Ugly Hedgehog - Photography Forum
Image Averaging – Noise Reduction
Jan 3, 2021 06:14:41   #
rlv567 Loc: Philippines
 
From a paper by Spencer Cox, in Photography Life”

“Almost all photographers know about panoramas and HDRs. Most also know about focus stacking. But how often do you hear photographers talk about a fourth method of blending photos together – image averaging? Depending on what subjects you photograph, image averaging can extend your shooting capabilities significantly.

As the name implies, image averaging involves stacking multiple photos on top of each other and averaging them together. Generally, all the images in question are taken from the same camera position using identical camera settings.

The main purpose of image averaging is to reduce noise. However, it can also be used to simulate motion blur, akin to using a longer shutter speed.

One problem with a smaller camera sensor (for instance, drones) is that, even at base ISO, you may still have high levels of noise in your photo. Image averaging can be a way to simulate a lower base ISO on such cameras. One of the biggest uses of image averaging is to capture large amounts of detail in the night sky.

If you take a series of photos with the same settings, the pattern of noise generally isn’t correlated from photo to photo. When you average multiple photos together, the overly bright or dark pixels will start to balance out, reducing the total level of noise in the image.

The more photos you average, the less noise will be in your final result. Each time you double the number of photos you average, you will improve the noise levels by one stop. By averaging together four, eight, sixteen, etc. photos, you can get vast improvements in the level of noise in your photos.”

In case this is a subject in which you’re interested, here is a good description of its use and how it’s accomplished:

https://photographylife.com/image-averaging-technique

Loren – in Beautiful Baguio City

| Reply
Jan 3, 2021 06:43:27   #
ELNikkor
 
Good point! I was aware of the night sky use of this technique, but hadn't considered using it for other purposes.

| Reply
Jan 3, 2021 08:01:20   #
R.G. Loc: Scotland
 
I recently decided to experiment with that technique and took multiple exposures of several different scenes. What I forgot to do was return the exposure setting from -1.3 to zero, which resulted in all of the exposures for that experiment being under-exposed by 1.3 stops. However, despite that I didn't have any problems brightening the resulting merges in PP by 1.3 stops and also subjecting them to the usual pushing and pulling - and all of that from a crop sensor camera (D5200). The shadows were smooth and noise-free and the highlights were strong and lacking any of that wishy-washy look they can have when you brighten an image.

| Reply
 
 
Jan 3, 2021 08:19:24   #
jackm1943 Loc: Omaha, Nebraska
 
rlv567 wrote:
From a paper by Spencer Cox, in Photography Life”

“Almost all photographers know about panoramas and HDRs. Most also know about focus stacking. But how often do you hear photographers talk about a fourth method of blending photos together – image averaging? Depending on what subjects you photograph, image averaging can extend your shooting capabilities significantly.

As the name implies, image averaging involves stacking multiple photos on top of each other and averaging them together. Generally, all the images in question are taken from the same camera position using identical camera settings.

The main purpose of image averaging is to reduce noise. However, it can also be used to simulate motion blur, akin to using a longer shutter speed.

One problem with a smaller camera sensor (for instance, drones) is that, even at base ISO, you may still have high levels of noise in your photo. Image averaging can be a way to simulate a lower base ISO on such cameras. One of the biggest uses of image averaging is to capture large amounts of detail in the night sky.

If you take a series of photos with the same settings, the pattern of noise generally isn’t correlated from photo to photo. When you average multiple photos together, the overly bright or dark pixels will start to balance out, reducing the total level of noise in the image.

The more photos you average, the less noise will be in your final result. Each time you double the number of photos you average, you will improve the noise levels by one stop. By averaging together four, eight, sixteen, etc. photos, you can get vast improvements in the level of noise in your photos.”

