Photo Cascadia Blog
Archive for the ‘Noise Removal’ Category
While starting out in photography I owned nothing but Canon cameras. I owned the Canon 5d Mk1 and moved way through the series. It all started with the Canon 5D Mark 2, where I would religiously bracket at least three images for every scene. The results I got with my Canon when I just shot a single exposure for the scene was often not enough coverage in terms of tonal range. What was happening was that I would end up either blowing highlights or blocking the shadows.
It was the general consensus that several years ago to capture dynamic range in any photography scene you needed to take several photos. This would range from-3 under exposure to +3 overexposure. I would bracket at 1-stop exposures so that I included the wide spectrum of tonal values in the scene. At the time, several third-party applications were coming out as well as Photoshop that would be capable of merging several images into a single Image. When arrival of these applications, HDR (high dynamic range) became the big thing. It was really fascinating to capture so many images and merge all of them into a single file. The results for the time we’re Incredible and people started to compare it to medium format photography. But like a lot of things in photography, HDR went too far and quickly received a bad name.
So I chose to explore a few applications and found one called exposure fusion within a program called Photomatix Pro which merged all images together to include all the dynamic range and yet receive realistic result. I was very happy with the results once I learned to fine tune the application but nevertheless took a lot of time to post process images in HDR.
I realized that I was spending a good portion of time now post processing and lot less time in the field shooting so I was always looking for a better solution.
A few years had passed and lots of people still were involved heavily with HDR even though cameras were getting better and better. Digital cameras were now getting behind the movement of more megapixels. The change came for me when Nikon decided to put out a D800 at 36 megapixels. I waited and waited for Canon to follow suit but it never happened. At this time I chose to make the move over to Nikon from Canon because the Nikon D800 had been receiving rave reviews. For my business this was perfect. I could now make larger prints and have the option to crop within the image. This cropping would allow me to eliminate things from the image and still have enough resolution in the image to print large.
In the beginning, the transition was hard and slow moving from Canon to Nikon but eventually it was a saving grace. One of the most unexpected benefits was the dynamic range of the Nikon D800. I immediately noticed that I was capturing the whole scene in terms of tonal range in a single file. For a while I thought this might be a mistake. But exposure after exposure I was able to post process the images from the single file.
As time went on, I began to grow more confident in being able to take a single exposure. Eventually I was not even bracketing as a backup except for situations of extreme contrast. I even noticed I was able to under expose the image and then bring out the shadows in post-processing. With the Canon if I had tried to bring out the shadows I would always have noise that would show throughout the image. This was not the case with the Nikon, which was pretty amazing to see and still is nice to demonstrate to people who shoot Canon.
Shooting one single exposure allows you really to focus on the composition and light. It has also had the added benefit of really allowing me to get into the scene and not worry whether I have everything I need in terms of exposure.
I now own the Nikon D810, which is even better and such an amazing camera when it comes to controlling highlights and avoiding blocked shadows in a single exposure. I shoot freely in low light situations and don’t worry about covering the tonal range of a scene.
When it comes to the histogram on the Nikon D810 I am often asked one should look for when shooting one exposure. Through trial and error I have found excellent results with the histogram if I aim to most of the data and information just to the left of the middle, which would mean I slightly underexpose the image. I know this goes against everything we have been taught from before with exposing to the right, but the Nikon D810 is a revolution and is changing the game. Aiming to have my information on my histogram slightly to the left I do need to make sure I don’t have any clipped highlights as the Nikon is much better at shadows then it is with clipped highlights. So when focusing on a landscape scene I generally will set the exposure on my foreground and go about a half stop to one stop under while making sure I don’t blow the highlights in the sky. I always like to review my histogram after each image just to make sure all the information is present.
The Original Camera Raw Image One Stop Underexposed
Histogram With Information Underexposed 1 Stop to the left
The Long Exposure On The Water
I will be very curious to see in the upcoming months if the new series of Canon 50 Megapixel cameras will focus on dynamic range or will it just be a megapixel monster?
