Have you ever returned from a photo shoot only to be disappointed in the results because your images lack spark? Perhaps, they had low contrast, and the color seemed dingy, washed-out and clumpy. This is often a result of digital image noise.
For those of you who are unfamiliar with what noise is – Here is a technical explanation for you.
Image noise is a random (not present in the object imaged) variation of brightness or color information in images, and is usually an aspect of electronic noise. It can be produced by the sensor and circuitry of a scanner or digital camera. Image noise can also originate in film grain and in the unavoidable shot noise of an ideal photon detector. Image noise is an undesirable by-product of image capture that adds spurious and extraneous information. (courtesy of Wikipedia)
Now let's transfer that into something we can all understand. Noise is created by the circuitry of your digital camera. That's it. That's all you need to know is that your camera is the culprit. Now, some cameras are better than others when it comes to the creation of noise, but all cameras create some image noise.
When noise becomes bothersome is when it becomes noticeable. Noise can be quite subtle and apparent only to the trained eye when viewing photographs electronically (on your computer), or, it can look like your photograph was taken in a sand storm.
Let's talk about what causes noise, how to prevent it, and what to do about it once it's present in your images.
What Causes Noise?
- The quality level of your camera
- The ISO setting on your camera
- The lighting conditions of your subject matter
- The accuracy of your exposure
The quality level of your camera contributes significantly to the noise levels produced in your images. It goes without saying that an SLR body that costs thousands of dollars will produce less noise than a point-and-shoot that cost a hundred dollars. But, even the most expensive cameras will still produce noise under certain conditions.
What Are Those Conditions?
The ISO setting on your camera is one of the primary causes of noise. In general, the higher the ISO setting, the more noise produced. Some cameras will maintain an undetectable noise level up to a certain ISO setting and from that point on the noise will increase exponentially.
The lighting conditions of your subject matter also play a part in generating noise. Modern digital cameras perform optimally under good light. The camera sensor simply cannot resolve as clear a picture under poor lighting. Camera manufacturers are constantly working to improve this. But, if your camera is even a few years old, you will see an increase in noise under poor, flat lighting.
The accuracy of your exposure will also play a large role in the generation of noise. The problem is underexposure, and the noise will rear it's ugly head when you attempt to lighten the photograph to an acceptable exposure.
Tips to Combat Noise Generation
- Buy the best camera that you can afford. Ask your salesperson about the low light capability of the camera you're considering. Read reviews online about the noise levels of that camera. Just because a camera is expensive doesn't always mean it handles noise well. A few years ago we purchased a highly touted SLR body by a major manufacturer and the noise levels of this camera were completely unacceptable. So much so that we quickly re-sold the camera and moved to a different model. Do your homework.
- When shooting your photographs keep your ISO setting as low as possible. If you're in an auto mode, be aware of what ISO the camera is using. If the camera is trying to set the ISO too high, go into manual mode and set it yourself. Here is one ISO tip that many photographers aren't aware of: keep your ISO setting to the industry standard numbers – 100, 200, 400, 800, 1600, 3200, etc. All digital cameras offer ISO settings between these numbers (such as ISO 250). But, what is really happening (with many camera models) is that the camera has electronically boosted the sensitivity of the sensor to ISO 400, then cut it back to ISO 250 using an algorithm, and this will increase the noise beyond what you might expect for that particular ISO setting.
- The lighting conditions of your photograph is (most likely) the one element you will have the least control over. Unless, you provide your own lighting. Just bear in mind that the lower the light intensity, and a flat quality of light (an extremely overcast day for example) the more noise you may have to deal with in post-production.
- Your exposure is an element of noise production that you DO have control over. Remember this! Underexposure leads to noise. If you're not good at understanding exposure, try this. Set the LCD panel on the back of your camera to show you the histogram (check your manual). When you make an exposure the histogram will show something that looks like a mountain range with a solid line below it. You do not want that mountain range to bleed past the line on the left hand side. This indicates underexposure.
