Skylum Updates Luminar 2018 Raw Processor/Image Editor with Artificial Intelligence-Driven AI Sky Enhancer

While it seems that most Australian professional photographers of my acquaintance depend on Adobe Bridge’s Camera Raw module in conjunction with Photoshop as their first choice for raw processing and imaged editing, followed by Adobe Lightroom in order of popularity, there are alternative products and alternative software companies.

One of the most creative is Skylum, formerly named Macphun, maker of Aurora HDR and Luminar, the first a high dynamic range image merging and editing application and the second a raw image processing and image editing application the features of which are ever-growing and unlike any other image editor in their scope and innovation. 

Luminar 2018 recently gained an artificial-intelligence driven feature, the AI Sky Enhancer, and the long-awaited major update, Luminar with Libraries, due sometime in December 2018, will add sync and batch editing, image organizing, rating, labelling and tagging to improve Luminar’s photo management and editing editing workflow.  

_dsf6264_iridientxtransformer_1920px_60pc
Skylum Luminar AI Sky Enhancer filter applied to a Fujifilm X-Pro2 raw file of a local landscape after initial processing with Iridient X-Transformer. Photograph © copyright Karin Gottschalk 2018. All rights reserved.

Ever since Apple signalled the coming end of its support for Aperture, one of the best media management, picture sorting and editing, raw processing and image editing applications ever, relied upon by photographers as well as picture editors, magazine publishers, advertising agencies, deign firms and more, photographers have been searching for a direct replacement and the available solutions have been found wanting to various degrees.

Since then I have tried a number of media management applications and modules built into raw processing and image editing software and none of them have filled the bill in exactly the way I need them to.

Skylum has verb working on a media management solution aka library for Luminar for some time and it looks like it will finally make its appearance soon.

I cannot want to try it out and with luck it will be the media manager cum image library I have been after for all these years.

Skylum Luminar 2018 AI Sky Enhancer Before and After

How Does AI Sky Enhancer Work?

skylum_luminar2018_skyenhancer_09_1920px_60pc
Sky detection & object recognition: With the help of our deep neural network, Luminar analyzes the image and detects the sky. This neural network had been trained using hundreds of thousands of images with different amounts and different types of sky, whether it’s a tiny patch of blue peeking through a skylight, a cloudy sky flaunting sunset colors, or a dark, ominous sky signaling a storm.
skylum_luminar2018_skyenhancer_08_1920px_60pc
Semantic Segmentation: Once image analysis is complete, Luminar performs what we call semantic segmentation, separating the image into different layers, based on the semantic and contextual meaning of the objects it detects. This analysis allows the most precise and intelligent sky enhancement with minimal noise, halos, and negative impact on other areas of the image. The AI detects objects in the foreground, defines edges, and analyzes the textures and tones of the sky itself. The deep neural network that powers AI Sky Enhancer had been trained using thousands of real life examples we’ve either taken ourselves or obtained from other professional photographers.
skylum_luminar2018_skyenhancer_07_1920px_60pc
Smart masking: Following sky detection and segmentation, Luminar applies an automatic mask to the sky that’s invisible to you as the user. When you move the slider to the right, you only notice how the sky becomes more and more beautiful. Just like a professional photographer, AI Sky Enhancer treats different images differently. It applies a custom set of adjustments to a sky, depending on its look. This means that a blue sky will get a treatment far different from a grey sky, and a sunset sky will be enhanced differently from a mid-morning one.

LUTs in Skylum Luminar 2018 and Aurora HDR 2019

I have been developing an approach to portrait photography based on five to seven bracket images batch processed in Iridient X-Transformer then merged in Skylum Aurora HDR 2019 with maximum image editing done in Aurora’s 16-bit colour space including applying film emulation and looks LUTs.

This has only been possible in the way I have long envisaged it since the release of Aurora HDR 2019 and its amazing realistic automatic tone-mapping, a huge evolutionary leap beyond previous versions of Aurora HDR and other HDR software I have used over the years.

The addition of the LUTs feature in both items of Skylum software is welcome as I have assembled an enormous collection of camera profile, film emulation and looks LUTs over the years and enjoy applying the film emulation LUTs in particular to portraits.

The challenge when editing with LUTs is to choose exactly the right one, or two or more of them in combination, to communicate the information and emotions I visualized for the finished image when I made the exposure.

Right now selecting that one or more perfect LUT from a big collection of them is a time-consuming process of trial and error, loading and looking, rejecting then choosing again.

Some video editing applications and colour grading applications and plug-ins have LUT library previewing capabilities that speed up the processing of choosing the best LUT for the job and I want to see the same functionality added to Skylum Luminar 2018 and Aurora HDR 2019.

In illustration, the above three portraits have been graded with three different film emulation LUTs, one from a medium-sized collection and the other two from a far bigger set of LUTs.

Choosing the looks I wanted took far too long and I skipped over thousands of other possible candidates.

Small previews of how each LUT would render the image would have sped up the process considerably.

Links