Posted on Leave a comment

Topaz Tutorial: Editing Cityscapes with Miroslav Petrasko

When editing cityscape photos, Topaz software can be a big help. There are many sharp edges and details in cityscapes you’ll want to preserve. To achieve this, you can get a lot of help from Topaz DeNoise AI and Topaz Gigapixel AI. Both are useful with reducing noise while preserving sharpness and adding clarity. And today I will show you how.

Cityscape Fireworks

Let’s look at one of my cityscape photos with fireworks. When I capture photos, I prefer to do multiple exposures, so I can blend them later in post-processing. Unfortunately, this does not work that well with fireworks. So, there are two ways I can approach it: I either take separate photos of the scenery before the fireworks and blend them in, or I just work with a single exposure and try to make the best of it. Over the years, I’ve settled on the second option.

Working with only one exposure creates some other issues though, namely noise. Since you can’t predict the brightness an explosion will create in a scene, you will end up with a lot of underexposed photos. You don’t want to overexpose, since you can’t easily fix overexposed areas. What I like to do is double-process the photo: I create a copy of its RAW file and process it once for the highlights and once for the shadows. Then, I put them back together in Photoshop.

Let’s look at a photo to understand it better.

Fireworks in Budapest

Base RAW Image (Click to view at 100%)

This fireworks photo was taken in Budapest, Hungary, during the St. Stephens celebrations there. It’s the biggest holiday in the country and always ends with huge fireworks over the Danube River.

This is the base RAW image I captured. As you can see, while the fireworks look a bit overexposed, the foreground feels dark. So, let’s break it into two files and edit them both.

(Click to view at 100%)

The photo above is edited for highlights, where I toned down the bright areas a bit.

The photo below is edited for shadows, where I opened the dark areas a lot. 

(Click to view at 100%)

Now I can put them into layers in Photoshop and using luminosity selections, I select the shadow areas and paint in the brighter version.

(Click to view at 100%)

While I won’t focus on this technique here, you can find a detailed description on my blog.

I also did a few tweaks to open the shadows even more. One thing I like to do in photos like this one is to brighten the lightest areas of the fireworks, to make them stand out even more.

Now it’s time to fix a few issues that were introduced with this post-processing. Since we had to brighten the shadow areas, they now have much more noise than the rest of the photo. This can be fixed by using Topaz DeNoise AI. We can either use it on the whole photo or just on the parts where it’s mostly visible.

Noise Reduction with Topaz DeNoise AI (Click to view at 100%)

Let’s open the image in the Topaz DeNoise AI plug-in. The automatic processing worked quite well here, but let’s also manually move the Remove Noise slider to 0.25 to get rid of a bit more. I like it when the clouds and smoke in the sky feel soft, so I want to remove the graininess that was created during editing.

Below is a comparison of the results of DeNoise AI in multiple areas of the photo. All examples are at 200% zoom. It looks good, so I will keep it for the whole photo as it is.

(Click to view at 100%)
(Click to view at 100%)
(Click to view at 100%)

We need to save the photo in Photoshop before moving on to the next step. Let’s save it as a TIFF file, as a copy, and without layers.

Now, let’s add some clarity in the photo. While Topaz Gigapixel AI is an image enlargement solution, and not specifically designed for this task, it can be used in this way. What we want to do is use oversampling here. Basically, we’ll enlarge the photo using Topaz Gigapixel AI, and then scale it back down in Photoshop.

I open it in Gigapixel AI and enlarge it by the 4x multiplier.

Photo enlarging in Topaz Gigapixel AI (Click to view at 100%)

The other settings can stay as they are since they will have little effect when we return the photo to its original size. But if your photo is a little out of focus, or the camera moved while you were taking it, you can try and use a higher setting for the Remove Blur option. Click “Start” to begin the image enlargement processing.

Once the process ends, I open the result back into Photoshop. The original photo’s width was 5,161 pixels. The new one is 22,000 pixels (that’s the limit for a TIFF file). This can now be resized in Photoshop back to the original 5,161 pixels to get our clearer result.

Resizing in Photoshop (Click to view at 100%)

Below are a few specific areas of the photo to see it before and after. All of these are at 200% zoom. As you can easily see, the details are much better, and the overall clarity has been greatly enhanced. Since Topaz DeNoise AI also adds a bit of clarity, you can see their cumulative effect. 

(Click to view at 100%)
(Click to view at 100%)
(Click to view at 100%)

We don’t need to use this on the whole photo, with areas where there are no details — we don’t need more clarity. If we just overlay the result with the original edit, we can use masking to apply it only to the areas where it is needed.

Clean Up in Photoshop (Click to view at 100%)

We are almost done here, except for a little cleanup. The trees in the top left and the light streaks in the bottom left have to be removed, together with few dust spots. Once it’s done, we have our final, enhanced and improved result below.