No matter how much you try – You cannot expect to eliminate all noise from your photographs. It is simply a fact of life with digital photography. Do what you can to keep noise under control.
Now, Let's Talk About What to Do Once Noise is Present in Your Photographs
Let's examine this photograph as an example.
Red Shirts by Kent DuFault, on Flickr
This photograph had several issues that contributed to noise. It was taken with an older digital camera. The lighting was flat and low in contrast. The image is slightly underexposed. The ISO setting was higher than the photographer would normally like; but the situation didn't allow the photographer time to be fussy. The camera was quickly set and the picture was taken.
You will probably notice right away is that the noise isn't immediately apparent. This can fool new photographers. When viewing your photographs electronically (on a monitor), you must look at them at a magnification of 100% or more. If you don't, and there is noise, and you choose that beautiful vacation memory to make a 24″ x 36″ print for your living room wall, you will be sorrily disappointed.
Let's Look into the Shadows of this image at 300% magnification
Noise by Kent DuFault, on Flickr
There it is! Noise! Lurking in those shadows.
Now that we know it's there, what do we do about it? We turn to software solutions in post-processing. Today, we're going to talk about three software solutions. There are undoubtedly many other programs available to help you with noise; these are our favorites.
Software Solutions
- Photoshop RAW window
- NIK Define software
- Topaz DeNoise Software
We have found the Photoshop RAW window to be acceptable for minor repairs of noise. Let's take a look at the RAW window –
Noise reduction in RAW processing by Kent DuFault, on Flickr
You can find the “noise reduction” sliders under the “Detail” icon. We have found this method to be effective if you can eliminate the noise at a setting of zero to forty. If you have to go any higher than that, you're better off going to a third party software program.
We keep two software solutions in our tool chest: Define by NIK software and DeNoise by Topaz software.
Here is a look at the NIK Define software window –
NIK Define Software by Kent DuFault, on Flickr
When you look at the preview window, you will notice three square boxes. This is where the software automatically detected noise. What we like about the Define software is that 95% of the time you can just set it to auto mode and let it do it's thing. However, you can manually override the software with relative ease. We have found the NIK software to be our #1 choice when global removal of noise is required.
Here is a look at the Topaz DeNoise software window –
Topaz DeNoise Software by Kent DuFault, on Flickr
When you look at the preview window, you will notice that the Topaz software also offers an opportunity for automatic global changes. But we have found their algorithms to be a little heavy-handed for our taste. Where the DeNoise software really shines is when applying noise reduction to specific areas of your photographs.
Both of these software programs have a learning curve. But we feel that the NIK Define software has an edge over it's competitors for ease of use. If you're a beginner, we would definitely go with NIK Define software and then add the Topaz software as your skills increase.
Let's take a look at the noise now that we've processed our photograph –
Post Noise Processed by Kent DuFault, on Flickr
And there you have it! No more noise. After some other minor adjustments to exposure, contrast, color balance, and sharpness: our image is now ready to be printed, framed, and hung on our living room wall.
Red Shirts by Kent DuFault, on Flickr
16 Comments
Good tips on noise reduction.
Great article. One “noise maker” I’m curious to hear your thoughts and possible (?) solutions of, is heat/high temps (dslrs) and increased noise. The same dynamic applies to (mostly) everything in the digital world- degraded performance and increase in errors e.g.; lengthy+poorly shielded/insulated cables transferring data (both analogue and digital data.
I’d like to know if any test exist, showing differences of a photo shoot taken, for example-
on location (portrait) over the course of many hours with ambient temps in the 100F+direct sunlight vs a similar (closest settings) shoot /w same gear taken with ambient temps in the 60f range. Camera bodies such as a canon 1DX or equiv. high end shooter, would (IMO) be shielded/insulated to reduce the heat-noise causing effects, somewhat evinced by recommend operating range (temp)…but for those who have 5d mk3 “level” bodies (and Nikon equiv) would like to know your thoughts/XP on the “hot” subject.
Thanks!