Needless to say I am a very happy with the results of the dynamic range of the Nikon and look forward to get out and shooting lots more now that I don’t have to bracket!
By Adrian Klein
I touched on this topic in early 2012 on this blog with the post New Photography Copyright Laws – Where Do You Stand when SOPA and PIPA were the buzz acronyms making their way through the online world. Although that buzz died down the subject of image copyright continues to be a topic that brings many passionate opinions. The vast majority of us share our work online for good or for bad. Frankly I am not sure how you would maximize your photo business starting out today without an online presence including social media. Meaning we need to be taking proactive measures to protecting our work even though the reality is if you want 100% assurance it’s never used without your knowledge you simply cannot post it online.
On this topic a few weeks back I decided to do a little searching with tools at my disposal. Most links took me to social media sites where people marked my work as a favorite or shared it with their friends or otherwise benign uses. Whether you agree with me or not I am not going to get wrapped up in every little usage like some folks do, it’s simply not worth my time. I am primarily concerned with businesses that would be using it in way that can help with their profit or cause (using on website, in publication, etc). That said it did not take too much time to uncover uses of my work that I did not agree with and was unaware of.
Looking for my images online I used the two most common solutions, Google Images and TinEye. They seem to do pretty well and comparing the two I found more links with my work on Google Images than TinEye . Likely you already know of these yet if have not used them before and you have images posted online it’s worth some time to see if what they come back with your images. I do question the relevance of Google’s “Visually Similar Images” search returns. Often it was similar in colors but nothing more.
I then reached out to the offenders found during my searches and infringing uses were removed promptly. In one case I had good dialogue with a office manager that seemed to genuinely lack an understanding on which images are okay or not okay to use online. Regardless if the company is playing ignorant or not, they are paying my invoice for what I feel is a small sum for illegal use of my image by a business.
Regardless of what I talk about in the below sections one step is register your work with the US Copyright Office (for those of you reading this in the United States). This is the legal way to show you as the copyright owner of the photographs/images. Yes I know the moment an image is captured by a photographer it’s copyright is owned by that person yet legally that is not enough for most lawyers to take a case. There is always the chance images can be used illegally before you submit if you post through social media,blogs or websites frequently and submit infrequently to the copyright office. However the cost is cheap ($35) plus it’s easy to do, and it could help you down the road. I admit to falling behind on this and the images I had issues with were more recent and I had not registered them. Updated submission is complete and in the mail.
Another solution is applying a digital watermark to an image, in essence a fingerprint. Then a system crawls the web like a spider to return where your image is found (sounds like someone talking late 90’s world wide web speak). There are a number of solutions out there for putting “invisibile” watermarks on images yet for this post I will cover Digimarc for Images (DFI) by Digimarc. I put invisible in quotes since it’s not completely invisible in all cases which I get into detail below. I went with Digimarc for a number of reasons including how long they have been a leader in this space.
As a Photoshop user (most reading this are) you will already see Digmarc as an option under Filters. The thing is that unless you subscribe to the service the tracking data associated to the watermark cannot be modified and you cannot access their search database online. Both pretty important pieces. In order to get full functionality and access to their search database you need to signup.
For the majority of us posting quality looking web sized images online is important. If it’s too soft or too sharp or too pixaleted or other image degradation issues it’s not very inviting to potential viewers. That is why it’s important to pay close attention when applying the watermark.
Digmarc has videos and FAQs to help but here my steps to add:
1) Complete all your normal processing to the point you are ready to save for web (the digital watermark should be your last step).
2) In Photoshop go to Filter –> Digimarc –> Embed Watermark
3) Here choose what settings you want (if any) besides your Digimarc ID and the strength of the watermark. I normally choose strength 3 (default)
4) Click okay and the watermark is applied to the image
Here is where you need to decide whether to make additional changes to the file before saving it. Personally there places on certain images where it’s too noticeable for my liking (I wish they could truly make it invisible). The problem areas I notice most is clear skies and dark portions of an image. I don’t like seeing it, I paint out the watermark in these spots and leave it on the rest. After doing this I do the following to test the strength of the watermark.