Final Result (Click to view at 100%)

You can drag the interactive, white slider bar across the image to see the improvements.

Editing Workflow

This way of post-processing works on most cityscape and similar photos. Usually, with cityscape photos, you have bright areas (e.g. artificial lights, windows, sky) and many dark shadow areas. Either by splitting one RAW into multiple images, or by having taken multiple exposures, you can properly expose both. By putting them together, you will create a nice, evenly-exposed photo that you can then reduce noise with Topaz DeNoise AI and add clarity with Topaz Gigapixel AI. I prefer my photos to look perfect at any size and this workflow allows me to achieve that.

Here is one more example below, where I used Topaz DeNoise AI and Topaz Gigapixel AI. You have here a full image and then a few before/after detail shots.

(Click to view at 100%)
(Click to view at 100%)
(Click to view at 100%)

About Miroslav Petrasko

While I started as a game designer, I switched to photography around 10 years ago. Since then I have been working with various luxury travel brands and almost daily and stubbornly updating my blog at with new photos, articles, and guides. It really is not an easy task.

Let’s end with a few more cityscapes. The first one is from my hometown of Bratislava in Slovakia. The other ones are: a sunset in Paris, France, looking up under the Burj Khalifa in Dubai, UAE, then one from Nur-Sultan in Kazakhstan and lastly, the Charles Bridge in Prague, Czech Republic.

Get the Topaz Applications Featured

Topaz DeNoise AI

Shoot anywhere in any light with no reservations. Eliminate noise and recover crisp detail in your images with the first AI-powered noise reduction tool.

Topaz Gigapixel AI

Enlarging your image without losing detail has always been impossible… until now. Upscale your photos by up to 600% while perfectly preserving image quality.
Posted on Leave a comment

Sounding Off on Noise in Images

What do all the images below have in common?

Low light + High ISO  = Noise²

They all contain some amount of noise. But what causes noise and how can you avoid it?
Albert Einstein by Arthur Sasse

In a galaxy far, far, away… the (extremely) wide depth of field causes an unavoidable amount of noise.

Pillars by NASA’s Spitzer Space Telescope

A legendary slam dunk in motion leads to noisy crowds and a noisy image.

Michael Jordan by John Swart

So, what do all these images have in common? Well, besides all being very iconic?

…They all contain some level of noise.

Perfect conditions may not always exist to avoid capturing a noisy image, but here are a few pointers to help during your next shoot and post-processing.

Understanding What Causes Noise and How to Avoid It

You can generally point a finger at a high ISO setting if you’re experiencing noise in your images. High ISO is the most common contributor to image noise in photography.

The ISO setting defines how sensitive your camera sensor is to the amount of light present in your scene. It can also vary by camera model. A standard range of ISO is typically between 200 to 1600.

The lower ISOs are ideal for well-lit or sunny environments, or when your camera is stationary.  

A higher ISO setting will cause the sensor to be more sensitive, allowing you to compensate for specific scenarios, but will also create more noise.

Uses and Benefits of High ISO Settings

A higher ISO enables:

–Faster shutter speeds to freeze motion (e.g. 1600+ ISO for indoor sports photography)
–Better performance in low light (e.g. nighttime cityscapes, Milky Way shoots, dimly lit indoor shoots)
–Reduced image blur when shooting handheld


Other Factors Affecting Exposure

In conjunction with ISO, there are other factors weighing on exposure, like aperture. Think of aperture like the pupil of your camera lens.

Much like our very own pupils in our eyes and how they respond to light and dark environments, you can adjust the aperture to allow more or less light into your camera’s sensor. The smaller the aperture, the less light is allowed in and the larger the depth of field. A larger aperture allows for more light and a more shallow depth of field.
Measured in fractions, the smaller number is actually a larger opening, so f/4 is bigger than f/16, with the “f” standing for the focal length of the lens.

And finally, the last major factor in exposure is shutter speed, the duration in which the shutter is open, exposing the sensor. Measured in a fraction of a second (e.g. 1/125, 1/500), shutter speed can compensate for images that appear too light or dark at your selected aperture.

When you double the ISO (e.g. from 400 to 800), your camera only requires half as much light for the same exposure. For example, if the shutter speed is 1/250 at 400 ISO, upping to 800 ISO will produce the same exposure at 1/500 second (with a static aperture).

Working in tandem, these three settings will help you develop better images and make post-processing simpler. But for those necessary shots in less-than-perfect scenarios, we have a new solution to reduce noise and restore your photo to its intended glory!

Meet our latest AI-powered solution, DeNoise AI.


How does DeNoise AI work?

After months of training and testing, we are now able to use a bigger AI network that handles more varieties of noise.

