About Me

My photo
Spofford, New Hampshire, United States
Jeff Newcomer has been a physician practicing in New Hampshire and Vermont for over 30 years. Over that time, as a member of the Conservation Commission in his home of Chesterfield New Hampshire, he has used his photography to promote the protection and appreciation of the town's wild lands. In recent years he has been transitioning his focus from medicine to photography, writing and teaching. Jeff enjoys photographing throughout New England, but has concentrated on the Monadnock Region and southern Vermont and has had a long term artistic relationship with Mount Monadnock. He is a featured artist in a number of local galleries and his work is often seen in regional print, web publications and in business installations throughout the country. For years Jeff has published a calendar celebrating the beauty of The New England country-side in all seasons. All of the proceeds from his New England Reflections Calendar have gone to support the Pulmonary Rehabilitation Program at the Cheshire Medical Center. Jeff has a strong commitment to sharing his excitement about the special beauty of our region and publishes a weekly blog about photography in New England.

Sunday, January 29, 2017

Noise Reduction in Lightroom CC



Stonewall Farm Barn, Noise reduced, Canon G11, Small Sensor

Last week I discussed sharpening in Lightroom CC.  It seems logical that this week I travel down the “Detail” workflow to the next set of tools controlling noise reduction.  Actually, I think that Adobe should have placed noise reduction above sharpening.  In my work-flow, I typically sharpen only after I have taken a first pass at reducing noise.  The goal is to avoid allowing sharpening to accentuate the underlying image noise. Regardless of the order in the development module, lets first look at why noise forms in digital images and what can be done to reduce it before the image gets to Lightroom.


What is Digital Noise 
Noise generally appears as tiny dots of contrasting tone or color across an image area that should show smooth tones.  The pixels in a sensor are not perfect and they all emit a degree of noise along with the light signal they detect.  Several factors can make this more apparent in the final image.



Colonial Marquee, Noise removed and sharpened





Bright vs Dark Areas
Luminance & Color Noise in Shadows
Noise tends to be more obvious in darker areas of an image especially when the dark areas are lightened in post-processing.  It is part of the nature of digital sensors that the dark areas of an image are recorded with a smaller tonal range and thus a lower signal to noise ratio (S/N).  In these areas, the noise can overwhelm the recorded brightness.  The noise in the bright areas is the same, but it is overwhelmed by the strength of the actual signal. This is one reason why it is recommended that images be “exposed to the right”.  “Blowing out” of the highlight should be avoided, but, the brighter that the shadows are recorded, the better will be the signal to noise and less noise will be apparent in the vulnerable dark regions.  In the final image, the shadows can be easily darkened in post processing.

ISO

ISO 1600 Noise, Obama in Keene NH, 2007
Noise becomes more apparent as ISO is increased.  Elevated ISOs work by increasing the signal intensity, but the noise is also increased.  Many new cameras have built-in noise reduction for high ISOs, but it is always best to shoot at the lowest ISO compatible with the shooting situation.


Pixel Size and Density
Stonewall Barn,  Canon G11, 10MP Small Sensor
As a rule, larger sensors produce images with less noise.  If you plan to blow-up your image to 20”x30”, don’t shoot it with your smart phone.  This related to the size of the individual pixels and how closely they are packed on the sensor.  A 14MP sensor crammed into a smart phone must have pixels which are much smaller than those in a 14 MP sensor on a full frame DSLR.  The small pixels each “see” less light and therefore must amplify the signal – and inevitably the noise.  The image from a 21MP sensor may have higher resolution than a 12MP sensor of the same size, but the 12 MP pixels can be bigger and therefore result in less noise.

Exposure Length
One Second Exposure, Lower Purgatory Falls
Longer exposures can introduce more noise into the image.  During long exposures, the pixels heat-up which results in more noise being produced. Also, during longer exposures, there is more opportunity for random static to add noise.  The impact of long exposures varies from camera to camera and some experimenting might be necessary to discover your camera’s tolerance for prolonged shutter speeds.


Balancing Factors
Lower Purgatory Falls, Wilton, NH
 


All photography is compromise.  If you lower your ISO to reduce noise, you may need to increase the shutter speed which will have the opposite effect.  When not constrained by the need to freeze action, I generally lean toward longer exposures, since it may take a shutter speed of a minute or more before heat will build-up to the point of causing significant noise.

--------------------------------------------------------------------------------

Noise Reduction in Lightroom CC

Detail Panel


Ok, you have pulled out your full sized sensor camera, lowered your ISO and exposed to the right, now what can Lightroom do to reduce any residual noise?

The Noise Reduction Panel performs separate adjustments on two types of noise, Luminance and color.  Let’s start with the easier of the two.




Color Noise
Color & Luminance Noise
Color noise appears as spots of aberrant color seen best when zoomed in to darker areas of the image. This noise often goes unseen since Lightroom, by default, presets color noise reduction to 25.  This is often sufficient to control the color artifacts and further adjustments may be unnecessary.  If you reduce the setting to zero you can see the underlying noise and higher levels of correction may rarely be required in especially noisy images.  The other two sliders seldom require adjustment.  The Detail Slider sets the threshold for what Lightroom will consider as noise.  Generally the default of 50 works well.  The Smoothness slider helps to smooth out larger blotches of color noise.

 




Luminance Noise

Color Noise Reduced, Luminance Noise Remains
Luminance noise is more difficult, both because it is generally more prominent, and because its limitation must be balanced against a loss of image detail.  Here, as in so much of photography, only you can decide where the best balance lies.  The major adjustment occurs with the Luminance Slider.  As it is increased, the coarse appearance of the noise fads, but along with this effect comes a steady loss of image detail, to the point that the image can have an overly smooth plastic look.  It is helpful to judge this effect both zoomed in and on the full image.  Some areas of
Luminance & Color Noise Reduced
smooth tone, such as the sky or skin may benefit from great greater noise reduction, but remember Lightroom's noise reduction is a global tool.  Both noise reduction and sharpening can be applied to select areas of the image with the local adjustment tools, but the effects don’t have the fine controls available in the Detail Panel.  In Photoshop, selective masking allows better control of local detail adjustments.


Local Brush to apply Noise Reduction Selectively to the Background (Red Mask)



The other two sliders in the Luminance Panel control the amount of preservation of image detail and contrast and can be adjusted to taste.

Final Obama Image, Keene State College Rally 2007

This is just a brief review of the sophisticated noise controls in Lightroom CC.  A more detailed discussion could not replace what can be learned by simply playing with the sliders.  Try them from their lowest to maximum levels, studying their effects at various magnifications.  Have fun and remember, this is Lightroom, everything is reversible, you can’t explode your precious image.

Jeff Newcomer
partridgebrookreflections.com
603-363-8338

1 comment:

  1. Really good to see this informative post. You have described everything in detail, loved every bit of it. Thanks for sharing the article

    ReplyDelete