1) Go to Filter –> Digimarc –> Read Watermark
2) Ideally it does not erase very much leaving strength in medium to high range. If the strength moves to low then you need to decide what is more important, the watermark strength or presentation.
Here are examples of images where you can clearly see the difference with and without watermark and other images where is almost no difference to the eye.
So how is Digimarc panning out for me? That is a good question I can answer down the road. I just started using the service and they say it takes at least a month before a user can potentially start seeing results through their online reporting tool. Additionally most of my work does not have DFI watermarking, meaning images already out there won’t found. I figure if they can report even a small number of unknown uses that other free tools are not picking up it’s worth the reasonable subscription price.
Use Discount Code = Cascadia20 to get 20% off a Digimarc subscription. If you have questions once you are signed up their customer service is great. I have had several questions and all were answered in significant detail within a couple business days.
Beyond using digital watermarking there are two other things I feel everyone should continue to do. Although neither of these will stop anyone from using your image they help keep honest people honest. It’s like locking the doors to your car.
1) Metadata: Ensure you have correct copyright and contact info in the metadata of your file. I have my name, website and copyright along with title, description and keywords.
2) Visible Watermark: Although the watermark was Photoshopped or cropped in cases I discovered that led me to this blog post, I still believe a relatively non-intrusive watermark of your name or company is beneficial.
Do you agree we should proactively take steps to protect or track our images before posting online? Have you cancelled your social media accounts over fears of illegal usage? What are your experiences and thoughts on this subject?
Recently I revisited some older images from the past few years. I am sure we all have some images in our past that we wish we could have done something different. For me, many of my images were underexposed to get detail in the sky we no regard for the foreground. Things have changed significantly with my post processing and advancements in Photoshop and Camera RAW. So much has happened behind the scenes in terms of the engine and workings in Camera Raw in the last few years. In this article I am going to try to explain how I now go about removing noise from my images.
To begin with I am using Adobe Photoshop CS6 and Camera Raw 7.0. To find out what version of Camera Raw you are using check the top of the Camera Raw box where it tells you as well as the camera model. To apply the best results in the noise removal sliders you must be using the latest version 7.0. As previously mentioned, with every new Camera Raw and Photoshop comes a remarkably improved feature that makes rescuing older images possible. Without a doubt one of those features that has changes my post processing methods is the noise removal within the actual Camera Raw. Let me start off my mentioning the fact that I never throw away my images and keep the original raw images in a separate folder. I name this folder so I never lost the originals. With so much happening in terms of digital processing it is smart to keep the original files to revisit later. We have no idea of the potential of where digital processing might lead to in the future. For the very reason it is critical to me to keep everything I shoot. Many photographers have different methods and will disagree with this but this is what works for me. There are as many right and wrong answers in how to do thing. It has to be very confusing for someone just beginning digital photography. I have always encouraged my students to do what works best for them and to stay consistent with the process.
The biggest improvement in Camera Raw for me is the improved features of Noise Removal. Previous to the latest Camera Raw I used different third party noise removal programs like Noise Ninja, Neat Image, and Topaz just to name a few. They all have done a good job in the past with certain images but others not so much. So with latest Camera Raw I was curious to see the changes in the new and improved Noise Removal. When opening up older files in the new version of Camera Raw, there is a highlighted exclamation point in the bottom right corner.
Important click on the image to see the full image
If you look to the right at the sliders before updating the image you will see different options. These are the older sliders that did not do such a good job of noise removal as seen in the image. The image is especially evident in the sky.
This is Camera Raw asking you whether you would like to update the older file with the new advancements. Right away after clicking on this update button you will see a huge improvement. The noise in the sky has almost been all removed by just updating the image.