Heat is a factor in the generation of noise. I do not know of any specific tests that show the difference in noise from one ambient temperature to another. It has always been my assumption that the noise generated by heat is inconsequential compared to the other factors listed in the article. I’ve done extensive work in South America and The Deep South of the United States (Temps 90 – 105 F) without seeing any noticeable increase in noise on the cameras that I own.
wtf. Once again it seems that you got tired of typing and just ended the article. The before photo looks the same as the after. Just when I think that you are going to explain how to use nik or topaz the article ends.
This was a good overview, but for someone just beginning, you need to mention that the NIK tools are plug-ins, so unless your user has Lightroom, Aperture, or Photoshop (including Elements), they might have more software to buy. I’m not sure about Topaz, I never used that. So having said that, another solution is to look at NeatImage. (Google it.) This is a great stand-alone tool that will take care of many of the classic noise problems. It’s affordable and there is light version that is free if you are simply saving jpgs. I used this as the first step in my post production workflow before I bought the NIK package. And I still use it today in some cases if I don’t plan to do further editing with Photoshop. It gives you a lot of control, but the default settings work about 90% of the time with great results. This will be WAY more cost effective for someone that just wants to eliminate some basic sky or shadow noise but is on a budget. (And really, who isn’t on a budget nowadays?)
Why not spend a couple of extra bucks for a faster lens? A good lens on a bad camera will take better pictures than a bad lens on a good camera.
Not really. Nik collection installs in your Applications folder (in Mac, don’t know about Windows). You can click on any photo, select Open with and choose Dfine2 and use to reduce noise, then save it. Same with the rest of the Nik Collection plug-ins /apps.
You can use the same noise reduction, photoshop raw methods described in the above tutorial using lightroom 4 or 5. In the Develop module, Detail tab. Its the same as the adobe raw in photoshop.
Personally, I get really tired of the photographers obsession with noise. I see so many photos with details smeared, and frankly ruined, all in the name of noise reduction.
Today’s cameras control noise very well. Unless your photographing with a cheapo compact or your iPhone, noise really is not an issues, if it’s there you’re better of leaving it along that messing with it.
My two cents.
Nice article on software noise reduction. Would like to see more tips on preventing noise in the first place – I sometimes find noise creeps in even on ISO 200, and shots often become unusable at ISO 640 or higher. Am I being too picky?
Nice article, thanks for sharing with us!!
But for me, there is an obcession about noise in the photography world. It is going far beyond the point of evaluation if the noise is taking away or not the details, the colors, the life on the photography.
And, once obcessed with it and limited with a non-state-of-the-art camera, many people are loosing the opportunity to register a really nice moment and/or place, to do only “coffee with milk” photos. It may be limiting the creativity a bit.
Just a thought!
So its not about the photographer and his choices, but more about if he or she can deal with a computer program?
I disagree, its like music today, no talent needed just computer skills.
By that logic, Ansel Adams was a bad photographer.
few things to add with this article.. if you are increasing the ISO when the light is low or u want to avoid handshake.. be sure of changing the picture more from jpeg small to big, large or RAW. it works with the combination of Mega pixel of the camera.. Obviously one can find more ISO options with higher mega pixels cameras.. e.g. if you are shooting at ISO 100 u can opt any format from RAW to jpeg small.. but when u shoot higher ISO like 1600 or 3200 change to RAW to avoid noise. The author totally forgot about cropping factors too.. if your pictures taken in RAW format you have an option of cropping smaller portion of the frame.. but even at clicked at ISO 100 but in jpeg small be careful on cropping the image or frame well..
For static scenes, such as the one used above, multiple iterations of the same scene can be stacked to reduce noise.
This is because noise is random per iteration.
Noise reduction occurs as: 1/sqrt # iterations.
For example, had the photo shown been taken four times (using a tripod) then those images could be stacked and the noise would be reduced by 1/2 or 50%.
Obviously, this solution has limited applications but can still be extremely useful for landscapes and still lifes.
People dont see noise they only see a good picture or a bad picture. Plus it’s very rare to see an image at 100%.