All you have to do is import your image. DeNoise AI will target the noise in your image and intelligently apply the best solution for removing it without sacrificing too much detail loss.

Though we’ve created DeNoise AI to be an easy, automatic solution, there are three adjustment sliders that allow you to make manual edits beyond the automated detection. With the Remove Noise slider you can apply even stronger noise reduction. Pushing the sliders to the extreme level will not give you the best output all the time. If your noise level is low, then keeping the slider in the lower range should give the best output.

The Enhance Sharpness slider allows you to sharpen images that become blurry after denoising. If you’re looking to add back specific details in your image that may have been removed by denoising, you can try the Restore Detail slider to use alpha-mask blending to restore and improve details.



Where does DeNoise AI fall the best in my workflow?

We recommend using DeNoise AI at the beginning of your workflow. Eliminating noise as a first step is key because it will ensure that you are working on a clean image! It is always important to remove any damaging defects before applying any color, detail or creative adjustments to your images.


P.S. Don’t miss out on our DeNoise AI special pricing. Download a free trial, see for yourself, and then snag DeNoise AI on sale for $59.99 (typically $79.99) until May 3!

Posted on 21 Comments

The Story Behind DeNoise AI

Why did we introduce Topaz DeNoise AI?

As photographers, we all have situations where we end up with noisy photos, like when we’re shooting in low lighting or shooting fast actions.  Even for more traditional photography, we sometimes find noise in shadow areas when adjusted in post editing. Over the last decade, Topaz Labs has been dedicated to developing the best possible tools to address this problem, along with other common editing hardships.  Topaz Denoise AI is our latest effort.

As many of our users know, our existing products “Topaz DeNoise” and “AI Clear” are commonly considered the best in photo noise reduction.  So why did we produce another one? The short answer is that the new DeNoise AI is even better, both in image quality and ease of integration with your photo editing workflow. Since we’re constantly looking to provide a better customer experience, this was the logical next step in noise reduction software.

Topaz DeNoise is one of our old plugins. This program used traditional image processing techniques and has many fail cases when the noise amount and types of images vary. Later in AI Clear, we saw that AI technology based in machine learning was very effective at removing noise and  blur, while also giving us crisp and detailed images for different types of noises. AI Clear exists as an adjustment within the framework of our editing software Topaz Studio.After a time, we found we were not quite happy that we could not put our best AI network in AI Clear because of this limitation. This was for two reasons. One, the AI technology demands a lot of resources and we were limiting the resources available by using the software within Topaz Studio. Two, AI Clear had to be interactive and work seamlessly with other studio adjustments. These reasons prevented us from using the most powerful AI networks at our disposal. Therefore, we decided that we would come up with a standalone and plugin version of this software where we would not compromise on the AI itself.

In DeNoise AI, we utilized a new software architecture to accommodate our best AI network. Albert, our CTO, dedicated all of our training machines to train different network variations. Our developer Bowen implemented all network variations in the Topaz AI inference engine and validated the quality and performance of the networks. After months of training and testing, we are now able to use a bigger network that handles more varieties of noise. DeNoise AI is trained to remove high ISO noise, sensor noise, thermal noise, banding noise, and noise from some scanned images. You can also provide hints about the type of noise through two sliders if not satisfied with the default output to improve the processing. Let’s look at some examples below,

The original image on the left has very strong noise and banding artifacts. The DeNoise output is clean and some details are generated in the very dark noise area.

This image has very strong banding artifacts in different levels. The AI removed those banding artifacts and generated some convincing details.  

Finally, check out two more examples of high ISO noise and the output from DeNoise AI in default settings.

Now the question is, in terms of noise and artifact removal, how can you get the most out of DeNoise AI? To do that, I’ll explain what is going on behind the scenes. In order to train DeNoise AI effectively, we used a lot of clean and noisy image pairs as examples to train an AI network. This is so that later when the AI sees a noisy image, it can predict the clean version of it. Mathematically, we mimic the way that real-world noise is added to the image, during capture or digitization. Our noise model generated millions of noisy-clean image pairs, which were used to train the network. The DeNoise AI network is also able to receive “hints” about the noise level and type from outside as a help.

There are three sliders you can also control from the interface if you don’t like the automated detection. Pushing the sliders to the extreme level will not give you the best output all the time. If your noise level is low then keeping the slider in the lower range should give the best output. With DeNoise AI, you now you have more fine control on the output through the slider rather than just a few steps.

So, is DeNoise AI better than AI Clear? What we can tell you is that they are totally different types of AI network. Personally, I like to play with the Remove Noise and Enhance Sharpness sliders and see the outcome in different images. In our experience, DeNoise AI will typically provide a far better output than AI Clear. But you never know! We encourage you to give both a try.

If you have already played with the trial version or purchased your own copy, let us know what you think. We’re eager to hear how DeNoise AI works for you!