Without having to do any adjustments you will see an immediate result in tonal contrast, better colors, and most importantly a massive improvement in the noise removal before even having to do any adjustments. I will start with the process of how I implemented these changes into my post processing workflow.
As you can see from the Details Tab in the previous example the Noise Reduction works in conjunction with the Sharpening. These two are meant to work together and should be applied by adjusting both to get the best results. I always start with the Noise Reduction Sliders and get a good baseline of minimal noise before adjusting the sharpening sliders. I will talk how I go about sharpening images in a future blog. Each image is different when it comes to noise so there are no default settings that I could tell you to do, as each image would have to be evaluated separately.
Taking a closer look at the noise removal sliders we have two main categories; Luminance Noise and Color Noise. Luminance Noise is the digital equivalent of film grain. In the example below you can see the noise in the sky where the pixels of variation in the sky are evident. The image below that shows an example of color noise where this is most apparent in the most underexposed areas of the image. This is especially evident in the dark trees and water.
Once you can better identify the difference between the two then it becomes easier to use the sliders to your benefit. In most cases you will find both types of noise in the image and must be dealt with according to the appropriate slider. Before I adjust the sliders I always zoom in to 100% magnification or more to really zoom in on the noise. I always look at the shadows within the image and this is where you find most of your noise especially with underexposed images. Never adjust the sliders at less than 100% magnification. Start moving the sliders around until noise is reduced but the quality of the image is not degraded. I have found in most examples it is important to be conservative with adjustments and to leave a slight amount of noise rather then overdoing it and getting soft or blurry results. The reason you want to be careful of using too much noise removal is that image can become texture less and take on painterly characteristics. This gets away from the image looking like a photographic image. The following example shows how the water and trees have taken on a blurry look and almost distorted look. The results don’t look natural.
In the example above restraint has been shown so that a small amount of noise remains but the details in the trees and especially the water have are still intact. Remember that textures are an important aspect of a good photograph and really enhance impact and depth in the image. The noise removal tools in Camera Raw do a great job of removing noise even in higher settings. The important elements of an image such as saturation, sharpness, and hue variety are not harmed.
Once you are happy with the results of the noise removal zoom back out to reevaluate the results. This is an important step in the process. In previous version of Camera Raw with the noise removal sliders there was only the option of a luminance and color slider, which in most cases really flattened the image due to the lost detail in the image.
With Adobe Photoshop and more specifically Camera Raw 7.0 the addition of the sliders that reintroduce detail and contrast have been added. Now with the Luminance Detail and Luminance Contrast sliders as well as the Color detail slider we get excellent results.
This is the reason why I now use the noise removal sliders in Camera Raw 7.0 as my main tool for removing noise. I start with the Luminance slider and remove the noise to a satisfactory level then use the Luminance Detail and Luminance Contrast to bring back details into the image. I then deal with the color slider separately and then use the color detail slider to bring back detail in the areas lost to the color slider. As you can see in the example below the before image has the tree outline morphed and blurry almost like a painting. The after image brings back all the details to the edges of the tree while still keeping the noise removed from the sky.
Noise is always most visible in the shadows so make sure to really zoom in to take a closer look at the shadows. Once I am satisfied with the noise removal in the image I will move onto my Sharpening Details to really bring back some of the lost edges due to the removal of noise. I will talk about sharpening in a future article.
The most important aspect of the noise removal in Camera Raw is that it is non destructive versus third party noise removal applications which are destructive.
Can the new Camera Raw save every image? The answer to that is no but it does do a great job on most images.
It is still critical to get the exposure right in camera and apply the basic principles of photography exposure. Can you use Camera Raw in combination with third party noise removal programs like Noise Ninja? I find that some images are completely removed of noise in Camera Raw and other images need more help. The addition of third party applications like Noise Ninja can help an image with certain areas of noise that Camera Raw is unable to. Using both applications would ideally be your best bet in my opinion. Just like anything in photography there are many ways to achieve the same results.