Posted on Leave a comment

How we engineer cutting-edge AI applications

Here at Topaz Labs, we train modern AI models to “auto-magically” enhance images and video, and deliver them to our customers in desktop applications that run on their own hardware.   This provides a lot of value to our customers, but raises new challenges to our engineers in ensuring reliable software delivery.

We recently released new versions of our Sharpen AI and DeNoise AI applications using a next-generation version of our AI Engine.  The AI Engine is the part of our software which takes in the AI model, and processes an image through it using a variety of hardware-optimized libraries.  This new version of the AI Engine has resulted in some incredible improvements in speed, but unfortunately the launch of the 3.0 versions of Sharpen AI and DeNoise AI was rougher than expected.

Problems

We ran into problems where after the initial release, certain combinations of operating systems, CPUs, GPUs, and drivers resulted in crashes or sub-optimal performance. Our customers posted several issues on our forums, and our support ticket volumes sky-rocketed.

Fortunately, our team rallied, and isolated a few of the major reproducible issues. And our engineers quickly diagnosed and patched numerous issues in rapid order. However, for many of our customers, the damage was already done: they had gone through an upgrade of our software, and gotten a new version that was worse than what they had had before.

The issues in our release products affected our customers across three dimensions:

The severity of the issue for a given user
The number of users affected
The length of time that a given user is affected

Each one of those dimensions has a multiplying effect on the negative impact of a given released bug, and unfortunately these releases suffered from larger than hoped-for effects across all three dimensions.

Solutions

Now that the dust is settling on those recent releases – the team met and we came up with some ways to improve the quality of our product releases in a way that also still enables us to innovate quickly on new enhancements to our products:

Better Testing

We recently had engaged a new Quality Assurance partner. But unfortunately we didn’t ask them to test across a diverse enough set of operating systems, CPUs, and GPUs. We also made large changes to our software, and the way that it is installed for our users, after the testing runs had already completed.

To address these issues, we’ve updated the testing plans to incorporate 5-times the number of tested types of machines, and instituted a stricter feature-freeze period between testing and release. This will ensure that we catch more issues earlier, and that we minimize the chances that new issues are introduced after testing.

Smaller Impact

We’ve also adjusted our release schedule, so that products with similar new functionality will go out at different times. This means that if issues with a new common piece of functionality, like this new AI Engine, are encountered, that they one affect one product at a time, and can be fixed for that one product.

Faster Fixes

One of the new features that we rolled out with the new product installers, is the ability to optimize software updates. This means that when you update a piece of Topaz software, that the installer only downloads the parts of the application which have changed since the last version.

We’ve also changed our release processes to enable more “hotfixes”, which represent a fix that the engineers feel is highly impactful for our customers, and has a low risk of introducing new problems. Close followers of past Topaz releases may have seen us do these before, we’re just further enabling the team to be able to get these fixes out faster to our customers.

Continuous Improvement through Transparency

At Topaz Labs, we feel that some of the best lessons are learned when things do not go well.  So we endeavour to foster open and honest communication internally towards learning those lessons as a team.  We also aim to keep our promise of high-quality software delivery to our customers, and be honest when we feel we have fallen short.

Hopefully others may find some of the lessons we’ve learned useful for their own projects.  We’ve certainly been encouraged that we can take steps towards even better releases in the future.

Posted on Leave a comment

How to make better holiday photos with AI-powered editing software

As we wrap-up the festive season and review the many family photos snapped throughout all of our holiday traditions, you may come across a few shots that are almost perfect.

Photographing holidays, like all live events, calls for a reportage style of photography: You’ll shoot quickly to catch the perfect moment, but sometimes the light will be a little low, or your shutter speed a bit slow, or the subject doesn’t quite fill the frame. Authentic moments unfold spontaneously, in real-time, and there are no “do-overs”. 

When you make that almost perfect shot — a great moment, with some technical flaws — all you need is a little help in post-production to make that image to shine. Topaz Labs offers a whole collection of AI-powered photo editing apps to help you rescue your photos.

