the same pixel will give a false signal (almost) every time. Random noise is when any pixel on the sensor can give a false signal. DSS uses dark frames to identify the systematic noise - and subtract that from the signals in all of your light frames.
But it needs to be able to distinguish between random and systematic noise. It does this by combining multiple dark frames. Systematic noise will be present in the same pixels, so combining them will boost the signal at those locations. With random noise only some of the dark frames will have a false signal, so combining them will result in those signals being 'diluted' with nice dark signals. This results in a difference between the intensity of the systematic and random noise levels. Statistically, that difference should increase by the square root of the number of dark frames. So, shooting 9 dark frames will make the systematic noise 3 times 'brighter' than the random noise. Shooting 49 darks will make them 7 times brighter. The returns start diminishing fairly quickly. All you need is enough darks for the software to easily identify the systematic noise. Any more is not going to give any better results. Like ohata0 says - increasing the number of light frames will make more of a difference (although that to is subject to the square-root law of dimishing returns). |
|
|
|
| sponsored links |
|
|
#4 |
|
Member
Join Date: Jan 2012
Location: auburn alabama
Posts: 615
|
!!! Guys you are awesome this is going to help tremendously! Thanks, the mathmatical factor is exactly what I was hoping for! One more question. I have been shooting raw but i notice that most stack programs don't like RAW files so should I just go back to shooting jpeg? Also for dark frames is it ok if I switch to the smallest size jpeg file option available on my camera or do I need to maintain the same file size as my original shots?
__________________
Wes ----------- Gear: canon t3i / 600d, 18-55mm kit, tokina 11-16mm, reversed: pentax-A 28mm, sigma APO 70-300, Sigma UC 70-210mm, Carl Zeiss 8x30b |
|
|
|
|
|
#5 | |
|
Goldmember
|
Quote:
These images were shot at ISO 12,800 and then cleaned with Imagenomic Noiseware. ![]() ![]() Along with Noiseware, there are several other programs that accomplish the same task. Some of them offer free trial versions. |
|
|
|
|
|
|
#6 |
|
Member
Join Date: Jan 2012
Location: auburn alabama
Posts: 615
|
Yes I have attempted to manually remove the noise but stacking is my first priority as it seems to defeat the noise without softening the image. I have been trying to figure out how to properly stack for a while now and I think this thread may be the last bit of info I need hopefully to get good results. What camera did you use to take those basketball shots?
__________________
Wes ----------- Gear: canon t3i / 600d, 18-55mm kit, tokina 11-16mm, reversed: pentax-A 28mm, sigma APO 70-300, Sigma UC 70-210mm, Carl Zeiss 8x30b |
|
|
|
|
|
#7 |
|
Cream of the Crop
|
it really is fairly irrelevant. Shrinking an 18 megapixel image down to less than one megapixel is another good way of removing noise from any image.
|
|
|
|
|
|
#8 | |
|
Goldmember
|
Quote:
With so much unhappiness over image noise, it's good to have software solutions that ease the use of high ISO's with no extreme effort. |
|
|
|
|
|
|
#9 | |
|
Member
Join Date: Jan 2010
Posts: 770
|
Quote:
Using something like Topaz Denoise 5 and nudging the "remove blur" control up a bit to get back edges without sharpening noise can help a bit. Careful how you process the images - using a "Nearest Neighbor" type of pixel count reduction actually increases noise, as does unnecessary sharpening (get it as sharp as possible in-camera). |
|
|
|
|
|
|
#10 |
|
Member
|
Convert your RAWS to Tiffs I think is the answer to your other question... looking at the date on this thread you may have already found it... Hope it helps
__________________
https://picasaweb.google.com/1131475...eat=directlink |
|
|
|
|
|
#11 |
|
Member
Join Date: Jan 2012
Location: auburn alabama
Posts: 615
|
everything is good now, got an eq mount so no need to shoot 12,800 anymore
__________________
Wes ----------- Gear: canon t3i / 600d, 18-55mm kit, tokina 11-16mm, reversed: pentax-A 28mm, sigma APO 70-300, Sigma UC 70-210mm, Carl Zeiss 8x30b |
|
|
|
![]() |
| Thread Tools | |
| Display Modes | |
|
|
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| HIGH ISO Noise removal | mr2step | Performing Arts Talk | 13 | 31st of July 2011 (Sun) 13:56 |
| PSP x2 noise removal | Nickc84 | RAW, Post Processing and Printing | 1 | 5th of August 2010 (Thu) 17:48 |
| Noise Ninja vs. CS3's noise removal filter | GM_of_OLC | RAW, Post Processing and Printing | 12 | 22nd of July 2008 (Tue) 23:48 |
| NOISE REMOVAL | uttershutternut | General Photography Talk | 2 | 13th of January 2007 (Sat) 13:27 |
| What noise removal do you use, and why? | jaymrobinson | RAW, Post Processing and Printing | 23 | 2nd of December 2006 (Sat) 12:21 |