You are here
Home > DPR > Open-source Mac app Burst Photo promises to reduce noise significantly

Open-source Mac app Burst Photo promises to reduce noise significantly

Burst Photo is an open-source application available for free on macOS that promises to significantly reduce noise and improve overall image quality for images captured using any camera.

The app, written in Swift/SwiftUI/Metal, implements a simplified version of HDR+, which is the computational photography pipeline included in Google Pixel phones. One of the tenets of HDR+ is that combining a burst of high ISO, noisy images makes it possible to align and merge the images to produce a single composite image with an expanded dynamic range and reduced visible noise. If you’ve ever done night sky stacking, it’s a similar idea, although you can stack handheld images with Burst Photo.

As of now, Burst Photo only works with DNG file formats. However, you can use Adobe DNG Converter or another app, like Lightroom, to convert Raw (or JPEG) images to DNG. It’s a step in the workflow but not especially challenging to overcome. The app aligns and merges a burst of similar images, and the developer, Martin Marek, suggests capturing at least 5-10 sharp images ‘without significant movement between them.’ They don’t need to be captured using a tripod, and it’s okay if the subject moves slightly, but it’s important not to capture significant movement between frames or any motion blur. If you’re capturing a low-light portrait, the subject must not change their pose or expression.

Once you have your source images, you open the free app on your Mac and then drag and drop a selected burst of images. The app then aligns and merges them and doesn’t require an active internet connection to perform any processing. You can see an example below. This has been cropped to a different aspect ratio. To view the original, click here.

The final image results from a 51-image burst shot at ISO 51,200 on a Sony A7S III, which was then boosted by 1.35EV during post-processing. On the left is a single frame from the burst.

Since the app uses DNG (Raw) image files, the processing happens directly on the Raw images, even before they’re demosaiced. This allows Burst Photo to bypass the color tint that can occur with very high ISO images. The tint is often seen as green or magenta.

Of course, it’s one thing to promise results, it’s another to deliver. You can find out for yourself if Burst Photo lives up to its billing by downloading the app for free. Again, it’s an open-source app. If you’d like to learn more about the principle behind Burst Photo, you can read this Google Research paper titled ‘Burst photography for high dynamic range and low-light imaging on mobile cameras.’


Images courtesy of Burst Photo

Top