Tackle low-light photos

Holiday lights look stunning in a dimly-lit room, but they are difficult to photograph. When you’re shooting in low light without a flash, you’ll struggle to shoot at a fast shutter speed unless you raise your ISO.

Higher ISOs increase the sensitivity of your camera’s sensor, so you can freeze motion even in low light. But this higher sensitivity comes at a price: noise. Noise is like a veil of static across an entire photo, muddling the details and depth of your image.

DeNoise AI uses machine learning tech to tackle heavy image noise, so your photos looks crisp and clean even at ISO 10,000 and above. It’s the perfect app to finish your nighttime holiday photos.

Sharpen blurry images

You’ve never seen a kid fly across a room faster than on Christmas morning! They’re all smiles as they tear into their gifts under the tree. These are the best moments filled with pure joy and lightning speed. But capturing those happy moments are difficult, because quick-moving subjects often leave us with a little motion blur in the final photo.

Correcting missed focus and motion blur is easy with Sharpen AI. You can choose from three AI Models Sharpen, Stabilize, or Focus to correct a soft image in one click. The results will give you more detail, for crystal-clear holiday memories. 

Make your photos perfect for printing

Sometimes you’ll shoot the perfect portrait if only your subject was filling the frame! If you need to crop-in to your photo to recompose the perfect shot, how many pixels are you losing? How much resolution do you sacrifice by cropping?

The solution is simple. Gigapixel AI makes it possible to upscale your image up to 600% with enhanced detail and sharpness. Achieve cleaner, crisper, bigger photos, ideal for print.

What's your biggest photo challenge?

Whether it’s capturing crisp photos or creating the perfect composition, we all have our own challenges with photography. What’s your biggest photo challenge? Let us know in the comments!

Posted on Leave a comment

How To Crop Your Photos and Recover 100% Resolution with Gigapixel AI

Sometimes you want to work with a portion of a photo, but "cropping in" reduces resolution. Gigapixel AI helps you crop images without compromising your pixel count.

It’s a great feeling to find the perfect image for your project, but what if you only want to work with a portion of the photo? Cropping into an image reduces the overall resolution which limits your use—especially for print, where every pixel counts. Let’s take a look at how Gigapixel AI can help you upsize your cropped photos by up to 600% and restore full-size resolution.

Crop and Resize with Gigapixel AI

Let’s start with a great image that would be made better with cropping. You can follow along yourself by downloading a totally free trial of Gigapixel AI, and downloading our free demo image.

The image we’re working with is 1920 x 1276 pixels, which is suitable for most uses but we want to crop tightly on the model’s face. 

Launch the image in Gigapixel AI, and click on the crop tool to select your crop. Apply your selection. After cropping, our image is now only 484 x 294 pixels—not much to work with here! Let’s see how Gigapixel can improve this crop.

In the main Gigapixel AI interface, we see a preview of 4X upscaling, which gives us an output of 1936 x 1176 pixels. You’ll see that the compression artifacts are gone, the model’s skin is smooth, and the details are sharp. But can we go even further?

We bump up our photo enlargement to 6X, or 600% larger than the cropped resolution. 

This gives us an output of 2904 x 1764 pixels — larger than our original uncropped image — and yet the output is still silky smooth with sharp detail in the right places. 

Gigapixel AI proves yet again to be an incredible image upscaling tool. But how does it compare to Photoshop?

Gigapixel AI vs Photoshop

 Let’s see whether Gigapixel can beat Photoshop, using the latest 2021 edition of Photoshop and its best image enlargement mode.

Using the same image and crop, we attempted 600% image upscaling in Photoshop 2021 using the Bicubic Smoother (enlargement) resampling mode.

The upscaling results in Photoshop were completely underwhelming. Compression artifacts are everywhere, and there is a lack of detail across the entire image. 

A side-by-side comparison says it all. Gigapixel drastically outperforms Photoshop even up to 6X image upscaling. 

Unlike older image editing algorithms, Gigapixel uses machine-learning technology to analyze your photo alongside millions of images to intelligently calculate the best output — this is why the results continue to shock and impress photographers all over the world. Try it totally free and see for yourself!