In case this is a subject in which you’re interested, here is a good description of its use and how it’s accomplished:

https://photographylife.com/image-averaging-technique

Loren – in Beautiful Baguio City
From a paper by Spencer Cox, in Photography Life” ... (show quote)

I did some tests a year or so ago to check this out. What I found was that, yes, image stacking at high ISO's significantly reduced noise and as few as three images was all it took. However, the final image was never quite as good as one long exposure at low ISO. I've never found a practical use for this technique but it may have some in astro photography which I've never tried.

| Reply
Jan 3, 2021 10:56:28   #
hpucker99 Loc: Anchorage, Alaska
 
jackm1943 wrote:
I did some tests a year or so ago to check this out. What I found was that, yes, image stacking at high ISO's significantly reduced noise and as few as three images was all it took. However, the final image was never quite as good as one long exposure at low ISO. I've never found a practical use for this technique but it may have some in astro photography which I've never tried.


I have found the same for astronomical and other low light situations; long exposures with low ISO is better than multiple shots with high ISO and shorter shutter speeds.

| Reply
Jan 3, 2021 11:00:02   #
bleirer
 
I wasn't clear how he did the math when he told us how many stops of noise reduction you got from a given number of averaged shots. I thought I knew that it ran with the inverse square of the number of images, so 4 shots averaged would give you half the noise compared to a single image. But what did he mean when he said it would reduce the noise by a full stop?

| Reply
Jan 3, 2021 11:06:34   #
jackm1943 Loc: Omaha, Nebraska
 
hpucker99 wrote:
I have found the same for astronomical and other low light situations; long exposures with low ISO is better than multiple shots with high ISO and shorter shutter speeds.


I just remembered the possible application for which I checked it out: Shooting a stationary subject in low light while hand holding. If you hold the camera fairly still, the software will align the separate images in good fashion.

| Reply
 
 
Jan 3, 2021 13:52:58   #
R.G. Loc: Scotland
 
jackm1943 wrote:
I just remembered the possible application for which I checked it out: Shooting a stationary subject in low light while hand holding. If you hold the camera fairly still, the software will align the separate images in good fashion.


If you want a high res image you can take multi exposures hand-held, upsize them and then merge them. That will produce a high res image that will have more small detail than any of the individual exposures and it will be more resistant to pixelation when cropped or printed large. It will also have far less noise than any of the individual exposures, and if you apply any downsizing to it, that will further reduce any noise.

| Reply
Jan 4, 2021 10:18:26   #
JBRIII
 
For those interested, like everything else, there are different ways to do this, including how the data is averaged together. For example, a lot of astro stacking software allows for statistical analysis of other means to select the best of X images to stack. A lot of imaging is now done with video, so 1000's of images can be rapidly acquired. Planets as recently discussed can be affected by astmospherics, so take 1000 's and find the best 10% to average. Astro photographers also often take dark frames, etc. to characterize sensor noise, light response, not pixels, etc. Some of this is similar to noise correction in at least my Canon, takes duplicate with shutter closed and subtracts, but much is well beyond this simple approach. Special software written with all this in mind, automated the processing of all the different calibration files. Most astro photography is "low light", so much of this is probably not generally needed for non-astro type, but never know, especially for low light stuff or maybe IR or UV.

From what I have read here on this site, I would say this is an area more explored by astro photographers who have pushed the limits harder to solve such problems, including cooled film (dry ice blocks in camera), hydrogen sensitized film, pre-flashing? film, and now thermal-electrically cooled sensors, and multiple types of calibration files and slight camera movement for resolution enhancement. I spent the summer and fall getting things finally where I can try my hand at some of this, if only the sky would return, 1 evening of clear skies in last 3 weeks, 10 am now and looks like half an hour after sunset.

I have seen plane streaks in my photos averaged out in short order by averaging, but turn the stars don't change their relative positions like bugs, people, etc. Finally, special effects are often added for Jupiter at least, as it rotates fast enough to show up in photos over relatively short periods, day is 10 hrs long I believe, so edges change quickly. This also happens with the sun over a day. I tried Zerene? I believe on several solar images and sent them some photos and they (the owner?) said it appeared things had moved (correct, and my dumb). Easy to see day to day if there are any sunspots near the edge appearing or disappearing.

| Reply
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.
UglyHedgehog.com - Forum
Copyright 2011-2021 Ugly Hedgehog, Inc.